agora inbox for [email protected]  

help / color / mirror / Atom feed

[PATCH v8 2/8] explain to show tuplehash bucket and memory stats..
265+ messages / 2 participants
[nested] [flat]

* [PATCH v8 2/8] explain to show tuplehash bucket and memory stats..
@ 2020-01-01 00:49 Justin Pryzby <[email protected]>
  0 siblings, 0 replies; 265+ messages in thread

From: Justin Pryzby @ 2020-01-01 00:49 UTC (permalink / raw)

Note that hashed SubPlan and recursiveUnion aren't affected in explain output,
probably since hashtables aren't allocated at that point.

Discussion: https://www.postgresql.org/message-id/flat/[email protected]
---
 .../postgres_fdw/expected/postgres_fdw.out    |  56 ++++--
 src/backend/commands/explain.c                | 168 ++++++++++++++----
 src/backend/executor/execGrouping.c           |  33 ++++
 src/backend/executor/nodeAgg.c                |  15 +-
 src/backend/executor/nodeRecursiveunion.c     |   3 +
 src/backend/executor/nodeSetOp.c              |   1 +
 src/backend/executor/nodeSubplan.c            |   3 +
 src/include/executor/executor.h               |   1 +
 src/include/nodes/execnodes.h                 |  10 +-
 9 files changed, 227 insertions(+), 63 deletions(-)

diff --git a/contrib/postgres_fdw/expected/postgres_fdw.out b/contrib/postgres_fdw/expected/postgres_fdw.out
index 62c2697920..2ddae83178 100644
--- a/contrib/postgres_fdw/expected/postgres_fdw.out
+++ b/contrib/postgres_fdw/expected/postgres_fdw.out
@@ -2086,9 +2086,11 @@ SELECT t1c1, avg(t1c1 + t2c1) FROM (SELECT t1.c1, t2.c1 FROM ft1 t1 JOIN ft2 t2
          ->  HashAggregate
                Output: t1.c1, avg((t1.c1 + t2.c1))
                Group Key: t1.c1
+               Buckets: 256
                ->  HashAggregate
                      Output: t1.c1, t2.c1
                      Group Key: t1.c1, t2.c1
+                     Buckets: 4096
                      ->  Append
                            ->  Foreign Scan
                                  Output: t1.c1, t2.c1
@@ -2098,7 +2100,7 @@ SELECT t1c1, avg(t1c1 + t2c1) FROM (SELECT t1.c1, t2.c1 FROM ft1 t1 JOIN ft2 t2
                                  Output: t1_1.c1, t2_1.c1
                                  Relations: (public.ft1 t1_1) INNER JOIN (public.ft2 t2_1)
                                  Remote SQL: SELECT r1."C 1", r2."C 1" FROM ("S 1"."T 1" r1 INNER JOIN "S 1"."T 1" r2 ON (((r1."C 1" = r2."C 1"))))
-(20 rows)
+(22 rows)
 
 SELECT t1c1, avg(t1c1 + t2c1) FROM (SELECT t1.c1, t2.c1 FROM ft1 t1 JOIN ft2 t2 ON (t1.c1 = t2.c1) UNION SELECT t1.c1, t2.c1 FROM ft1 t1 JOIN ft2 t2 ON (t1.c1 = t2.c1)) AS t (t1c1, t2c1) GROUP BY t1c1 ORDER BY t1c1 OFFSET 100 LIMIT 10;
  t1c1 |         avg          
@@ -2129,11 +2131,12 @@ SELECT t1."C 1" FROM "S 1"."T 1" t1, LATERAL (SELECT DISTINCT t2.c1, t3.c1 FROM
          ->  HashAggregate
                Output: t2.c1, t3.c1
                Group Key: t2.c1, t3.c1
+               Buckets: 2
                ->  Foreign Scan
                      Output: t2.c1, t3.c1
                      Relations: (public.ft1 t2) INNER JOIN (public.ft2 t3)
                      Remote SQL: SELECT r1."C 1", r2."C 1" FROM ("S 1"."T 1" r1 INNER JOIN "S 1"."T 1" r2 ON (((r1."C 1" = r2."C 1")) AND ((r1.c2 = $1::integer))))
-(13 rows)
+(14 rows)
 
 SELECT t1."C 1" FROM "S 1"."T 1" t1, LATERAL (SELECT DISTINCT t2.c1, t3.c1 FROM ft1 t2, ft2 t3 WHERE t2.c1 = t3.c1 AND t2.c2 = t1.c2) q ORDER BY t1."C 1" OFFSET 10 LIMIT 10;
  C 1 
@@ -2610,10 +2613,11 @@ select c2 * (random() <= 1)::int as c2 from ft2 group by c2 * (random() <= 1)::i
    ->  HashAggregate
          Output: ((c2 * ((random() <= '1'::double precision))::integer))
          Group Key: (ft2.c2 * ((random() <= '1'::double precision))::integer)
+         Buckets: 2
          ->  Foreign Scan on public.ft2
                Output: (c2 * ((random() <= '1'::double precision))::integer)
                Remote SQL: SELECT c2 FROM "S 1"."T 1"
-(9 rows)
+(10 rows)
 
 -- GROUP BY clause in various forms, cardinal, alias and constant expression
 explain (verbose, costs off)
@@ -2713,11 +2717,12 @@ select sum(c1) from ft1 group by c2 having avg(c1 * (random() <= 1)::int) > 100
    ->  HashAggregate
          Output: sum(c1), c2
          Group Key: ft1.c2
+         Buckets: 16
          Filter: (avg((ft1.c1 * ((random() <= '1'::double precision))::integer)) > '100'::numeric)
          ->  Foreign Scan on public.ft1
                Output: c1, c2
                Remote SQL: SELECT "C 1", c2 FROM "S 1"."T 1"
-(10 rows)
+(11 rows)
 
 -- Remote aggregate in combination with a local Param (for the output
 -- of an initplan) can be trouble, per bug #15781
@@ -2963,10 +2968,11 @@ select sum(c1) filter (where (c1 / c1) * random() <= 1) from ft1 group by c2 ord
    ->  HashAggregate
          Output: sum(c1) FILTER (WHERE ((((c1 / c1))::double precision * random()) <= '1'::double precision)), c2
          Group Key: ft1.c2
+         Buckets: 16
          ->  Foreign Scan on public.ft1
                Output: c1, c2
                Remote SQL: SELECT "C 1", c2 FROM "S 1"."T 1"
-(9 rows)
+(10 rows)
 
 explain (verbose, costs off)
 select sum(c2) filter (where c2 in (select c2 from ft1 where c2 < 5)) from ft1;
@@ -3229,6 +3235,7 @@ select count(*), x.b from ft1, (select c2 a, sum(c1) b from ft1 group by c2) x w
    ->  HashAggregate
          Output: count(*), x.b
          Group Key: x.b
+         Buckets: 16
          ->  Hash Join
                Output: x.b
                Inner Unique: true
@@ -3244,7 +3251,7 @@ select count(*), x.b from ft1, (select c2 a, sum(c1) b from ft1 group by c2) x w
                                  Output: ft1_1.c2, (sum(ft1_1.c1))
                                  Relations: Aggregate on (public.ft1 ft1_1)
                                  Remote SQL: SELECT c2, sum("C 1") FROM "S 1"."T 1" GROUP BY 1
-(21 rows)
+(22 rows)
 
 select count(*), x.b from ft1, (select c2 a, sum(c1) b from ft1 group by c2) x where ft1.c2 = x.a group by x.b order by 1, 2;
  count |   b   
@@ -3449,11 +3456,12 @@ select c2, sum(c1) from ft1 where c2 < 3 group by rollup(c2) order by 1 nulls la
    ->  MixedAggregate
          Output: c2, sum(c1)
          Hash Key: ft1.c2
+         Buckets: 16
          Group Key: ()
          ->  Foreign Scan on public.ft1
                Output: c2, c1
                Remote SQL: SELECT "C 1", c2 FROM "S 1"."T 1" WHERE ((c2 < 3))
-(10 rows)
+(11 rows)
 
 select c2, sum(c1) from ft1 where c2 < 3 group by rollup(c2) order by 1 nulls last;
  c2 |  sum   
@@ -3474,11 +3482,12 @@ select c2, sum(c1) from ft1 where c2 < 3 group by cube(c2) order by 1 nulls last
    ->  MixedAggregate
          Output: c2, sum(c1)
          Hash Key: ft1.c2
+         Buckets: 16
          Group Key: ()
          ->  Foreign Scan on public.ft1
                Output: c2, c1
                Remote SQL: SELECT "C 1", c2 FROM "S 1"."T 1" WHERE ((c2 < 3))
-(10 rows)
+(11 rows)
 
 select c2, sum(c1) from ft1 where c2 < 3 group by cube(c2) order by 1 nulls last;
  c2 |  sum   
@@ -3499,11 +3508,13 @@ select c2, c6, sum(c1) from ft1 where c2 < 3 group by grouping sets(c2, c6) orde
    ->  HashAggregate
          Output: c2, c6, sum(c1)
          Hash Key: ft1.c2
+         Buckets: 16
          Hash Key: ft1.c6
+         Buckets: 16
          ->  Foreign Scan on public.ft1
                Output: c2, c6, c1
                Remote SQL: SELECT "C 1", c2, c6 FROM "S 1"."T 1" WHERE ((c2 < 3))
-(10 rows)
+(12 rows)
 
 select c2, c6, sum(c1) from ft1 where c2 < 3 group by grouping sets(c2, c6) order by 1 nulls last, 2 nulls last;
  c2 | c6 |  sum  
@@ -3526,10 +3537,11 @@ select c2, sum(c1), grouping(c2) from ft1 where c2 < 3 group by c2 order by 1 nu
    ->  HashAggregate
          Output: c2, sum(c1), GROUPING(c2)
          Group Key: ft1.c2
+         Buckets: 16
          ->  Foreign Scan on public.ft1
                Output: c2, c1
                Remote SQL: SELECT "C 1", c2 FROM "S 1"."T 1" WHERE ((c2 < 3))
-(9 rows)
+(10 rows)
 
 select c2, sum(c1), grouping(c2) from ft1 where c2 < 3 group by c2 order by 1 nulls last;
  c2 |  sum  | grouping 
@@ -7147,13 +7159,14 @@ select * from bar where f1 in (select f1 from foo) for update;
                ->  HashAggregate
                      Output: foo.ctid, foo.f1, foo.*, foo.tableoid
                      Group Key: foo.f1
+                     Buckets: 256
                      ->  Append
                            ->  Seq Scan on public.foo foo_1
                                  Output: foo_1.ctid, foo_1.f1, foo_1.*, foo_1.tableoid
                            ->  Foreign Scan on public.foo2 foo_2
                                  Output: foo_2.ctid, foo_2.f1, foo_2.*, foo_2.tableoid
                                  Remote SQL: SELECT f1, f2, f3, ctid FROM public.loct1
-(23 rows)
+(24 rows)
 
 select * from bar where f1 in (select f1 from foo) for update;
  f1 | f2 
@@ -7185,13 +7198,14 @@ select * from bar where f1 in (select f1 from foo) for share;
                ->  HashAggregate
                      Output: foo.ctid, foo.f1, foo.*, foo.tableoid
                      Group Key: foo.f1
+                     Buckets: 256
                      ->  Append
                            ->  Seq Scan on public.foo foo_1
                                  Output: foo_1.ctid, foo_1.f1, foo_1.*, foo_1.tableoid
                            ->  Foreign Scan on public.foo2 foo_2
                                  Output: foo_2.ctid, foo_2.f1, foo_2.*, foo_2.tableoid
                                  Remote SQL: SELECT f1, f2, f3, ctid FROM public.loct1
-(23 rows)
+(24 rows)
 
 select * from bar where f1 in (select f1 from foo) for share;
  f1 | f2 
@@ -7222,6 +7236,7 @@ update bar set f2 = f2 + 100 where f1 in (select f1 from foo);
                ->  HashAggregate
                      Output: foo.ctid, foo.f1, foo.*, foo.tableoid
                      Group Key: foo.f1
+                     Buckets: 256
                      ->  Append
                            ->  Seq Scan on public.foo foo_1
                                  Output: foo_1.ctid, foo_1.f1, foo_1.*, foo_1.tableoid
@@ -7240,13 +7255,14 @@ update bar set f2 = f2 + 100 where f1 in (select f1 from foo);
                ->  HashAggregate
                      Output: foo.ctid, foo.f1, foo.*, foo.tableoid
                      Group Key: foo.f1
+                     Buckets: 256
                      ->  Append
                            ->  Seq Scan on public.foo foo_1
                                  Output: foo_1.ctid, foo_1.f1, foo_1.*, foo_1.tableoid
                            ->  Foreign Scan on public.foo2 foo_2
                                  Output: foo_2.ctid, foo_2.f1, foo_2.*, foo_2.tableoid
                                  Remote SQL: SELECT f1, f2, f3, ctid FROM public.loct1
-(39 rows)
+(41 rows)
 
 update bar set f2 = f2 + 100 where f1 in (select f1 from foo);
 select tableoid::regclass, * from bar order by 1,2;
@@ -8751,12 +8767,13 @@ SELECT a, sum(b), min(b), count(*) FROM pagg_tab GROUP BY a HAVING avg(b) < 22 O
    Sort Key: pagg_tab.a
    ->  HashAggregate
          Group Key: pagg_tab.a
+         Buckets: 64
          Filter: (avg(pagg_tab.b) < '22'::numeric)
          ->  Append
                ->  Foreign Scan on fpagg_tab_p1 pagg_tab_1
                ->  Foreign Scan on fpagg_tab_p2 pagg_tab_2
                ->  Foreign Scan on fpagg_tab_p3 pagg_tab_3
-(9 rows)
+(10 rows)
 
 -- Plan with partitionwise aggregates is enabled
 SET enable_partitionwise_aggregate TO true;
@@ -8799,6 +8816,7 @@ SELECT a, count(t1) FROM pagg_tab t1 GROUP BY a HAVING avg(b) < 22 ORDER BY 1;
          ->  HashAggregate
                Output: t1.a, count(((t1.*)::pagg_tab))
                Group Key: t1.a
+               Buckets: 16
                Filter: (avg(t1.b) < '22'::numeric)
                ->  Foreign Scan on public.fpagg_tab_p1 t1
                      Output: t1.a, t1.*, t1.b
@@ -8806,6 +8824,7 @@ SELECT a, count(t1) FROM pagg_tab t1 GROUP BY a HAVING avg(b) < 22 ORDER BY 1;
          ->  HashAggregate
                Output: t1_1.a, count(((t1_1.*)::pagg_tab))
                Group Key: t1_1.a
+               Buckets: 16
                Filter: (avg(t1_1.b) < '22'::numeric)
                ->  Foreign Scan on public.fpagg_tab_p2 t1_1
                      Output: t1_1.a, t1_1.*, t1_1.b
@@ -8813,11 +8832,12 @@ SELECT a, count(t1) FROM pagg_tab t1 GROUP BY a HAVING avg(b) < 22 ORDER BY 1;
          ->  HashAggregate
                Output: t1_2.a, count(((t1_2.*)::pagg_tab))
                Group Key: t1_2.a
+               Buckets: 16
                Filter: (avg(t1_2.b) < '22'::numeric)
                ->  Foreign Scan on public.fpagg_tab_p3 t1_2
                      Output: t1_2.a, t1_2.*, t1_2.b
                      Remote SQL: SELECT a, b, c FROM public.pagg_tab_p3
-(25 rows)
+(28 rows)
 
 SELECT a, count(t1) FROM pagg_tab t1 GROUP BY a HAVING avg(b) < 22 ORDER BY 1;
  a  | count 
@@ -8839,18 +8859,22 @@ SELECT b, avg(a), max(a), count(*) FROM pagg_tab GROUP BY b HAVING sum(a) < 700
    Sort Key: pagg_tab.b
    ->  Finalize HashAggregate
          Group Key: pagg_tab.b
+         Buckets: 64
          Filter: (sum(pagg_tab.a) < 700)
          ->  Append
                ->  Partial HashAggregate
                      Group Key: pagg_tab.b
+                     Buckets: 64
                      ->  Foreign Scan on fpagg_tab_p1 pagg_tab
                ->  Partial HashAggregate
                      Group Key: pagg_tab_1.b
+                     Buckets: 64
                      ->  Foreign Scan on fpagg_tab_p2 pagg_tab_1
                ->  Partial HashAggregate
                      Group Key: pagg_tab_2.b
+                     Buckets: 64
                      ->  Foreign Scan on fpagg_tab_p3 pagg_tab_2
-(15 rows)
+(19 rows)
 
 -- ===================================================================
 -- access rights and superuser
diff --git a/src/backend/commands/explain.c b/src/backend/commands/explain.c
index 58141d8393..1d9623619b 100644
--- a/src/backend/commands/explain.c
+++ b/src/backend/commands/explain.c
@@ -18,6 +18,7 @@
 #include "commands/createas.h"
 #include "commands/defrem.h"
 #include "commands/prepare.h"
+#include "executor/nodeAgg.h"
 #include "executor/nodeHash.h"
 #include "foreign/fdwapi.h"
 #include "jit/jit.h"
@@ -86,12 +87,14 @@ static void show_merge_append_keys(MergeAppendState *mstate, List *ancestors,
 								   ExplainState *es);
 static void show_agg_keys(AggState *astate, List *ancestors,
 						  ExplainState *es);
-static void show_grouping_sets(PlanState *planstate, Agg *agg,
+static void show_grouping_sets(AggState *aggstate, Agg *agg,
 							   List *ancestors, ExplainState *es);
-static void show_grouping_set_keys(PlanState *planstate,
+static void show_grouping_set_info(AggState *aggstate,
 								   Agg *aggnode, Sort *sortnode,
 								   List *context, bool useprefix,
-								   List *ancestors, ExplainState *es);
+								   List *ancestors,
+								   HashTableInstrumentation *inst,
+								   ExplainState *es);
 static void show_group_keys(GroupState *gstate, List *ancestors,
 							ExplainState *es);
 static void show_sort_group_keys(PlanState *planstate, const char *qlabel,
@@ -104,7 +107,8 @@ static void show_tablesample(TableSampleClause *tsc, PlanState *planstate,
 							 List *ancestors, ExplainState *es);
 static void show_sort_info(SortState *sortstate, ExplainState *es);
 static void show_hash_info(HashState *hashstate, ExplainState *es);
-static void show_hashagg_info(AggState *hashstate, ExplainState *es);
+static void show_tuplehash_info(HashTableInstrumentation *inst, AggState *as,
+		ExplainState *es);
 static void show_tidbitmap_info(BitmapHeapScanState *planstate,
 								ExplainState *es);
 static void show_instrumentation_count(const char *qlabel, int which,
@@ -1490,6 +1494,7 @@ ExplainNode(PlanState *planstate, List *ancestors,
 					appendStringInfo(es->str, " %s", setopcmd);
 				else
 					ExplainPropertyText("Command", setopcmd, es);
+				// show strategy in text mode ?
 			}
 			break;
 		default:
@@ -1883,11 +1888,24 @@ ExplainNode(PlanState *planstate, List *ancestors,
 		case T_Agg:
 			show_agg_keys(castNode(AggState, planstate), ancestors, es);
 			show_upper_qual(plan->qual, "Filter", planstate, ancestors, es);
-			show_hashagg_info((AggState *) planstate, es);
 			if (plan->qual)
 				show_instrumentation_count("Rows Removed by Filter", 1,
 										   planstate, es);
 			break;
+		case T_SetOp:
+			{
+				SetOpState *sos = castNode(SetOpState, planstate);
+				if (sos->hashtable)
+					show_tuplehash_info(&sos->hashtable->instrument, NULL, es);
+			}
+			break;
+		case T_RecursiveUnion:
+			{
+				RecursiveUnionState *rus = (RecursiveUnionState *)planstate;
+				if (rus->hashtable)
+					show_tuplehash_info(&rus->hashtable->instrument, NULL, es);
+			}
+			break;
 		case T_Group:
 			show_group_keys(castNode(GroupState, planstate), ancestors, es);
 			show_upper_qual(plan->qual, "Filter", planstate, ancestors, es);
@@ -2264,24 +2282,31 @@ show_agg_keys(AggState *astate, List *ancestors,
 		ancestors = lcons(plan, ancestors);
 
 		if (plan->groupingSets)
-			show_grouping_sets(outerPlanState(astate), plan, ancestors, es);
+			show_grouping_sets(astate, plan, ancestors, es);
 		else
+		{
 			show_sort_group_keys(outerPlanState(astate), "Group Key",
 								 plan->numCols, plan->grpColIdx,
 								 NULL, NULL, NULL,
 								 ancestors, es);
+			Assert(astate->num_hashes <= 1);
+			if (astate->num_hashes)
+				show_tuplehash_info(&astate->perhash[0].hashtable->instrument, astate, es);
+		}
 
 		ancestors = list_delete_first(ancestors);
 	}
 }
 
 static void
-show_grouping_sets(PlanState *planstate, Agg *agg,
+show_grouping_sets(AggState *aggstate, Agg *agg,
 				   List *ancestors, ExplainState *es)
 {
+	PlanState	*planstate = outerPlanState(aggstate);
 	List	   *context;
 	bool		useprefix;
 	ListCell   *lc;
+	int			setno = 0;
 
 	/* Set up deparsing context */
 	context = set_deparse_context_plan(es->deparse_cxt,
@@ -2291,27 +2316,41 @@ show_grouping_sets(PlanState *planstate, Agg *agg,
 
 	ExplainOpenGroup("Grouping Sets", "Grouping Sets", false, es);
 
-	show_grouping_set_keys(planstate, agg, NULL,
-						   context, useprefix, ancestors, es);
+	show_grouping_set_info(aggstate, agg, NULL, context, useprefix, ancestors,
+			aggstate->num_hashes ?
+			&aggstate->perhash[setno++].hashtable->instrument : NULL,
+			es);
 
 	foreach(lc, agg->chain)
 	{
 		Agg		   *aggnode = lfirst(lc);
 		Sort	   *sortnode = (Sort *) aggnode->plan.lefttree;
+		HashTableInstrumentation *inst = NULL;
+
+		if (aggnode->aggstrategy == AGG_HASHED ||
+				aggnode->aggstrategy == AGG_MIXED)
+		{
+			Assert(setno < aggstate->num_hashes);
+			inst = &aggstate->perhash[setno++].hashtable->instrument;
+		}
 
-		show_grouping_set_keys(planstate, aggnode, sortnode,
-							   context, useprefix, ancestors, es);
+		show_grouping_set_info(aggstate, aggnode, sortnode,
+							   context, useprefix, ancestors,
+							   inst, es);
 	}
 
 	ExplainCloseGroup("Grouping Sets", "Grouping Sets", false, es);
 }
 
+/* Show keys and any hash instrumentation for a grouping set */
 static void
-show_grouping_set_keys(PlanState *planstate,
+show_grouping_set_info(AggState *aggstate,
 					   Agg *aggnode, Sort *sortnode,
 					   List *context, bool useprefix,
-					   List *ancestors, ExplainState *es)
+					   List *ancestors, HashTableInstrumentation *inst,
+					   ExplainState *es)
 {
+	PlanState	*planstate = outerPlanState(aggstate);
 	Plan	   *plan = planstate->plan;
 	char	   *exprstr;
 	ListCell   *lc;
@@ -2375,6 +2414,10 @@ show_grouping_set_keys(PlanState *planstate,
 
 	ExplainCloseGroup(keysetname, keysetname, false, es);
 
+	if (aggnode->aggstrategy == AGG_HASHED ||
+			aggnode->aggstrategy == AGG_MIXED)
+		show_tuplehash_info(inst, NULL, es);
+
 	if (sortnode && es->format == EXPLAIN_FORMAT_TEXT)
 		es->indent--;
 
@@ -2772,37 +2815,73 @@ show_hash_info(HashState *hashstate, ExplainState *es)
 }
 
 /*
- * Show information on hash aggregate memory usage and batches.
+ * Show hash bucket stats and (optionally) memory.
  */
 static void
-show_hashagg_info(AggState *aggstate, ExplainState *es)
+show_tuplehash_info(HashTableInstrumentation *inst, AggState *aggstate, ExplainState *es)
 {
-	Agg		*agg	   = (Agg *)aggstate->ss.ps.plan;
-	long	 memPeakKb = (aggstate->hash_mem_peak + 1023) / 1024;
+	size_t	spacePeakKb_tuples = (inst->space_peak_tuples + 1023) / 1024,
+		spacePeakKb_hash = (inst->space_peak_hash + 1023) / 1024;
 
-	Assert(IsA(aggstate, AggState));
-
-	if (agg->aggstrategy != AGG_HASHED &&
-		agg->aggstrategy != AGG_MIXED)
-		return;
-
-	if (es->costs && aggstate->hash_planned_partitions > 0)
-	{
+	if (es->costs && aggstate!=NULL && aggstate->hash_planned_partitions > 0)
 		ExplainPropertyInteger("Planned Partitions", NULL,
 							   aggstate->hash_planned_partitions, es);
-	}
 
 	if (!es->analyze)
 		return;
 
-	/* EXPLAIN ANALYZE */
-	ExplainPropertyInteger("Peak Memory Usage", "kB", memPeakKb, es);
-	if (aggstate->hash_batches_used > 0)
+	if (es->format != EXPLAIN_FORMAT_TEXT)
 	{
-		ExplainPropertyInteger("Disk Usage", "kB",
-							   aggstate->hash_disk_used, es);
-		ExplainPropertyInteger("HashAgg Batches", NULL,
-							   aggstate->hash_batches_used, es);
+		ExplainPropertyInteger("Hash Buckets", NULL,
+							   inst->nbuckets, es);
+		ExplainPropertyInteger("Original Hash Buckets", NULL,
+							   inst->nbuckets_original, es);
+		ExplainPropertyInteger("Peak Memory Usage (hashtable)", "kB",
+							   spacePeakKb_hash, es);
+		ExplainPropertyInteger("Peak Memory Usage (tuples)", "kB",
+							   spacePeakKb_tuples, es);
+		if (aggstate != NULL)
+		{
+			ExplainPropertyInteger("Disk Usage", "kB",
+								   aggstate->hash_disk_used, es);
+			ExplainPropertyInteger("HashAgg Batches", NULL,
+								   aggstate->hash_batches_used, es);
+		}
+	}
+	else if (!inst->nbuckets)
+		; /* Do nothing */
+	else
+	{
+		if (inst->nbuckets_original != inst->nbuckets)
+		{
+			ExplainIndentText(es);
+			appendStringInfo(es->str,
+						"Buckets: %lld (originally %lld)",
+						(long long)inst->nbuckets,
+						(long long)inst->nbuckets_original);
+		}
+		else
+		{
+			ExplainIndentText(es);
+			appendStringInfo(es->str,
+						"Buckets: %lld",
+						(long long)inst->nbuckets);
+		}
+
+		appendStringInfoChar(es->str, '\n');
+		ExplainIndentText(es);
+		appendStringInfo(es->str,
+				"Peak Memory Usage: hashtable: %lldkB, tuples: %lldkB",
+				(long long)spacePeakKb_hash, (long long)spacePeakKb_tuples);
+		appendStringInfoChar(es->str, '\n');
+
+		if (aggstate!=NULL && aggstate->hash_batches_used > 0)
+		{
+			ExplainPropertyInteger("Disk Usage", "kB",
+								   aggstate->hash_disk_used, es);
+			ExplainPropertyInteger("HashAgg Batches", NULL,
+								   aggstate->hash_batches_used, es);
+		}
 	}
 }
 
@@ -3473,6 +3552,29 @@ ExplainSubPlans(List *plans, List *ancestors,
 
 		ExplainNode(sps->planstate, ancestors,
 					relationship, sp->plan_name, es);
+		if (sps->hashtable)
+		{
+			ExplainOpenGroup("Hashtable", "Hashtable", true, es);
+			if (es->format == EXPLAIN_FORMAT_TEXT)
+			{
+				ExplainIndentText(es);
+				appendStringInfoString(es->str, "Hashtable: ");
+			}
+			show_tuplehash_info(&sps->hashtable->instrument, NULL, es);
+			ExplainCloseGroup("Hashtable", "Hashtable", true, es);
+		}
+
+		if (sps->hashnulls)
+		{
+			ExplainOpenGroup("Null Hashtable", "Null Hashtable", true, es);
+			if (es->format == EXPLAIN_FORMAT_TEXT)
+			{
+				ExplainIndentText(es);
+				appendStringInfoString(es->str, "Null Hashtable: ");
+			}
+			show_tuplehash_info(&sps->hashnulls->instrument, NULL, es);
+			ExplainCloseGroup("Null Hashtable", "Null Hashtable", true, es);
+		}
 
 		ancestors = list_delete_first(ancestors);
 	}
diff --git a/src/backend/executor/execGrouping.c b/src/backend/executor/execGrouping.c
index 009d27b9a8..10276d3f58 100644
--- a/src/backend/executor/execGrouping.c
+++ b/src/backend/executor/execGrouping.c
@@ -188,6 +188,7 @@ BuildTupleHashTableExt(PlanState *parent,
 	hashtable->inputslot = NULL;
 	hashtable->in_hash_funcs = NULL;
 	hashtable->cur_eq_func = NULL;
+	memset(&hashtable->instrument, 0, sizeof(hashtable->instrument));
 
 	/*
 	 * If parallelism is in use, even if the master backend is performing the
@@ -203,6 +204,7 @@ BuildTupleHashTableExt(PlanState *parent,
 		hashtable->hash_iv = 0;
 
 	hashtable->hashtab = tuplehash_create(metacxt, nbuckets, hashtable);
+	UpdateTupleHashTableStats(hashtable, true);
 
 	/*
 	 * We copy the input tuple descriptor just for safety --- we assume all
@@ -281,9 +283,40 @@ BuildTupleHashTable(PlanState *parent,
 void
 ResetTupleHashTable(TupleHashTable hashtable)
 {
+	UpdateTupleHashTableStats(hashtable, false);
 	tuplehash_reset(hashtable->hashtab);
 }
 
+/* Update instrumentation stats */
+void
+UpdateTupleHashTableStats(TupleHashTable hashtable, bool initial)
+{
+	hashtable->instrument.nbuckets = hashtable->hashtab->size;
+	if (initial)
+	{
+		hashtable->instrument.nbuckets_original = hashtable->hashtab->size;
+		// hashtable->instrument.space_peak_hash = hashtable->hashtab->size *
+			// sizeof(TupleHashEntryData);
+		hashtable->instrument.space_peak_hash =
+			MemoryContextMemAllocated(hashtable->hashtab->ctx, true);
+		hashtable->instrument.space_peak_tuples = 0;
+	}
+	else
+	{
+		/* hashtable->entrysize includes additionalsize */
+		size_t hash_size = MemoryContextMemAllocated(hashtable->hashtab->ctx, true);
+		size_t tuple_size = MemoryContextMemAllocated(hashtable->tablecxt, true);
+
+		hashtable->instrument.space_peak_hash = Max(
+			hashtable->instrument.space_peak_hash,
+			hash_size);
+
+		hashtable->instrument.space_peak_tuples = Max(
+			hashtable->instrument.space_peak_tuples, tuple_size);
+				// hashtable->hashtab->members * hashtable->entrysize);
+	}
+}
+
 /*
  * Find or create a hashtable entry for the tuple group containing the
  * given tuple.  The tuple must be the same type as the hashtable entries.
diff --git a/src/backend/executor/nodeAgg.c b/src/backend/executor/nodeAgg.c
index 1d319f49d0..daa82cdee2 100644
--- a/src/backend/executor/nodeAgg.c
+++ b/src/backend/executor/nodeAgg.c
@@ -1840,36 +1840,25 @@ hash_agg_enter_spill_mode(AggState *aggstate)
 static void
 hash_agg_update_metrics(AggState *aggstate, bool from_tape, int npartitions)
 {
-	Size	meta_mem = 0;
 	Size	hash_mem = 0;
 	Size	buffer_mem;
-	Size	total_mem;
 
 	if (aggstate->aggstrategy != AGG_MIXED &&
 		aggstate->aggstrategy != AGG_HASHED)
 		return;
 
-
 	for (int i = 0; i < aggstate->num_hashes; ++i)
 	{
-		/* memory for the hash table itself */
-		meta_mem += MemoryContextMemAllocated(
-			aggstate->perhash[i].hash_metacxt, true);
-		/* memory for the group keys and transition states */
 		hash_mem += MemoryContextMemAllocated(
 			aggstate->perhash[i].hashcontext->ecxt_per_tuple_memory, true);
+		UpdateTupleHashTableStats(aggstate->perhash[i].hashtable, false);
 	}
 
-	/* memory for read/write tape buffers, if spilled */
+	/* memory for read/write tape buffers, if spilled XXX */
 	buffer_mem = npartitions * HASHAGG_WRITE_BUFFER_SIZE;
 	if (from_tape)
 		buffer_mem += HASHAGG_READ_BUFFER_SIZE;
 
-	/* update peak mem */
-	total_mem = meta_mem + hash_mem + buffer_mem;
-	if (total_mem > aggstate->hash_mem_peak)
-		aggstate->hash_mem_peak = total_mem;
-
 	/* update disk usage */
 	if (aggstate->hash_tapeinfo != NULL)
 	{
diff --git a/src/backend/executor/nodeRecursiveunion.c b/src/backend/executor/nodeRecursiveunion.c
index 620414a1ed..93272c28b1 100644
--- a/src/backend/executor/nodeRecursiveunion.c
+++ b/src/backend/executor/nodeRecursiveunion.c
@@ -156,6 +156,9 @@ ExecRecursiveUnion(PlanState *pstate)
 		return slot;
 	}
 
+	if (node->hashtable)
+		UpdateTupleHashTableStats(node->hashtable, false);
+
 	return NULL;
 }
 
diff --git a/src/backend/executor/nodeSetOp.c b/src/backend/executor/nodeSetOp.c
index bfd148a41a..9c0e0ab96e 100644
--- a/src/backend/executor/nodeSetOp.c
+++ b/src/backend/executor/nodeSetOp.c
@@ -415,6 +415,7 @@ setop_fill_hash_table(SetOpState *setopstate)
 
 	setopstate->table_filled = true;
 	/* Initialize to walk the hash table */
+	UpdateTupleHashTableStats(setopstate->hashtable, false);
 	ResetTupleHashIterator(setopstate->hashtable, &setopstate->hashiter);
 }
 
diff --git a/src/backend/executor/nodeSubplan.c b/src/backend/executor/nodeSubplan.c
index 298b7757f5..22c32612ba 100644
--- a/src/backend/executor/nodeSubplan.c
+++ b/src/backend/executor/nodeSubplan.c
@@ -621,6 +621,9 @@ buildSubPlanHash(SubPlanState *node, ExprContext *econtext)
 	ExecClearTuple(node->projRight->pi_state.resultslot);
 
 	MemoryContextSwitchTo(oldcontext);
+	UpdateTupleHashTableStats(node->hashtable, false);
+	if (node->hashnulls)
+		UpdateTupleHashTableStats(node->hashnulls, false);
 }
 
 /*
diff --git a/src/include/executor/executor.h b/src/include/executor/executor.h
index 94890512dc..f4f2ede207 100644
--- a/src/include/executor/executor.h
+++ b/src/include/executor/executor.h
@@ -150,6 +150,7 @@ extern TupleHashEntry FindTupleHashEntry(TupleHashTable hashtable,
 										 ExprState *eqcomp,
 										 FmgrInfo *hashfunctions);
 extern void ResetTupleHashTable(TupleHashTable hashtable);
+extern void UpdateTupleHashTableStats(TupleHashTable hashtable, bool initial);
 
 /*
  * prototypes from functions in execJunk.c
diff --git a/src/include/nodes/execnodes.h b/src/include/nodes/execnodes.h
index ddf0b43916..63bbf22955 100644
--- a/src/include/nodes/execnodes.h
+++ b/src/include/nodes/execnodes.h
@@ -691,6 +691,14 @@ typedef struct TupleHashEntryData
 #define SH_DECLARE
 #include "lib/simplehash.h"
 
+typedef struct HashTableInstrumentation
+{
+	size_t	nbuckets;				/* number of buckets at end of execution */
+	size_t	nbuckets_original;		/* planned number of buckets */
+	size_t	space_peak_hash;		/* peak memory usage in bytes */
+	size_t	space_peak_tuples;		/* peak memory usage in bytes */
+} HashTableInstrumentation;
+
 typedef struct TupleHashTableData
 {
 	tuplehash_hash *hashtab;	/* underlying hash table */
@@ -709,6 +717,7 @@ typedef struct TupleHashTableData
 	ExprState  *cur_eq_func;	/* comparator for input vs. table */
 	uint32		hash_iv;		/* hash-function IV */
 	ExprContext *exprcontext;	/* expression context */
+	HashTableInstrumentation instrument;
 }			TupleHashTableData;
 
 typedef tuplehash_iterator TupleHashIterator;
@@ -2091,7 +2100,6 @@ typedef struct AggState
 	int			hash_planned_partitions; /* number of partitions planned
 											for first pass */
 	double		hashentrysize;	/* estimate revised during execution */
-	Size		hash_mem_peak;	/* peak hash table memory usage */
 	uint64		hash_ngroups_current;	/* number of groups currently in
 										   memory in all hash tables */
 	uint64		hash_disk_used; /* kB of disk space used */
-- 
2.17.0


--ZljC5FVPx7rxDQQ8
Content-Type: text/x-diff; charset=us-ascii
Content-Disposition: attachment;
 filename="v8-0003-refactor-show_grouping_set_keys.patch"



^ permalink  raw  reply  [nested|flat] 265+ messages in thread

* [PATCH v6] README for 64bit xid
@ 2022-01-10 19:20 Pavel Borisov <[email protected]>
  0 siblings, 0 replies; 265+ messages in thread

From: Pavel Borisov @ 2022-01-10 19:20 UTC (permalink / raw)

Authors:
- Pavel Borisov <[email protected]>
- Maxim Orlov <[email protected]>
- Yura Sokolov <[email protected]> <[email protected]>
---
 src/backend/access/heap/README.XID64 | 128 +++++++++++++++++++++++++++
 1 file changed, 128 insertions(+)
 create mode 100644 src/backend/access/heap/README.XID64

diff --git a/src/backend/access/heap/README.XID64 b/src/backend/access/heap/README.XID64
new file mode 100644
index 00000000000..457ba9b9ef5
--- /dev/null
+++ b/src/backend/access/heap/README.XID64
@@ -0,0 +1,128 @@
+src/backend/access/heap/README.XID64
+
+64-bit Transaction ID's (XID)
+=============================
+
+A limited number (N = 2^32) of XID's required to do vacuum freeze to prevent
+wraparound every N/2 transactions. This causes performance degradation due
+to the need to exclusively lock tables while being vacuumed. In each
+wraparound cycle, SLRU buffers are also being cut.
+
+With 64-bit XID's wraparound is effectively postponed to a very distant
+future. Even in highly loaded systems that had 2^32 transactions per day
+it will take huge 2^31 days before the first enforced "vacuum to prevent
+wraparound"). Buffers cutting and routine vacuum are not enforced, and DBA
+can plan them independently at the time with the least system load and least
+critical for database performance. Also, it can be done less frequently
+(several times a year vs every several days) on systems with transaction rates
+similar to those mentioned above.
+
+On-disk tuple and page format
+-----------------------------
+
+On-disk tuple format remains unchanged. 32-bit t_xmin and t_xmax store the
+lower parts of 64-bit XMIN and XMAX values. Each heap page has additional
+64-bit pd_xid_base and pd_multi_base which are common for all tuples on a page.
+They are placed into a pd_special area - 16 bytes in the end of a heap page.
+Actual XMIN/XMAX for a tuple are calculated upon reading a tuple from a page
+as follows:
+
+XMIN = t_xmin + pd_xid_base. 					(1)
+XMAX = t_xmax + pd_xid_base/pd_multi_base.		(2)
+
+"Double XMAX" page format
+---------------------------------
+
+At first read of a heap page after pg_upgrade from 32-bit XID PostgreSQL
+version pd_special area with a size of 16 bytes should be added to a page.
+Though a page may not have space for this. Then it can be converted to a
+temporary format called "double XMAX".
+
+All tuples after pg-upgrade would necessarily have xmin = FrozenTransactionId.
+So we don't need tuple header t_xmin field and we reuse t_xmin to store higher
+32 bits of its XMAX.
+
+Double XMAX format is only for full pages that don't have 16 bytes for
+pd_special. So it neither has a�place for a single tuple. Insert and HOT update
+for double XMAX pages is impossible and not supported. We can only read or
+delete tuples from it.
+
+When we are able to prune page double XMAX it will be converted from it to
+general 64-bit XID page format with all operations on its tuples supported.
+
+In-memory tuple format
+----------------------
+
+In-memory tuple representation consists of two parts:
+- HeapTupleHeader from disk page (contains all heap tuple contents, not only
+header)
+- HeapTuple with additional in-memory fields
+
+HeapTuple for each tuple in memory stores t_xid_base/t_multi_base - a copies of
+page's pd_xid_base/pd_multi_base. With tuple's 32-bit t_xmin and t_xmax from
+HeapTupleHeader they are used to calculate actual 64-bit XMIN and XMAX:
+
+XMIN = t_xmin + t_xid_base. 					(3)
+XMAX = t_xmax + t_xid_base/t_multi_base.		(4)
+
+The downside of this is that we can not use tuple's XMIN and XMAX right away.
+We often need to re-read t_xmin and t_xmax - which could actually be pointers
+into a page in shared buffers and therefore they could be updated by any other
+backend.
+
+Update/delete with 64-bit XIDs and 32-bit t_xmin/t_xmax
+--------------------------------------------------------------
+
+When we try to delete/update a tuple, we check that XMAX for a page fits (2).
+I.e. that t_xmax will not be over MaxShortTransactionId relative to
+pd_xid_base/pd_multi_base of a its page.
+
+If the current XID doesn't fit a range
+(pd_xid_base, pd_xid_base + MaxShortTransactionId) (5):
+
+- heap_page_prepare_for_xid() will try to increase pd_xid_base/pd_multi_base on
+a page and update all t_xmin/t_xmax of the other tuples on the page to
+correspond new pd_xid_base/pd_multi_base.
+
+- If it was impossible, it will try to prune and freeze tuples on a page.
+
+- If this is unsuccessful it will throw an error. Normally this is very
+unlikely but if there is a very old living transaction with an age of around
+2^32 this can arise. Basically, this is a behavior similar to one during the
+vacuum to prevent wraparound when XID was 32-bit. Dba should take care and
+avoid very-long-living transactions with an age close to 2^32. So long-living
+transactions often they are most likely defunct.
+
+Insert with 64-bit XIDs and 32-bit t_xmin/t_xmax
+------------------------------------------------
+
+On insert we check if current XID fits a range (5). Otherwise:
+
+- heap_page_prepare_for_xid() will try to increase pd_xid_base for t_xmin will
+not be over MaxShortTransactionId.
+
+- If it is impossible, then it will try to prune and freeze tuples on a page.
+
+Known issue: if pd_xid_base could not be shifted to accommodate a tuple being
+inserted due to a very long-running transaction, we just throw an error. We
+neither try to insert a�tuple into another page nor mark the current page as
+full. So, in this (unlikely) case we will get regular insert errors on the next
+tries to insert to the page 'locked' by this very long-running transaction.
+
+Upgrade from 32-bit XID versions
+--------------------------------
+
+pg_upgrade doesn't change pages format itself. It is done lazily after.
+
+1. At first heap page read, tuples on a page are repacked to free 16 bytes
+at the end of a page, possibly freeing space from dead tuples.
+
+2A. 16 bytes of pd_special is added if there is a place for it
+
+2B. Page is converted to "Double XMAX" format if there is no place for
+pd_special
+
+3. If a page is in double XMAX format after its first read, and vacuum (or
+micro-vacuum at select query) could prune some tuples and free space for
+pd_special, prune_page will add pd_special and convert page from double XMAX
+to general 64-bit XID page format.
-- 
2.24.3 (Apple Git-128)


--cpok4wp6gsarlzvp--





^ permalink  raw  reply  [nested|flat] 265+ messages in thread

* [PATCH v6] README for 64bit xid
@ 2022-01-10 19:20 Pavel Borisov <[email protected]>
  0 siblings, 0 replies; 265+ messages in thread

From: Pavel Borisov @ 2022-01-10 19:20 UTC (permalink / raw)

Authors:
- Pavel Borisov <[email protected]>
- Maxim Orlov <[email protected]>
- Yura Sokolov <[email protected]> <[email protected]>
---
 src/backend/access/heap/README.XID64 | 128 +++++++++++++++++++++++++++
 1 file changed, 128 insertions(+)
 create mode 100644 src/backend/access/heap/README.XID64

diff --git a/src/backend/access/heap/README.XID64 b/src/backend/access/heap/README.XID64
new file mode 100644
index 00000000000..457ba9b9ef5
--- /dev/null
+++ b/src/backend/access/heap/README.XID64
@@ -0,0 +1,128 @@
+src/backend/access/heap/README.XID64
+
+64-bit Transaction ID's (XID)
+=============================
+
+A limited number (N = 2^32) of XID's required to do vacuum freeze to prevent
+wraparound every N/2 transactions. This causes performance degradation due
+to the need to exclusively lock tables while being vacuumed. In each
+wraparound cycle, SLRU buffers are also being cut.
+
+With 64-bit XID's wraparound is effectively postponed to a very distant
+future. Even in highly loaded systems that had 2^32 transactions per day
+it will take huge 2^31 days before the first enforced "vacuum to prevent
+wraparound"). Buffers cutting and routine vacuum are not enforced, and DBA
+can plan them independently at the time with the least system load and least
+critical for database performance. Also, it can be done less frequently
+(several times a year vs every several days) on systems with transaction rates
+similar to those mentioned above.
+
+On-disk tuple and page format
+-----------------------------
+
+On-disk tuple format remains unchanged. 32-bit t_xmin and t_xmax store the
+lower parts of 64-bit XMIN and XMAX values. Each heap page has additional
+64-bit pd_xid_base and pd_multi_base which are common for all tuples on a page.
+They are placed into a pd_special area - 16 bytes in the end of a heap page.
+Actual XMIN/XMAX for a tuple are calculated upon reading a tuple from a page
+as follows:
+
+XMIN = t_xmin + pd_xid_base. 					(1)
+XMAX = t_xmax + pd_xid_base/pd_multi_base.		(2)
+
+"Double XMAX" page format
+---------------------------------
+
+At first read of a heap page after pg_upgrade from 32-bit XID PostgreSQL
+version pd_special area with a size of 16 bytes should be added to a page.
+Though a page may not have space for this. Then it can be converted to a
+temporary format called "double XMAX".
+
+All tuples after pg-upgrade would necessarily have xmin = FrozenTransactionId.
+So we don't need tuple header t_xmin field and we reuse t_xmin to store higher
+32 bits of its XMAX.
+
+Double XMAX format is only for full pages that don't have 16 bytes for
+pd_special. So it neither has a�place for a single tuple. Insert and HOT update
+for double XMAX pages is impossible and not supported. We can only read or
+delete tuples from it.
+
+When we are able to prune page double XMAX it will be converted from it to
+general 64-bit XID page format with all operations on its tuples supported.
+
+In-memory tuple format
+----------------------
+
+In-memory tuple representation consists of two parts:
+- HeapTupleHeader from disk page (contains all heap tuple contents, not only
+header)
+- HeapTuple with additional in-memory fields
+
+HeapTuple for each tuple in memory stores t_xid_base/t_multi_base - a copies of
+page's pd_xid_base/pd_multi_base. With tuple's 32-bit t_xmin and t_xmax from
+HeapTupleHeader they are used to calculate actual 64-bit XMIN and XMAX:
+
+XMIN = t_xmin + t_xid_base. 					(3)
+XMAX = t_xmax + t_xid_base/t_multi_base.		(4)
+
+The downside of this is that we can not use tuple's XMIN and XMAX right away.
+We often need to re-read t_xmin and t_xmax - which could actually be pointers
+into a page in shared buffers and therefore they could be updated by any other
+backend.
+
+Update/delete with 64-bit XIDs and 32-bit t_xmin/t_xmax
+--------------------------------------------------------------
+
+When we try to delete/update a tuple, we check that XMAX for a page fits (2).
+I.e. that t_xmax will not be over MaxShortTransactionId relative to
+pd_xid_base/pd_multi_base of a its page.
+
+If the current XID doesn't fit a range
+(pd_xid_base, pd_xid_base + MaxShortTransactionId) (5):
+
+- heap_page_prepare_for_xid() will try to increase pd_xid_base/pd_multi_base on
+a page and update all t_xmin/t_xmax of the other tuples on the page to
+correspond new pd_xid_base/pd_multi_base.
+
+- If it was impossible, it will try to prune and freeze tuples on a page.
+
+- If this is unsuccessful it will throw an error. Normally this is very
+unlikely but if there is a very old living transaction with an age of around
+2^32 this can arise. Basically, this is a behavior similar to one during the
+vacuum to prevent wraparound when XID was 32-bit. Dba should take care and
+avoid very-long-living transactions with an age close to 2^32. So long-living
+transactions often they are most likely defunct.
+
+Insert with 64-bit XIDs and 32-bit t_xmin/t_xmax
+------------------------------------------------
+
+On insert we check if current XID fits a range (5). Otherwise:
+
+- heap_page_prepare_for_xid() will try to increase pd_xid_base for t_xmin will
+not be over MaxShortTransactionId.
+
+- If it is impossible, then it will try to prune and freeze tuples on a page.
+
+Known issue: if pd_xid_base could not be shifted to accommodate a tuple being
+inserted due to a very long-running transaction, we just throw an error. We
+neither try to insert a�tuple into another page nor mark the current page as
+full. So, in this (unlikely) case we will get regular insert errors on the next
+tries to insert to the page 'locked' by this very long-running transaction.
+
+Upgrade from 32-bit XID versions
+--------------------------------
+
+pg_upgrade doesn't change pages format itself. It is done lazily after.
+
+1. At first heap page read, tuples on a page are repacked to free 16 bytes
+at the end of a page, possibly freeing space from dead tuples.
+
+2A. 16 bytes of pd_special is added if there is a place for it
+
+2B. Page is converted to "Double XMAX" format if there is no place for
+pd_special
+
+3. If a page is in double XMAX format after its first read, and vacuum (or
+micro-vacuum at select query) could prune some tuples and free space for
+pd_special, prune_page will add pd_special and convert page from double XMAX
+to general 64-bit XID page format.
-- 
2.24.3 (Apple Git-128)


--cpok4wp6gsarlzvp--





^ permalink  raw  reply  [nested|flat] 265+ messages in thread

* [PATCH v6] README for 64bit xid
@ 2022-01-10 19:20 Pavel Borisov <[email protected]>
  0 siblings, 0 replies; 265+ messages in thread

From: Pavel Borisov @ 2022-01-10 19:20 UTC (permalink / raw)

Authors:
- Pavel Borisov <[email protected]>
- Maxim Orlov <[email protected]>
- Yura Sokolov <[email protected]> <[email protected]>
---
 src/backend/access/heap/README.XID64 | 128 +++++++++++++++++++++++++++
 1 file changed, 128 insertions(+)
 create mode 100644 src/backend/access/heap/README.XID64

diff --git a/src/backend/access/heap/README.XID64 b/src/backend/access/heap/README.XID64
new file mode 100644
index 00000000000..457ba9b9ef5
--- /dev/null
+++ b/src/backend/access/heap/README.XID64
@@ -0,0 +1,128 @@
+src/backend/access/heap/README.XID64
+
+64-bit Transaction ID's (XID)
+=============================
+
+A limited number (N = 2^32) of XID's required to do vacuum freeze to prevent
+wraparound every N/2 transactions. This causes performance degradation due
+to the need to exclusively lock tables while being vacuumed. In each
+wraparound cycle, SLRU buffers are also being cut.
+
+With 64-bit XID's wraparound is effectively postponed to a very distant
+future. Even in highly loaded systems that had 2^32 transactions per day
+it will take huge 2^31 days before the first enforced "vacuum to prevent
+wraparound"). Buffers cutting and routine vacuum are not enforced, and DBA
+can plan them independently at the time with the least system load and least
+critical for database performance. Also, it can be done less frequently
+(several times a year vs every several days) on systems with transaction rates
+similar to those mentioned above.
+
+On-disk tuple and page format
+-----------------------------
+
+On-disk tuple format remains unchanged. 32-bit t_xmin and t_xmax store the
+lower parts of 64-bit XMIN and XMAX values. Each heap page has additional
+64-bit pd_xid_base and pd_multi_base which are common for all tuples on a page.
+They are placed into a pd_special area - 16 bytes in the end of a heap page.
+Actual XMIN/XMAX for a tuple are calculated upon reading a tuple from a page
+as follows:
+
+XMIN = t_xmin + pd_xid_base. 					(1)
+XMAX = t_xmax + pd_xid_base/pd_multi_base.		(2)
+
+"Double XMAX" page format
+---------------------------------
+
+At first read of a heap page after pg_upgrade from 32-bit XID PostgreSQL
+version pd_special area with a size of 16 bytes should be added to a page.
+Though a page may not have space for this. Then it can be converted to a
+temporary format called "double XMAX".
+
+All tuples after pg-upgrade would necessarily have xmin = FrozenTransactionId.
+So we don't need tuple header t_xmin field and we reuse t_xmin to store higher
+32 bits of its XMAX.
+
+Double XMAX format is only for full pages that don't have 16 bytes for
+pd_special. So it neither has a�place for a single tuple. Insert and HOT update
+for double XMAX pages is impossible and not supported. We can only read or
+delete tuples from it.
+
+When we are able to prune page double XMAX it will be converted from it to
+general 64-bit XID page format with all operations on its tuples supported.
+
+In-memory tuple format
+----------------------
+
+In-memory tuple representation consists of two parts:
+- HeapTupleHeader from disk page (contains all heap tuple contents, not only
+header)
+- HeapTuple with additional in-memory fields
+
+HeapTuple for each tuple in memory stores t_xid_base/t_multi_base - a copies of
+page's pd_xid_base/pd_multi_base. With tuple's 32-bit t_xmin and t_xmax from
+HeapTupleHeader they are used to calculate actual 64-bit XMIN and XMAX:
+
+XMIN = t_xmin + t_xid_base. 					(3)
+XMAX = t_xmax + t_xid_base/t_multi_base.		(4)
+
+The downside of this is that we can not use tuple's XMIN and XMAX right away.
+We often need to re-read t_xmin and t_xmax - which could actually be pointers
+into a page in shared buffers and therefore they could be updated by any other
+backend.
+
+Update/delete with 64-bit XIDs and 32-bit t_xmin/t_xmax
+--------------------------------------------------------------
+
+When we try to delete/update a tuple, we check that XMAX for a page fits (2).
+I.e. that t_xmax will not be over MaxShortTransactionId relative to
+pd_xid_base/pd_multi_base of a its page.
+
+If the current XID doesn't fit a range
+(pd_xid_base, pd_xid_base + MaxShortTransactionId) (5):
+
+- heap_page_prepare_for_xid() will try to increase pd_xid_base/pd_multi_base on
+a page and update all t_xmin/t_xmax of the other tuples on the page to
+correspond new pd_xid_base/pd_multi_base.
+
+- If it was impossible, it will try to prune and freeze tuples on a page.
+
+- If this is unsuccessful it will throw an error. Normally this is very
+unlikely but if there is a very old living transaction with an age of around
+2^32 this can arise. Basically, this is a behavior similar to one during the
+vacuum to prevent wraparound when XID was 32-bit. Dba should take care and
+avoid very-long-living transactions with an age close to 2^32. So long-living
+transactions often they are most likely defunct.
+
+Insert with 64-bit XIDs and 32-bit t_xmin/t_xmax
+------------------------------------------------
+
+On insert we check if current XID fits a range (5). Otherwise:
+
+- heap_page_prepare_for_xid() will try to increase pd_xid_base for t_xmin will
+not be over MaxShortTransactionId.
+
+- If it is impossible, then it will try to prune and freeze tuples on a page.
+
+Known issue: if pd_xid_base could not be shifted to accommodate a tuple being
+inserted due to a very long-running transaction, we just throw an error. We
+neither try to insert a�tuple into another page nor mark the current page as
+full. So, in this (unlikely) case we will get regular insert errors on the next
+tries to insert to the page 'locked' by this very long-running transaction.
+
+Upgrade from 32-bit XID versions
+--------------------------------
+
+pg_upgrade doesn't change pages format itself. It is done lazily after.
+
+1. At first heap page read, tuples on a page are repacked to free 16 bytes
+at the end of a page, possibly freeing space from dead tuples.
+
+2A. 16 bytes of pd_special is added if there is a place for it
+
+2B. Page is converted to "Double XMAX" format if there is no place for
+pd_special
+
+3. If a page is in double XMAX format after its first read, and vacuum (or
+micro-vacuum at select query) could prune some tuples and free space for
+pd_special, prune_page will add pd_special and convert page from double XMAX
+to general 64-bit XID page format.
-- 
2.24.3 (Apple Git-128)


--cpok4wp6gsarlzvp--





^ permalink  raw  reply  [nested|flat] 265+ messages in thread

* [PATCH v6] README for 64bit xid
@ 2022-01-10 19:20 Pavel Borisov <[email protected]>
  0 siblings, 0 replies; 265+ messages in thread

From: Pavel Borisov @ 2022-01-10 19:20 UTC (permalink / raw)

Authors:
- Pavel Borisov <[email protected]>
- Maxim Orlov <[email protected]>
- Yura Sokolov <[email protected]> <[email protected]>
---
 src/backend/access/heap/README.XID64 | 128 +++++++++++++++++++++++++++
 1 file changed, 128 insertions(+)
 create mode 100644 src/backend/access/heap/README.XID64

diff --git a/src/backend/access/heap/README.XID64 b/src/backend/access/heap/README.XID64
new file mode 100644
index 00000000000..457ba9b9ef5
--- /dev/null
+++ b/src/backend/access/heap/README.XID64
@@ -0,0 +1,128 @@
+src/backend/access/heap/README.XID64
+
+64-bit Transaction ID's (XID)
+=============================
+
+A limited number (N = 2^32) of XID's required to do vacuum freeze to prevent
+wraparound every N/2 transactions. This causes performance degradation due
+to the need to exclusively lock tables while being vacuumed. In each
+wraparound cycle, SLRU buffers are also being cut.
+
+With 64-bit XID's wraparound is effectively postponed to a very distant
+future. Even in highly loaded systems that had 2^32 transactions per day
+it will take huge 2^31 days before the first enforced "vacuum to prevent
+wraparound"). Buffers cutting and routine vacuum are not enforced, and DBA
+can plan them independently at the time with the least system load and least
+critical for database performance. Also, it can be done less frequently
+(several times a year vs every several days) on systems with transaction rates
+similar to those mentioned above.
+
+On-disk tuple and page format
+-----------------------------
+
+On-disk tuple format remains unchanged. 32-bit t_xmin and t_xmax store the
+lower parts of 64-bit XMIN and XMAX values. Each heap page has additional
+64-bit pd_xid_base and pd_multi_base which are common for all tuples on a page.
+They are placed into a pd_special area - 16 bytes in the end of a heap page.
+Actual XMIN/XMAX for a tuple are calculated upon reading a tuple from a page
+as follows:
+
+XMIN = t_xmin + pd_xid_base. 					(1)
+XMAX = t_xmax + pd_xid_base/pd_multi_base.		(2)
+
+"Double XMAX" page format
+---------------------------------
+
+At first read of a heap page after pg_upgrade from 32-bit XID PostgreSQL
+version pd_special area with a size of 16 bytes should be added to a page.
+Though a page may not have space for this. Then it can be converted to a
+temporary format called "double XMAX".
+
+All tuples after pg-upgrade would necessarily have xmin = FrozenTransactionId.
+So we don't need tuple header t_xmin field and we reuse t_xmin to store higher
+32 bits of its XMAX.
+
+Double XMAX format is only for full pages that don't have 16 bytes for
+pd_special. So it neither has a�place for a single tuple. Insert and HOT update
+for double XMAX pages is impossible and not supported. We can only read or
+delete tuples from it.
+
+When we are able to prune page double XMAX it will be converted from it to
+general 64-bit XID page format with all operations on its tuples supported.
+
+In-memory tuple format
+----------------------
+
+In-memory tuple representation consists of two parts:
+- HeapTupleHeader from disk page (contains all heap tuple contents, not only
+header)
+- HeapTuple with additional in-memory fields
+
+HeapTuple for each tuple in memory stores t_xid_base/t_multi_base - a copies of
+page's pd_xid_base/pd_multi_base. With tuple's 32-bit t_xmin and t_xmax from
+HeapTupleHeader they are used to calculate actual 64-bit XMIN and XMAX:
+
+XMIN = t_xmin + t_xid_base. 					(3)
+XMAX = t_xmax + t_xid_base/t_multi_base.		(4)
+
+The downside of this is that we can not use tuple's XMIN and XMAX right away.
+We often need to re-read t_xmin and t_xmax - which could actually be pointers
+into a page in shared buffers and therefore they could be updated by any other
+backend.
+
+Update/delete with 64-bit XIDs and 32-bit t_xmin/t_xmax
+--------------------------------------------------------------
+
+When we try to delete/update a tuple, we check that XMAX for a page fits (2).
+I.e. that t_xmax will not be over MaxShortTransactionId relative to
+pd_xid_base/pd_multi_base of a its page.
+
+If the current XID doesn't fit a range
+(pd_xid_base, pd_xid_base + MaxShortTransactionId) (5):
+
+- heap_page_prepare_for_xid() will try to increase pd_xid_base/pd_multi_base on
+a page and update all t_xmin/t_xmax of the other tuples on the page to
+correspond new pd_xid_base/pd_multi_base.
+
+- If it was impossible, it will try to prune and freeze tuples on a page.
+
+- If this is unsuccessful it will throw an error. Normally this is very
+unlikely but if there is a very old living transaction with an age of around
+2^32 this can arise. Basically, this is a behavior similar to one during the
+vacuum to prevent wraparound when XID was 32-bit. Dba should take care and
+avoid very-long-living transactions with an age close to 2^32. So long-living
+transactions often they are most likely defunct.
+
+Insert with 64-bit XIDs and 32-bit t_xmin/t_xmax
+------------------------------------------------
+
+On insert we check if current XID fits a range (5). Otherwise:
+
+- heap_page_prepare_for_xid() will try to increase pd_xid_base for t_xmin will
+not be over MaxShortTransactionId.
+
+- If it is impossible, then it will try to prune and freeze tuples on a page.
+
+Known issue: if pd_xid_base could not be shifted to accommodate a tuple being
+inserted due to a very long-running transaction, we just throw an error. We
+neither try to insert a�tuple into another page nor mark the current page as
+full. So, in this (unlikely) case we will get regular insert errors on the next
+tries to insert to the page 'locked' by this very long-running transaction.
+
+Upgrade from 32-bit XID versions
+--------------------------------
+
+pg_upgrade doesn't change pages format itself. It is done lazily after.
+
+1. At first heap page read, tuples on a page are repacked to free 16 bytes
+at the end of a page, possibly freeing space from dead tuples.
+
+2A. 16 bytes of pd_special is added if there is a place for it
+
+2B. Page is converted to "Double XMAX" format if there is no place for
+pd_special
+
+3. If a page is in double XMAX format after its first read, and vacuum (or
+micro-vacuum at select query) could prune some tuples and free space for
+pd_special, prune_page will add pd_special and convert page from double XMAX
+to general 64-bit XID page format.
-- 
2.24.3 (Apple Git-128)


--cpok4wp6gsarlzvp--





^ permalink  raw  reply  [nested|flat] 265+ messages in thread

* [PATCH v6] README for 64bit xid
@ 2022-01-10 19:20 Pavel Borisov <[email protected]>
  0 siblings, 0 replies; 265+ messages in thread

From: Pavel Borisov @ 2022-01-10 19:20 UTC (permalink / raw)

Authors:
- Pavel Borisov <[email protected]>
- Maxim Orlov <[email protected]>
- Yura Sokolov <[email protected]> <[email protected]>
---
 src/backend/access/heap/README.XID64 | 128 +++++++++++++++++++++++++++
 1 file changed, 128 insertions(+)
 create mode 100644 src/backend/access/heap/README.XID64

diff --git a/src/backend/access/heap/README.XID64 b/src/backend/access/heap/README.XID64
new file mode 100644
index 00000000000..457ba9b9ef5
--- /dev/null
+++ b/src/backend/access/heap/README.XID64
@@ -0,0 +1,128 @@
+src/backend/access/heap/README.XID64
+
+64-bit Transaction ID's (XID)
+=============================
+
+A limited number (N = 2^32) of XID's required to do vacuum freeze to prevent
+wraparound every N/2 transactions. This causes performance degradation due
+to the need to exclusively lock tables while being vacuumed. In each
+wraparound cycle, SLRU buffers are also being cut.
+
+With 64-bit XID's wraparound is effectively postponed to a very distant
+future. Even in highly loaded systems that had 2^32 transactions per day
+it will take huge 2^31 days before the first enforced "vacuum to prevent
+wraparound"). Buffers cutting and routine vacuum are not enforced, and DBA
+can plan them independently at the time with the least system load and least
+critical for database performance. Also, it can be done less frequently
+(several times a year vs every several days) on systems with transaction rates
+similar to those mentioned above.
+
+On-disk tuple and page format
+-----------------------------
+
+On-disk tuple format remains unchanged. 32-bit t_xmin and t_xmax store the
+lower parts of 64-bit XMIN and XMAX values. Each heap page has additional
+64-bit pd_xid_base and pd_multi_base which are common for all tuples on a page.
+They are placed into a pd_special area - 16 bytes in the end of a heap page.
+Actual XMIN/XMAX for a tuple are calculated upon reading a tuple from a page
+as follows:
+
+XMIN = t_xmin + pd_xid_base. 					(1)
+XMAX = t_xmax + pd_xid_base/pd_multi_base.		(2)
+
+"Double XMAX" page format
+---------------------------------
+
+At first read of a heap page after pg_upgrade from 32-bit XID PostgreSQL
+version pd_special area with a size of 16 bytes should be added to a page.
+Though a page may not have space for this. Then it can be converted to a
+temporary format called "double XMAX".
+
+All tuples after pg-upgrade would necessarily have xmin = FrozenTransactionId.
+So we don't need tuple header t_xmin field and we reuse t_xmin to store higher
+32 bits of its XMAX.
+
+Double XMAX format is only for full pages that don't have 16 bytes for
+pd_special. So it neither has a�place for a single tuple. Insert and HOT update
+for double XMAX pages is impossible and not supported. We can only read or
+delete tuples from it.
+
+When we are able to prune page double XMAX it will be converted from it to
+general 64-bit XID page format with all operations on its tuples supported.
+
+In-memory tuple format
+----------------------
+
+In-memory tuple representation consists of two parts:
+- HeapTupleHeader from disk page (contains all heap tuple contents, not only
+header)
+- HeapTuple with additional in-memory fields
+
+HeapTuple for each tuple in memory stores t_xid_base/t_multi_base - a copies of
+page's pd_xid_base/pd_multi_base. With tuple's 32-bit t_xmin and t_xmax from
+HeapTupleHeader they are used to calculate actual 64-bit XMIN and XMAX:
+
+XMIN = t_xmin + t_xid_base. 					(3)
+XMAX = t_xmax + t_xid_base/t_multi_base.		(4)
+
+The downside of this is that we can not use tuple's XMIN and XMAX right away.
+We often need to re-read t_xmin and t_xmax - which could actually be pointers
+into a page in shared buffers and therefore they could be updated by any other
+backend.
+
+Update/delete with 64-bit XIDs and 32-bit t_xmin/t_xmax
+--------------------------------------------------------------
+
+When we try to delete/update a tuple, we check that XMAX for a page fits (2).
+I.e. that t_xmax will not be over MaxShortTransactionId relative to
+pd_xid_base/pd_multi_base of a its page.
+
+If the current XID doesn't fit a range
+(pd_xid_base, pd_xid_base + MaxShortTransactionId) (5):
+
+- heap_page_prepare_for_xid() will try to increase pd_xid_base/pd_multi_base on
+a page and update all t_xmin/t_xmax of the other tuples on the page to
+correspond new pd_xid_base/pd_multi_base.
+
+- If it was impossible, it will try to prune and freeze tuples on a page.
+
+- If this is unsuccessful it will throw an error. Normally this is very
+unlikely but if there is a very old living transaction with an age of around
+2^32 this can arise. Basically, this is a behavior similar to one during the
+vacuum to prevent wraparound when XID was 32-bit. Dba should take care and
+avoid very-long-living transactions with an age close to 2^32. So long-living
+transactions often they are most likely defunct.
+
+Insert with 64-bit XIDs and 32-bit t_xmin/t_xmax
+------------------------------------------------
+
+On insert we check if current XID fits a range (5). Otherwise:
+
+- heap_page_prepare_for_xid() will try to increase pd_xid_base for t_xmin will
+not be over MaxShortTransactionId.
+
+- If it is impossible, then it will try to prune and freeze tuples on a page.
+
+Known issue: if pd_xid_base could not be shifted to accommodate a tuple being
+inserted due to a very long-running transaction, we just throw an error. We
+neither try to insert a�tuple into another page nor mark the current page as
+full. So, in this (unlikely) case we will get regular insert errors on the next
+tries to insert to the page 'locked' by this very long-running transaction.
+
+Upgrade from 32-bit XID versions
+--------------------------------
+
+pg_upgrade doesn't change pages format itself. It is done lazily after.
+
+1. At first heap page read, tuples on a page are repacked to free 16 bytes
+at the end of a page, possibly freeing space from dead tuples.
+
+2A. 16 bytes of pd_special is added if there is a place for it
+
+2B. Page is converted to "Double XMAX" format if there is no place for
+pd_special
+
+3. If a page is in double XMAX format after its first read, and vacuum (or
+micro-vacuum at select query) could prune some tuples and free space for
+pd_special, prune_page will add pd_special and convert page from double XMAX
+to general 64-bit XID page format.
-- 
2.24.3 (Apple Git-128)


--cpok4wp6gsarlzvp--





^ permalink  raw  reply  [nested|flat] 265+ messages in thread

* [PATCH v6] README for 64bit xid
@ 2022-01-10 19:20 Pavel Borisov <[email protected]>
  0 siblings, 0 replies; 265+ messages in thread

From: Pavel Borisov @ 2022-01-10 19:20 UTC (permalink / raw)

Authors:
- Pavel Borisov <[email protected]>
- Maxim Orlov <[email protected]>
- Yura Sokolov <[email protected]> <[email protected]>
---
 src/backend/access/heap/README.XID64 | 128 +++++++++++++++++++++++++++
 1 file changed, 128 insertions(+)
 create mode 100644 src/backend/access/heap/README.XID64

diff --git a/src/backend/access/heap/README.XID64 b/src/backend/access/heap/README.XID64
new file mode 100644
index 00000000000..457ba9b9ef5
--- /dev/null
+++ b/src/backend/access/heap/README.XID64
@@ -0,0 +1,128 @@
+src/backend/access/heap/README.XID64
+
+64-bit Transaction ID's (XID)
+=============================
+
+A limited number (N = 2^32) of XID's required to do vacuum freeze to prevent
+wraparound every N/2 transactions. This causes performance degradation due
+to the need to exclusively lock tables while being vacuumed. In each
+wraparound cycle, SLRU buffers are also being cut.
+
+With 64-bit XID's wraparound is effectively postponed to a very distant
+future. Even in highly loaded systems that had 2^32 transactions per day
+it will take huge 2^31 days before the first enforced "vacuum to prevent
+wraparound"). Buffers cutting and routine vacuum are not enforced, and DBA
+can plan them independently at the time with the least system load and least
+critical for database performance. Also, it can be done less frequently
+(several times a year vs every several days) on systems with transaction rates
+similar to those mentioned above.
+
+On-disk tuple and page format
+-----------------------------
+
+On-disk tuple format remains unchanged. 32-bit t_xmin and t_xmax store the
+lower parts of 64-bit XMIN and XMAX values. Each heap page has additional
+64-bit pd_xid_base and pd_multi_base which are common for all tuples on a page.
+They are placed into a pd_special area - 16 bytes in the end of a heap page.
+Actual XMIN/XMAX for a tuple are calculated upon reading a tuple from a page
+as follows:
+
+XMIN = t_xmin + pd_xid_base. 					(1)
+XMAX = t_xmax + pd_xid_base/pd_multi_base.		(2)
+
+"Double XMAX" page format
+---------------------------------
+
+At first read of a heap page after pg_upgrade from 32-bit XID PostgreSQL
+version pd_special area with a size of 16 bytes should be added to a page.
+Though a page may not have space for this. Then it can be converted to a
+temporary format called "double XMAX".
+
+All tuples after pg-upgrade would necessarily have xmin = FrozenTransactionId.
+So we don't need tuple header t_xmin field and we reuse t_xmin to store higher
+32 bits of its XMAX.
+
+Double XMAX format is only for full pages that don't have 16 bytes for
+pd_special. So it neither has a�place for a single tuple. Insert and HOT update
+for double XMAX pages is impossible and not supported. We can only read or
+delete tuples from it.
+
+When we are able to prune page double XMAX it will be converted from it to
+general 64-bit XID page format with all operations on its tuples supported.
+
+In-memory tuple format
+----------------------
+
+In-memory tuple representation consists of two parts:
+- HeapTupleHeader from disk page (contains all heap tuple contents, not only
+header)
+- HeapTuple with additional in-memory fields
+
+HeapTuple for each tuple in memory stores t_xid_base/t_multi_base - a copies of
+page's pd_xid_base/pd_multi_base. With tuple's 32-bit t_xmin and t_xmax from
+HeapTupleHeader they are used to calculate actual 64-bit XMIN and XMAX:
+
+XMIN = t_xmin + t_xid_base. 					(3)
+XMAX = t_xmax + t_xid_base/t_multi_base.		(4)
+
+The downside of this is that we can not use tuple's XMIN and XMAX right away.
+We often need to re-read t_xmin and t_xmax - which could actually be pointers
+into a page in shared buffers and therefore they could be updated by any other
+backend.
+
+Update/delete with 64-bit XIDs and 32-bit t_xmin/t_xmax
+--------------------------------------------------------------
+
+When we try to delete/update a tuple, we check that XMAX for a page fits (2).
+I.e. that t_xmax will not be over MaxShortTransactionId relative to
+pd_xid_base/pd_multi_base of a its page.
+
+If the current XID doesn't fit a range
+(pd_xid_base, pd_xid_base + MaxShortTransactionId) (5):
+
+- heap_page_prepare_for_xid() will try to increase pd_xid_base/pd_multi_base on
+a page and update all t_xmin/t_xmax of the other tuples on the page to
+correspond new pd_xid_base/pd_multi_base.
+
+- If it was impossible, it will try to prune and freeze tuples on a page.
+
+- If this is unsuccessful it will throw an error. Normally this is very
+unlikely but if there is a very old living transaction with an age of around
+2^32 this can arise. Basically, this is a behavior similar to one during the
+vacuum to prevent wraparound when XID was 32-bit. Dba should take care and
+avoid very-long-living transactions with an age close to 2^32. So long-living
+transactions often they are most likely defunct.
+
+Insert with 64-bit XIDs and 32-bit t_xmin/t_xmax
+------------------------------------------------
+
+On insert we check if current XID fits a range (5). Otherwise:
+
+- heap_page_prepare_for_xid() will try to increase pd_xid_base for t_xmin will
+not be over MaxShortTransactionId.
+
+- If it is impossible, then it will try to prune and freeze tuples on a page.
+
+Known issue: if pd_xid_base could not be shifted to accommodate a tuple being
+inserted due to a very long-running transaction, we just throw an error. We
+neither try to insert a�tuple into another page nor mark the current page as
+full. So, in this (unlikely) case we will get regular insert errors on the next
+tries to insert to the page 'locked' by this very long-running transaction.
+
+Upgrade from 32-bit XID versions
+--------------------------------
+
+pg_upgrade doesn't change pages format itself. It is done lazily after.
+
+1. At first heap page read, tuples on a page are repacked to free 16 bytes
+at the end of a page, possibly freeing space from dead tuples.
+
+2A. 16 bytes of pd_special is added if there is a place for it
+
+2B. Page is converted to "Double XMAX" format if there is no place for
+pd_special
+
+3. If a page is in double XMAX format after its first read, and vacuum (or
+micro-vacuum at select query) could prune some tuples and free space for
+pd_special, prune_page will add pd_special and convert page from double XMAX
+to general 64-bit XID page format.
-- 
2.24.3 (Apple Git-128)


--cpok4wp6gsarlzvp--





^ permalink  raw  reply  [nested|flat] 265+ messages in thread

* [PATCH v6] README for 64bit xid
@ 2022-01-10 19:20 Pavel Borisov <[email protected]>
  0 siblings, 0 replies; 265+ messages in thread

From: Pavel Borisov @ 2022-01-10 19:20 UTC (permalink / raw)

Authors:
- Pavel Borisov <[email protected]>
- Maxim Orlov <[email protected]>
- Yura Sokolov <[email protected]> <[email protected]>
---
 src/backend/access/heap/README.XID64 | 128 +++++++++++++++++++++++++++
 1 file changed, 128 insertions(+)
 create mode 100644 src/backend/access/heap/README.XID64

diff --git a/src/backend/access/heap/README.XID64 b/src/backend/access/heap/README.XID64
new file mode 100644
index 00000000000..457ba9b9ef5
--- /dev/null
+++ b/src/backend/access/heap/README.XID64
@@ -0,0 +1,128 @@
+src/backend/access/heap/README.XID64
+
+64-bit Transaction ID's (XID)
+=============================
+
+A limited number (N = 2^32) of XID's required to do vacuum freeze to prevent
+wraparound every N/2 transactions. This causes performance degradation due
+to the need to exclusively lock tables while being vacuumed. In each
+wraparound cycle, SLRU buffers are also being cut.
+
+With 64-bit XID's wraparound is effectively postponed to a very distant
+future. Even in highly loaded systems that had 2^32 transactions per day
+it will take huge 2^31 days before the first enforced "vacuum to prevent
+wraparound"). Buffers cutting and routine vacuum are not enforced, and DBA
+can plan them independently at the time with the least system load and least
+critical for database performance. Also, it can be done less frequently
+(several times a year vs every several days) on systems with transaction rates
+similar to those mentioned above.
+
+On-disk tuple and page format
+-----------------------------
+
+On-disk tuple format remains unchanged. 32-bit t_xmin and t_xmax store the
+lower parts of 64-bit XMIN and XMAX values. Each heap page has additional
+64-bit pd_xid_base and pd_multi_base which are common for all tuples on a page.
+They are placed into a pd_special area - 16 bytes in the end of a heap page.
+Actual XMIN/XMAX for a tuple are calculated upon reading a tuple from a page
+as follows:
+
+XMIN = t_xmin + pd_xid_base. 					(1)
+XMAX = t_xmax + pd_xid_base/pd_multi_base.		(2)
+
+"Double XMAX" page format
+---------------------------------
+
+At first read of a heap page after pg_upgrade from 32-bit XID PostgreSQL
+version pd_special area with a size of 16 bytes should be added to a page.
+Though a page may not have space for this. Then it can be converted to a
+temporary format called "double XMAX".
+
+All tuples after pg-upgrade would necessarily have xmin = FrozenTransactionId.
+So we don't need tuple header t_xmin field and we reuse t_xmin to store higher
+32 bits of its XMAX.
+
+Double XMAX format is only for full pages that don't have 16 bytes for
+pd_special. So it neither has a�place for a single tuple. Insert and HOT update
+for double XMAX pages is impossible and not supported. We can only read or
+delete tuples from it.
+
+When we are able to prune page double XMAX it will be converted from it to
+general 64-bit XID page format with all operations on its tuples supported.
+
+In-memory tuple format
+----------------------
+
+In-memory tuple representation consists of two parts:
+- HeapTupleHeader from disk page (contains all heap tuple contents, not only
+header)
+- HeapTuple with additional in-memory fields
+
+HeapTuple for each tuple in memory stores t_xid_base/t_multi_base - a copies of
+page's pd_xid_base/pd_multi_base. With tuple's 32-bit t_xmin and t_xmax from
+HeapTupleHeader they are used to calculate actual 64-bit XMIN and XMAX:
+
+XMIN = t_xmin + t_xid_base. 					(3)
+XMAX = t_xmax + t_xid_base/t_multi_base.		(4)
+
+The downside of this is that we can not use tuple's XMIN and XMAX right away.
+We often need to re-read t_xmin and t_xmax - which could actually be pointers
+into a page in shared buffers and therefore they could be updated by any other
+backend.
+
+Update/delete with 64-bit XIDs and 32-bit t_xmin/t_xmax
+--------------------------------------------------------------
+
+When we try to delete/update a tuple, we check that XMAX for a page fits (2).
+I.e. that t_xmax will not be over MaxShortTransactionId relative to
+pd_xid_base/pd_multi_base of a its page.
+
+If the current XID doesn't fit a range
+(pd_xid_base, pd_xid_base + MaxShortTransactionId) (5):
+
+- heap_page_prepare_for_xid() will try to increase pd_xid_base/pd_multi_base on
+a page and update all t_xmin/t_xmax of the other tuples on the page to
+correspond new pd_xid_base/pd_multi_base.
+
+- If it was impossible, it will try to prune and freeze tuples on a page.
+
+- If this is unsuccessful it will throw an error. Normally this is very
+unlikely but if there is a very old living transaction with an age of around
+2^32 this can arise. Basically, this is a behavior similar to one during the
+vacuum to prevent wraparound when XID was 32-bit. Dba should take care and
+avoid very-long-living transactions with an age close to 2^32. So long-living
+transactions often they are most likely defunct.
+
+Insert with 64-bit XIDs and 32-bit t_xmin/t_xmax
+------------------------------------------------
+
+On insert we check if current XID fits a range (5). Otherwise:
+
+- heap_page_prepare_for_xid() will try to increase pd_xid_base for t_xmin will
+not be over MaxShortTransactionId.
+
+- If it is impossible, then it will try to prune and freeze tuples on a page.
+
+Known issue: if pd_xid_base could not be shifted to accommodate a tuple being
+inserted due to a very long-running transaction, we just throw an error. We
+neither try to insert a�tuple into another page nor mark the current page as
+full. So, in this (unlikely) case we will get regular insert errors on the next
+tries to insert to the page 'locked' by this very long-running transaction.
+
+Upgrade from 32-bit XID versions
+--------------------------------
+
+pg_upgrade doesn't change pages format itself. It is done lazily after.
+
+1. At first heap page read, tuples on a page are repacked to free 16 bytes
+at the end of a page, possibly freeing space from dead tuples.
+
+2A. 16 bytes of pd_special is added if there is a place for it
+
+2B. Page is converted to "Double XMAX" format if there is no place for
+pd_special
+
+3. If a page is in double XMAX format after its first read, and vacuum (or
+micro-vacuum at select query) could prune some tuples and free space for
+pd_special, prune_page will add pd_special and convert page from double XMAX
+to general 64-bit XID page format.
-- 
2.24.3 (Apple Git-128)


--cpok4wp6gsarlzvp--





^ permalink  raw  reply  [nested|flat] 265+ messages in thread

* [PATCH v6] README for 64bit xid
@ 2022-01-10 19:20 Pavel Borisov <[email protected]>
  0 siblings, 0 replies; 265+ messages in thread

From: Pavel Borisov @ 2022-01-10 19:20 UTC (permalink / raw)

Authors:
- Pavel Borisov <[email protected]>
- Maxim Orlov <[email protected]>
- Yura Sokolov <[email protected]> <[email protected]>
---
 src/backend/access/heap/README.XID64 | 128 +++++++++++++++++++++++++++
 1 file changed, 128 insertions(+)
 create mode 100644 src/backend/access/heap/README.XID64

diff --git a/src/backend/access/heap/README.XID64 b/src/backend/access/heap/README.XID64
new file mode 100644
index 00000000000..457ba9b9ef5
--- /dev/null
+++ b/src/backend/access/heap/README.XID64
@@ -0,0 +1,128 @@
+src/backend/access/heap/README.XID64
+
+64-bit Transaction ID's (XID)
+=============================
+
+A limited number (N = 2^32) of XID's required to do vacuum freeze to prevent
+wraparound every N/2 transactions. This causes performance degradation due
+to the need to exclusively lock tables while being vacuumed. In each
+wraparound cycle, SLRU buffers are also being cut.
+
+With 64-bit XID's wraparound is effectively postponed to a very distant
+future. Even in highly loaded systems that had 2^32 transactions per day
+it will take huge 2^31 days before the first enforced "vacuum to prevent
+wraparound"). Buffers cutting and routine vacuum are not enforced, and DBA
+can plan them independently at the time with the least system load and least
+critical for database performance. Also, it can be done less frequently
+(several times a year vs every several days) on systems with transaction rates
+similar to those mentioned above.
+
+On-disk tuple and page format
+-----------------------------
+
+On-disk tuple format remains unchanged. 32-bit t_xmin and t_xmax store the
+lower parts of 64-bit XMIN and XMAX values. Each heap page has additional
+64-bit pd_xid_base and pd_multi_base which are common for all tuples on a page.
+They are placed into a pd_special area - 16 bytes in the end of a heap page.
+Actual XMIN/XMAX for a tuple are calculated upon reading a tuple from a page
+as follows:
+
+XMIN = t_xmin + pd_xid_base. 					(1)
+XMAX = t_xmax + pd_xid_base/pd_multi_base.		(2)
+
+"Double XMAX" page format
+---------------------------------
+
+At first read of a heap page after pg_upgrade from 32-bit XID PostgreSQL
+version pd_special area with a size of 16 bytes should be added to a page.
+Though a page may not have space for this. Then it can be converted to a
+temporary format called "double XMAX".
+
+All tuples after pg-upgrade would necessarily have xmin = FrozenTransactionId.
+So we don't need tuple header t_xmin field and we reuse t_xmin to store higher
+32 bits of its XMAX.
+
+Double XMAX format is only for full pages that don't have 16 bytes for
+pd_special. So it neither has a�place for a single tuple. Insert and HOT update
+for double XMAX pages is impossible and not supported. We can only read or
+delete tuples from it.
+
+When we are able to prune page double XMAX it will be converted from it to
+general 64-bit XID page format with all operations on its tuples supported.
+
+In-memory tuple format
+----------------------
+
+In-memory tuple representation consists of two parts:
+- HeapTupleHeader from disk page (contains all heap tuple contents, not only
+header)
+- HeapTuple with additional in-memory fields
+
+HeapTuple for each tuple in memory stores t_xid_base/t_multi_base - a copies of
+page's pd_xid_base/pd_multi_base. With tuple's 32-bit t_xmin and t_xmax from
+HeapTupleHeader they are used to calculate actual 64-bit XMIN and XMAX:
+
+XMIN = t_xmin + t_xid_base. 					(3)
+XMAX = t_xmax + t_xid_base/t_multi_base.		(4)
+
+The downside of this is that we can not use tuple's XMIN and XMAX right away.
+We often need to re-read t_xmin and t_xmax - which could actually be pointers
+into a page in shared buffers and therefore they could be updated by any other
+backend.
+
+Update/delete with 64-bit XIDs and 32-bit t_xmin/t_xmax
+--------------------------------------------------------------
+
+When we try to delete/update a tuple, we check that XMAX for a page fits (2).
+I.e. that t_xmax will not be over MaxShortTransactionId relative to
+pd_xid_base/pd_multi_base of a its page.
+
+If the current XID doesn't fit a range
+(pd_xid_base, pd_xid_base + MaxShortTransactionId) (5):
+
+- heap_page_prepare_for_xid() will try to increase pd_xid_base/pd_multi_base on
+a page and update all t_xmin/t_xmax of the other tuples on the page to
+correspond new pd_xid_base/pd_multi_base.
+
+- If it was impossible, it will try to prune and freeze tuples on a page.
+
+- If this is unsuccessful it will throw an error. Normally this is very
+unlikely but if there is a very old living transaction with an age of around
+2^32 this can arise. Basically, this is a behavior similar to one during the
+vacuum to prevent wraparound when XID was 32-bit. Dba should take care and
+avoid very-long-living transactions with an age close to 2^32. So long-living
+transactions often they are most likely defunct.
+
+Insert with 64-bit XIDs and 32-bit t_xmin/t_xmax
+------------------------------------------------
+
+On insert we check if current XID fits a range (5). Otherwise:
+
+- heap_page_prepare_for_xid() will try to increase pd_xid_base for t_xmin will
+not be over MaxShortTransactionId.
+
+- If it is impossible, then it will try to prune and freeze tuples on a page.
+
+Known issue: if pd_xid_base could not be shifted to accommodate a tuple being
+inserted due to a very long-running transaction, we just throw an error. We
+neither try to insert a�tuple into another page nor mark the current page as
+full. So, in this (unlikely) case we will get regular insert errors on the next
+tries to insert to the page 'locked' by this very long-running transaction.
+
+Upgrade from 32-bit XID versions
+--------------------------------
+
+pg_upgrade doesn't change pages format itself. It is done lazily after.
+
+1. At first heap page read, tuples on a page are repacked to free 16 bytes
+at the end of a page, possibly freeing space from dead tuples.
+
+2A. 16 bytes of pd_special is added if there is a place for it
+
+2B. Page is converted to "Double XMAX" format if there is no place for
+pd_special
+
+3. If a page is in double XMAX format after its first read, and vacuum (or
+micro-vacuum at select query) could prune some tuples and free space for
+pd_special, prune_page will add pd_special and convert page from double XMAX
+to general 64-bit XID page format.
-- 
2.24.3 (Apple Git-128)


--cpok4wp6gsarlzvp--





^ permalink  raw  reply  [nested|flat] 265+ messages in thread

* [PATCH v6] README for 64bit xid
@ 2022-01-10 19:20 Pavel Borisov <[email protected]>
  0 siblings, 0 replies; 265+ messages in thread

From: Pavel Borisov @ 2022-01-10 19:20 UTC (permalink / raw)

Authors:
- Pavel Borisov <[email protected]>
- Maxim Orlov <[email protected]>
- Yura Sokolov <[email protected]> <[email protected]>
---
 src/backend/access/heap/README.XID64 | 128 +++++++++++++++++++++++++++
 1 file changed, 128 insertions(+)
 create mode 100644 src/backend/access/heap/README.XID64

diff --git a/src/backend/access/heap/README.XID64 b/src/backend/access/heap/README.XID64
new file mode 100644
index 00000000000..457ba9b9ef5
--- /dev/null
+++ b/src/backend/access/heap/README.XID64
@@ -0,0 +1,128 @@
+src/backend/access/heap/README.XID64
+
+64-bit Transaction ID's (XID)
+=============================
+
+A limited number (N = 2^32) of XID's required to do vacuum freeze to prevent
+wraparound every N/2 transactions. This causes performance degradation due
+to the need to exclusively lock tables while being vacuumed. In each
+wraparound cycle, SLRU buffers are also being cut.
+
+With 64-bit XID's wraparound is effectively postponed to a very distant
+future. Even in highly loaded systems that had 2^32 transactions per day
+it will take huge 2^31 days before the first enforced "vacuum to prevent
+wraparound"). Buffers cutting and routine vacuum are not enforced, and DBA
+can plan them independently at the time with the least system load and least
+critical for database performance. Also, it can be done less frequently
+(several times a year vs every several days) on systems with transaction rates
+similar to those mentioned above.
+
+On-disk tuple and page format
+-----------------------------
+
+On-disk tuple format remains unchanged. 32-bit t_xmin and t_xmax store the
+lower parts of 64-bit XMIN and XMAX values. Each heap page has additional
+64-bit pd_xid_base and pd_multi_base which are common for all tuples on a page.
+They are placed into a pd_special area - 16 bytes in the end of a heap page.
+Actual XMIN/XMAX for a tuple are calculated upon reading a tuple from a page
+as follows:
+
+XMIN = t_xmin + pd_xid_base. 					(1)
+XMAX = t_xmax + pd_xid_base/pd_multi_base.		(2)
+
+"Double XMAX" page format
+---------------------------------
+
+At first read of a heap page after pg_upgrade from 32-bit XID PostgreSQL
+version pd_special area with a size of 16 bytes should be added to a page.
+Though a page may not have space for this. Then it can be converted to a
+temporary format called "double XMAX".
+
+All tuples after pg-upgrade would necessarily have xmin = FrozenTransactionId.
+So we don't need tuple header t_xmin field and we reuse t_xmin to store higher
+32 bits of its XMAX.
+
+Double XMAX format is only for full pages that don't have 16 bytes for
+pd_special. So it neither has a�place for a single tuple. Insert and HOT update
+for double XMAX pages is impossible and not supported. We can only read or
+delete tuples from it.
+
+When we are able to prune page double XMAX it will be converted from it to
+general 64-bit XID page format with all operations on its tuples supported.
+
+In-memory tuple format
+----------------------
+
+In-memory tuple representation consists of two parts:
+- HeapTupleHeader from disk page (contains all heap tuple contents, not only
+header)
+- HeapTuple with additional in-memory fields
+
+HeapTuple for each tuple in memory stores t_xid_base/t_multi_base - a copies of
+page's pd_xid_base/pd_multi_base. With tuple's 32-bit t_xmin and t_xmax from
+HeapTupleHeader they are used to calculate actual 64-bit XMIN and XMAX:
+
+XMIN = t_xmin + t_xid_base. 					(3)
+XMAX = t_xmax + t_xid_base/t_multi_base.		(4)
+
+The downside of this is that we can not use tuple's XMIN and XMAX right away.
+We often need to re-read t_xmin and t_xmax - which could actually be pointers
+into a page in shared buffers and therefore they could be updated by any other
+backend.
+
+Update/delete with 64-bit XIDs and 32-bit t_xmin/t_xmax
+--------------------------------------------------------------
+
+When we try to delete/update a tuple, we check that XMAX for a page fits (2).
+I.e. that t_xmax will not be over MaxShortTransactionId relative to
+pd_xid_base/pd_multi_base of a its page.
+
+If the current XID doesn't fit a range
+(pd_xid_base, pd_xid_base + MaxShortTransactionId) (5):
+
+- heap_page_prepare_for_xid() will try to increase pd_xid_base/pd_multi_base on
+a page and update all t_xmin/t_xmax of the other tuples on the page to
+correspond new pd_xid_base/pd_multi_base.
+
+- If it was impossible, it will try to prune and freeze tuples on a page.
+
+- If this is unsuccessful it will throw an error. Normally this is very
+unlikely but if there is a very old living transaction with an age of around
+2^32 this can arise. Basically, this is a behavior similar to one during the
+vacuum to prevent wraparound when XID was 32-bit. Dba should take care and
+avoid very-long-living transactions with an age close to 2^32. So long-living
+transactions often they are most likely defunct.
+
+Insert with 64-bit XIDs and 32-bit t_xmin/t_xmax
+------------------------------------------------
+
+On insert we check if current XID fits a range (5). Otherwise:
+
+- heap_page_prepare_for_xid() will try to increase pd_xid_base for t_xmin will
+not be over MaxShortTransactionId.
+
+- If it is impossible, then it will try to prune and freeze tuples on a page.
+
+Known issue: if pd_xid_base could not be shifted to accommodate a tuple being
+inserted due to a very long-running transaction, we just throw an error. We
+neither try to insert a�tuple into another page nor mark the current page as
+full. So, in this (unlikely) case we will get regular insert errors on the next
+tries to insert to the page 'locked' by this very long-running transaction.
+
+Upgrade from 32-bit XID versions
+--------------------------------
+
+pg_upgrade doesn't change pages format itself. It is done lazily after.
+
+1. At first heap page read, tuples on a page are repacked to free 16 bytes
+at the end of a page, possibly freeing space from dead tuples.
+
+2A. 16 bytes of pd_special is added if there is a place for it
+
+2B. Page is converted to "Double XMAX" format if there is no place for
+pd_special
+
+3. If a page is in double XMAX format after its first read, and vacuum (or
+micro-vacuum at select query) could prune some tuples and free space for
+pd_special, prune_page will add pd_special and convert page from double XMAX
+to general 64-bit XID page format.
-- 
2.24.3 (Apple Git-128)


--cpok4wp6gsarlzvp--





^ permalink  raw  reply  [nested|flat] 265+ messages in thread

* [PATCH v6] README for 64bit xid
@ 2022-01-10 19:20 Pavel Borisov <[email protected]>
  0 siblings, 0 replies; 265+ messages in thread

From: Pavel Borisov @ 2022-01-10 19:20 UTC (permalink / raw)

Authors:
- Pavel Borisov <[email protected]>
- Maxim Orlov <[email protected]>
- Yura Sokolov <[email protected]> <[email protected]>
---
 src/backend/access/heap/README.XID64 | 128 +++++++++++++++++++++++++++
 1 file changed, 128 insertions(+)
 create mode 100644 src/backend/access/heap/README.XID64

diff --git a/src/backend/access/heap/README.XID64 b/src/backend/access/heap/README.XID64
new file mode 100644
index 00000000000..457ba9b9ef5
--- /dev/null
+++ b/src/backend/access/heap/README.XID64
@@ -0,0 +1,128 @@
+src/backend/access/heap/README.XID64
+
+64-bit Transaction ID's (XID)
+=============================
+
+A limited number (N = 2^32) of XID's required to do vacuum freeze to prevent
+wraparound every N/2 transactions. This causes performance degradation due
+to the need to exclusively lock tables while being vacuumed. In each
+wraparound cycle, SLRU buffers are also being cut.
+
+With 64-bit XID's wraparound is effectively postponed to a very distant
+future. Even in highly loaded systems that had 2^32 transactions per day
+it will take huge 2^31 days before the first enforced "vacuum to prevent
+wraparound"). Buffers cutting and routine vacuum are not enforced, and DBA
+can plan them independently at the time with the least system load and least
+critical for database performance. Also, it can be done less frequently
+(several times a year vs every several days) on systems with transaction rates
+similar to those mentioned above.
+
+On-disk tuple and page format
+-----------------------------
+
+On-disk tuple format remains unchanged. 32-bit t_xmin and t_xmax store the
+lower parts of 64-bit XMIN and XMAX values. Each heap page has additional
+64-bit pd_xid_base and pd_multi_base which are common for all tuples on a page.
+They are placed into a pd_special area - 16 bytes in the end of a heap page.
+Actual XMIN/XMAX for a tuple are calculated upon reading a tuple from a page
+as follows:
+
+XMIN = t_xmin + pd_xid_base. 					(1)
+XMAX = t_xmax + pd_xid_base/pd_multi_base.		(2)
+
+"Double XMAX" page format
+---------------------------------
+
+At first read of a heap page after pg_upgrade from 32-bit XID PostgreSQL
+version pd_special area with a size of 16 bytes should be added to a page.
+Though a page may not have space for this. Then it can be converted to a
+temporary format called "double XMAX".
+
+All tuples after pg-upgrade would necessarily have xmin = FrozenTransactionId.
+So we don't need tuple header t_xmin field and we reuse t_xmin to store higher
+32 bits of its XMAX.
+
+Double XMAX format is only for full pages that don't have 16 bytes for
+pd_special. So it neither has a�place for a single tuple. Insert and HOT update
+for double XMAX pages is impossible and not supported. We can only read or
+delete tuples from it.
+
+When we are able to prune page double XMAX it will be converted from it to
+general 64-bit XID page format with all operations on its tuples supported.
+
+In-memory tuple format
+----------------------
+
+In-memory tuple representation consists of two parts:
+- HeapTupleHeader from disk page (contains all heap tuple contents, not only
+header)
+- HeapTuple with additional in-memory fields
+
+HeapTuple for each tuple in memory stores t_xid_base/t_multi_base - a copies of
+page's pd_xid_base/pd_multi_base. With tuple's 32-bit t_xmin and t_xmax from
+HeapTupleHeader they are used to calculate actual 64-bit XMIN and XMAX:
+
+XMIN = t_xmin + t_xid_base. 					(3)
+XMAX = t_xmax + t_xid_base/t_multi_base.		(4)
+
+The downside of this is that we can not use tuple's XMIN and XMAX right away.
+We often need to re-read t_xmin and t_xmax - which could actually be pointers
+into a page in shared buffers and therefore they could be updated by any other
+backend.
+
+Update/delete with 64-bit XIDs and 32-bit t_xmin/t_xmax
+--------------------------------------------------------------
+
+When we try to delete/update a tuple, we check that XMAX for a page fits (2).
+I.e. that t_xmax will not be over MaxShortTransactionId relative to
+pd_xid_base/pd_multi_base of a its page.
+
+If the current XID doesn't fit a range
+(pd_xid_base, pd_xid_base + MaxShortTransactionId) (5):
+
+- heap_page_prepare_for_xid() will try to increase pd_xid_base/pd_multi_base on
+a page and update all t_xmin/t_xmax of the other tuples on the page to
+correspond new pd_xid_base/pd_multi_base.
+
+- If it was impossible, it will try to prune and freeze tuples on a page.
+
+- If this is unsuccessful it will throw an error. Normally this is very
+unlikely but if there is a very old living transaction with an age of around
+2^32 this can arise. Basically, this is a behavior similar to one during the
+vacuum to prevent wraparound when XID was 32-bit. Dba should take care and
+avoid very-long-living transactions with an age close to 2^32. So long-living
+transactions often they are most likely defunct.
+
+Insert with 64-bit XIDs and 32-bit t_xmin/t_xmax
+------------------------------------------------
+
+On insert we check if current XID fits a range (5). Otherwise:
+
+- heap_page_prepare_for_xid() will try to increase pd_xid_base for t_xmin will
+not be over MaxShortTransactionId.
+
+- If it is impossible, then it will try to prune and freeze tuples on a page.
+
+Known issue: if pd_xid_base could not be shifted to accommodate a tuple being
+inserted due to a very long-running transaction, we just throw an error. We
+neither try to insert a�tuple into another page nor mark the current page as
+full. So, in this (unlikely) case we will get regular insert errors on the next
+tries to insert to the page 'locked' by this very long-running transaction.
+
+Upgrade from 32-bit XID versions
+--------------------------------
+
+pg_upgrade doesn't change pages format itself. It is done lazily after.
+
+1. At first heap page read, tuples on a page are repacked to free 16 bytes
+at the end of a page, possibly freeing space from dead tuples.
+
+2A. 16 bytes of pd_special is added if there is a place for it
+
+2B. Page is converted to "Double XMAX" format if there is no place for
+pd_special
+
+3. If a page is in double XMAX format after its first read, and vacuum (or
+micro-vacuum at select query) could prune some tuples and free space for
+pd_special, prune_page will add pd_special and convert page from double XMAX
+to general 64-bit XID page format.
-- 
2.24.3 (Apple Git-128)


--cpok4wp6gsarlzvp--





^ permalink  raw  reply  [nested|flat] 265+ messages in thread

* [PATCH v6] README for 64bit xid
@ 2022-01-10 19:20 Pavel Borisov <[email protected]>
  0 siblings, 0 replies; 265+ messages in thread

From: Pavel Borisov @ 2022-01-10 19:20 UTC (permalink / raw)

Authors:
- Pavel Borisov <[email protected]>
- Maxim Orlov <[email protected]>
- Yura Sokolov <[email protected]> <[email protected]>
---
 src/backend/access/heap/README.XID64 | 128 +++++++++++++++++++++++++++
 1 file changed, 128 insertions(+)
 create mode 100644 src/backend/access/heap/README.XID64

diff --git a/src/backend/access/heap/README.XID64 b/src/backend/access/heap/README.XID64
new file mode 100644
index 00000000000..457ba9b9ef5
--- /dev/null
+++ b/src/backend/access/heap/README.XID64
@@ -0,0 +1,128 @@
+src/backend/access/heap/README.XID64
+
+64-bit Transaction ID's (XID)
+=============================
+
+A limited number (N = 2^32) of XID's required to do vacuum freeze to prevent
+wraparound every N/2 transactions. This causes performance degradation due
+to the need to exclusively lock tables while being vacuumed. In each
+wraparound cycle, SLRU buffers are also being cut.
+
+With 64-bit XID's wraparound is effectively postponed to a very distant
+future. Even in highly loaded systems that had 2^32 transactions per day
+it will take huge 2^31 days before the first enforced "vacuum to prevent
+wraparound"). Buffers cutting and routine vacuum are not enforced, and DBA
+can plan them independently at the time with the least system load and least
+critical for database performance. Also, it can be done less frequently
+(several times a year vs every several days) on systems with transaction rates
+similar to those mentioned above.
+
+On-disk tuple and page format
+-----------------------------
+
+On-disk tuple format remains unchanged. 32-bit t_xmin and t_xmax store the
+lower parts of 64-bit XMIN and XMAX values. Each heap page has additional
+64-bit pd_xid_base and pd_multi_base which are common for all tuples on a page.
+They are placed into a pd_special area - 16 bytes in the end of a heap page.
+Actual XMIN/XMAX for a tuple are calculated upon reading a tuple from a page
+as follows:
+
+XMIN = t_xmin + pd_xid_base. 					(1)
+XMAX = t_xmax + pd_xid_base/pd_multi_base.		(2)
+
+"Double XMAX" page format
+---------------------------------
+
+At first read of a heap page after pg_upgrade from 32-bit XID PostgreSQL
+version pd_special area with a size of 16 bytes should be added to a page.
+Though a page may not have space for this. Then it can be converted to a
+temporary format called "double XMAX".
+
+All tuples after pg-upgrade would necessarily have xmin = FrozenTransactionId.
+So we don't need tuple header t_xmin field and we reuse t_xmin to store higher
+32 bits of its XMAX.
+
+Double XMAX format is only for full pages that don't have 16 bytes for
+pd_special. So it neither has a�place for a single tuple. Insert and HOT update
+for double XMAX pages is impossible and not supported. We can only read or
+delete tuples from it.
+
+When we are able to prune page double XMAX it will be converted from it to
+general 64-bit XID page format with all operations on its tuples supported.
+
+In-memory tuple format
+----------------------
+
+In-memory tuple representation consists of two parts:
+- HeapTupleHeader from disk page (contains all heap tuple contents, not only
+header)
+- HeapTuple with additional in-memory fields
+
+HeapTuple for each tuple in memory stores t_xid_base/t_multi_base - a copies of
+page's pd_xid_base/pd_multi_base. With tuple's 32-bit t_xmin and t_xmax from
+HeapTupleHeader they are used to calculate actual 64-bit XMIN and XMAX:
+
+XMIN = t_xmin + t_xid_base. 					(3)
+XMAX = t_xmax + t_xid_base/t_multi_base.		(4)
+
+The downside of this is that we can not use tuple's XMIN and XMAX right away.
+We often need to re-read t_xmin and t_xmax - which could actually be pointers
+into a page in shared buffers and therefore they could be updated by any other
+backend.
+
+Update/delete with 64-bit XIDs and 32-bit t_xmin/t_xmax
+--------------------------------------------------------------
+
+When we try to delete/update a tuple, we check that XMAX for a page fits (2).
+I.e. that t_xmax will not be over MaxShortTransactionId relative to
+pd_xid_base/pd_multi_base of a its page.
+
+If the current XID doesn't fit a range
+(pd_xid_base, pd_xid_base + MaxShortTransactionId) (5):
+
+- heap_page_prepare_for_xid() will try to increase pd_xid_base/pd_multi_base on
+a page and update all t_xmin/t_xmax of the other tuples on the page to
+correspond new pd_xid_base/pd_multi_base.
+
+- If it was impossible, it will try to prune and freeze tuples on a page.
+
+- If this is unsuccessful it will throw an error. Normally this is very
+unlikely but if there is a very old living transaction with an age of around
+2^32 this can arise. Basically, this is a behavior similar to one during the
+vacuum to prevent wraparound when XID was 32-bit. Dba should take care and
+avoid very-long-living transactions with an age close to 2^32. So long-living
+transactions often they are most likely defunct.
+
+Insert with 64-bit XIDs and 32-bit t_xmin/t_xmax
+------------------------------------------------
+
+On insert we check if current XID fits a range (5). Otherwise:
+
+- heap_page_prepare_for_xid() will try to increase pd_xid_base for t_xmin will
+not be over MaxShortTransactionId.
+
+- If it is impossible, then it will try to prune and freeze tuples on a page.
+
+Known issue: if pd_xid_base could not be shifted to accommodate a tuple being
+inserted due to a very long-running transaction, we just throw an error. We
+neither try to insert a�tuple into another page nor mark the current page as
+full. So, in this (unlikely) case we will get regular insert errors on the next
+tries to insert to the page 'locked' by this very long-running transaction.
+
+Upgrade from 32-bit XID versions
+--------------------------------
+
+pg_upgrade doesn't change pages format itself. It is done lazily after.
+
+1. At first heap page read, tuples on a page are repacked to free 16 bytes
+at the end of a page, possibly freeing space from dead tuples.
+
+2A. 16 bytes of pd_special is added if there is a place for it
+
+2B. Page is converted to "Double XMAX" format if there is no place for
+pd_special
+
+3. If a page is in double XMAX format after its first read, and vacuum (or
+micro-vacuum at select query) could prune some tuples and free space for
+pd_special, prune_page will add pd_special and convert page from double XMAX
+to general 64-bit XID page format.
-- 
2.24.3 (Apple Git-128)


--cpok4wp6gsarlzvp--





^ permalink  raw  reply  [nested|flat] 265+ messages in thread

* [PATCH v6] README for 64bit xid
@ 2022-01-10 19:20 Pavel Borisov <[email protected]>
  0 siblings, 0 replies; 265+ messages in thread

From: Pavel Borisov @ 2022-01-10 19:20 UTC (permalink / raw)

Authors:
- Pavel Borisov <[email protected]>
- Maxim Orlov <[email protected]>
- Yura Sokolov <[email protected]> <[email protected]>
---
 src/backend/access/heap/README.XID64 | 128 +++++++++++++++++++++++++++
 1 file changed, 128 insertions(+)
 create mode 100644 src/backend/access/heap/README.XID64

diff --git a/src/backend/access/heap/README.XID64 b/src/backend/access/heap/README.XID64
new file mode 100644
index 00000000000..457ba9b9ef5
--- /dev/null
+++ b/src/backend/access/heap/README.XID64
@@ -0,0 +1,128 @@
+src/backend/access/heap/README.XID64
+
+64-bit Transaction ID's (XID)
+=============================
+
+A limited number (N = 2^32) of XID's required to do vacuum freeze to prevent
+wraparound every N/2 transactions. This causes performance degradation due
+to the need to exclusively lock tables while being vacuumed. In each
+wraparound cycle, SLRU buffers are also being cut.
+
+With 64-bit XID's wraparound is effectively postponed to a very distant
+future. Even in highly loaded systems that had 2^32 transactions per day
+it will take huge 2^31 days before the first enforced "vacuum to prevent
+wraparound"). Buffers cutting and routine vacuum are not enforced, and DBA
+can plan them independently at the time with the least system load and least
+critical for database performance. Also, it can be done less frequently
+(several times a year vs every several days) on systems with transaction rates
+similar to those mentioned above.
+
+On-disk tuple and page format
+-----------------------------
+
+On-disk tuple format remains unchanged. 32-bit t_xmin and t_xmax store the
+lower parts of 64-bit XMIN and XMAX values. Each heap page has additional
+64-bit pd_xid_base and pd_multi_base which are common for all tuples on a page.
+They are placed into a pd_special area - 16 bytes in the end of a heap page.
+Actual XMIN/XMAX for a tuple are calculated upon reading a tuple from a page
+as follows:
+
+XMIN = t_xmin + pd_xid_base. 					(1)
+XMAX = t_xmax + pd_xid_base/pd_multi_base.		(2)
+
+"Double XMAX" page format
+---------------------------------
+
+At first read of a heap page after pg_upgrade from 32-bit XID PostgreSQL
+version pd_special area with a size of 16 bytes should be added to a page.
+Though a page may not have space for this. Then it can be converted to a
+temporary format called "double XMAX".
+
+All tuples after pg-upgrade would necessarily have xmin = FrozenTransactionId.
+So we don't need tuple header t_xmin field and we reuse t_xmin to store higher
+32 bits of its XMAX.
+
+Double XMAX format is only for full pages that don't have 16 bytes for
+pd_special. So it neither has a�place for a single tuple. Insert and HOT update
+for double XMAX pages is impossible and not supported. We can only read or
+delete tuples from it.
+
+When we are able to prune page double XMAX it will be converted from it to
+general 64-bit XID page format with all operations on its tuples supported.
+
+In-memory tuple format
+----------------------
+
+In-memory tuple representation consists of two parts:
+- HeapTupleHeader from disk page (contains all heap tuple contents, not only
+header)
+- HeapTuple with additional in-memory fields
+
+HeapTuple for each tuple in memory stores t_xid_base/t_multi_base - a copies of
+page's pd_xid_base/pd_multi_base. With tuple's 32-bit t_xmin and t_xmax from
+HeapTupleHeader they are used to calculate actual 64-bit XMIN and XMAX:
+
+XMIN = t_xmin + t_xid_base. 					(3)
+XMAX = t_xmax + t_xid_base/t_multi_base.		(4)
+
+The downside of this is that we can not use tuple's XMIN and XMAX right away.
+We often need to re-read t_xmin and t_xmax - which could actually be pointers
+into a page in shared buffers and therefore they could be updated by any other
+backend.
+
+Update/delete with 64-bit XIDs and 32-bit t_xmin/t_xmax
+--------------------------------------------------------------
+
+When we try to delete/update a tuple, we check that XMAX for a page fits (2).
+I.e. that t_xmax will not be over MaxShortTransactionId relative to
+pd_xid_base/pd_multi_base of a its page.
+
+If the current XID doesn't fit a range
+(pd_xid_base, pd_xid_base + MaxShortTransactionId) (5):
+
+- heap_page_prepare_for_xid() will try to increase pd_xid_base/pd_multi_base on
+a page and update all t_xmin/t_xmax of the other tuples on the page to
+correspond new pd_xid_base/pd_multi_base.
+
+- If it was impossible, it will try to prune and freeze tuples on a page.
+
+- If this is unsuccessful it will throw an error. Normally this is very
+unlikely but if there is a very old living transaction with an age of around
+2^32 this can arise. Basically, this is a behavior similar to one during the
+vacuum to prevent wraparound when XID was 32-bit. Dba should take care and
+avoid very-long-living transactions with an age close to 2^32. So long-living
+transactions often they are most likely defunct.
+
+Insert with 64-bit XIDs and 32-bit t_xmin/t_xmax
+------------------------------------------------
+
+On insert we check if current XID fits a range (5). Otherwise:
+
+- heap_page_prepare_for_xid() will try to increase pd_xid_base for t_xmin will
+not be over MaxShortTransactionId.
+
+- If it is impossible, then it will try to prune and freeze tuples on a page.
+
+Known issue: if pd_xid_base could not be shifted to accommodate a tuple being
+inserted due to a very long-running transaction, we just throw an error. We
+neither try to insert a�tuple into another page nor mark the current page as
+full. So, in this (unlikely) case we will get regular insert errors on the next
+tries to insert to the page 'locked' by this very long-running transaction.
+
+Upgrade from 32-bit XID versions
+--------------------------------
+
+pg_upgrade doesn't change pages format itself. It is done lazily after.
+
+1. At first heap page read, tuples on a page are repacked to free 16 bytes
+at the end of a page, possibly freeing space from dead tuples.
+
+2A. 16 bytes of pd_special is added if there is a place for it
+
+2B. Page is converted to "Double XMAX" format if there is no place for
+pd_special
+
+3. If a page is in double XMAX format after its first read, and vacuum (or
+micro-vacuum at select query) could prune some tuples and free space for
+pd_special, prune_page will add pd_special and convert page from double XMAX
+to general 64-bit XID page format.
-- 
2.24.3 (Apple Git-128)


--cpok4wp6gsarlzvp--





^ permalink  raw  reply  [nested|flat] 265+ messages in thread

* [PATCH v6] README for 64bit xid
@ 2022-01-10 19:20 Pavel Borisov <[email protected]>
  0 siblings, 0 replies; 265+ messages in thread

From: Pavel Borisov @ 2022-01-10 19:20 UTC (permalink / raw)

Authors:
- Pavel Borisov <[email protected]>
- Maxim Orlov <[email protected]>
- Yura Sokolov <[email protected]> <[email protected]>
---
 src/backend/access/heap/README.XID64 | 128 +++++++++++++++++++++++++++
 1 file changed, 128 insertions(+)
 create mode 100644 src/backend/access/heap/README.XID64

diff --git a/src/backend/access/heap/README.XID64 b/src/backend/access/heap/README.XID64
new file mode 100644
index 00000000000..457ba9b9ef5
--- /dev/null
+++ b/src/backend/access/heap/README.XID64
@@ -0,0 +1,128 @@
+src/backend/access/heap/README.XID64
+
+64-bit Transaction ID's (XID)
+=============================
+
+A limited number (N = 2^32) of XID's required to do vacuum freeze to prevent
+wraparound every N/2 transactions. This causes performance degradation due
+to the need to exclusively lock tables while being vacuumed. In each
+wraparound cycle, SLRU buffers are also being cut.
+
+With 64-bit XID's wraparound is effectively postponed to a very distant
+future. Even in highly loaded systems that had 2^32 transactions per day
+it will take huge 2^31 days before the first enforced "vacuum to prevent
+wraparound"). Buffers cutting and routine vacuum are not enforced, and DBA
+can plan them independently at the time with the least system load and least
+critical for database performance. Also, it can be done less frequently
+(several times a year vs every several days) on systems with transaction rates
+similar to those mentioned above.
+
+On-disk tuple and page format
+-----------------------------
+
+On-disk tuple format remains unchanged. 32-bit t_xmin and t_xmax store the
+lower parts of 64-bit XMIN and XMAX values. Each heap page has additional
+64-bit pd_xid_base and pd_multi_base which are common for all tuples on a page.
+They are placed into a pd_special area - 16 bytes in the end of a heap page.
+Actual XMIN/XMAX for a tuple are calculated upon reading a tuple from a page
+as follows:
+
+XMIN = t_xmin + pd_xid_base. 					(1)
+XMAX = t_xmax + pd_xid_base/pd_multi_base.		(2)
+
+"Double XMAX" page format
+---------------------------------
+
+At first read of a heap page after pg_upgrade from 32-bit XID PostgreSQL
+version pd_special area with a size of 16 bytes should be added to a page.
+Though a page may not have space for this. Then it can be converted to a
+temporary format called "double XMAX".
+
+All tuples after pg-upgrade would necessarily have xmin = FrozenTransactionId.
+So we don't need tuple header t_xmin field and we reuse t_xmin to store higher
+32 bits of its XMAX.
+
+Double XMAX format is only for full pages that don't have 16 bytes for
+pd_special. So it neither has a�place for a single tuple. Insert and HOT update
+for double XMAX pages is impossible and not supported. We can only read or
+delete tuples from it.
+
+When we are able to prune page double XMAX it will be converted from it to
+general 64-bit XID page format with all operations on its tuples supported.
+
+In-memory tuple format
+----------------------
+
+In-memory tuple representation consists of two parts:
+- HeapTupleHeader from disk page (contains all heap tuple contents, not only
+header)
+- HeapTuple with additional in-memory fields
+
+HeapTuple for each tuple in memory stores t_xid_base/t_multi_base - a copies of
+page's pd_xid_base/pd_multi_base. With tuple's 32-bit t_xmin and t_xmax from
+HeapTupleHeader they are used to calculate actual 64-bit XMIN and XMAX:
+
+XMIN = t_xmin + t_xid_base. 					(3)
+XMAX = t_xmax + t_xid_base/t_multi_base.		(4)
+
+The downside of this is that we can not use tuple's XMIN and XMAX right away.
+We often need to re-read t_xmin and t_xmax - which could actually be pointers
+into a page in shared buffers and therefore they could be updated by any other
+backend.
+
+Update/delete with 64-bit XIDs and 32-bit t_xmin/t_xmax
+--------------------------------------------------------------
+
+When we try to delete/update a tuple, we check that XMAX for a page fits (2).
+I.e. that t_xmax will not be over MaxShortTransactionId relative to
+pd_xid_base/pd_multi_base of a its page.
+
+If the current XID doesn't fit a range
+(pd_xid_base, pd_xid_base + MaxShortTransactionId) (5):
+
+- heap_page_prepare_for_xid() will try to increase pd_xid_base/pd_multi_base on
+a page and update all t_xmin/t_xmax of the other tuples on the page to
+correspond new pd_xid_base/pd_multi_base.
+
+- If it was impossible, it will try to prune and freeze tuples on a page.
+
+- If this is unsuccessful it will throw an error. Normally this is very
+unlikely but if there is a very old living transaction with an age of around
+2^32 this can arise. Basically, this is a behavior similar to one during the
+vacuum to prevent wraparound when XID was 32-bit. Dba should take care and
+avoid very-long-living transactions with an age close to 2^32. So long-living
+transactions often they are most likely defunct.
+
+Insert with 64-bit XIDs and 32-bit t_xmin/t_xmax
+------------------------------------------------
+
+On insert we check if current XID fits a range (5). Otherwise:
+
+- heap_page_prepare_for_xid() will try to increase pd_xid_base for t_xmin will
+not be over MaxShortTransactionId.
+
+- If it is impossible, then it will try to prune and freeze tuples on a page.
+
+Known issue: if pd_xid_base could not be shifted to accommodate a tuple being
+inserted due to a very long-running transaction, we just throw an error. We
+neither try to insert a�tuple into another page nor mark the current page as
+full. So, in this (unlikely) case we will get regular insert errors on the next
+tries to insert to the page 'locked' by this very long-running transaction.
+
+Upgrade from 32-bit XID versions
+--------------------------------
+
+pg_upgrade doesn't change pages format itself. It is done lazily after.
+
+1. At first heap page read, tuples on a page are repacked to free 16 bytes
+at the end of a page, possibly freeing space from dead tuples.
+
+2A. 16 bytes of pd_special is added if there is a place for it
+
+2B. Page is converted to "Double XMAX" format if there is no place for
+pd_special
+
+3. If a page is in double XMAX format after its first read, and vacuum (or
+micro-vacuum at select query) could prune some tuples and free space for
+pd_special, prune_page will add pd_special and convert page from double XMAX
+to general 64-bit XID page format.
-- 
2.24.3 (Apple Git-128)


--cpok4wp6gsarlzvp--





^ permalink  raw  reply  [nested|flat] 265+ messages in thread

* [PATCH v6] README for 64bit xid
@ 2022-01-10 19:20 Pavel Borisov <[email protected]>
  0 siblings, 0 replies; 265+ messages in thread

From: Pavel Borisov @ 2022-01-10 19:20 UTC (permalink / raw)

Authors:
- Pavel Borisov <[email protected]>
- Maxim Orlov <[email protected]>
- Yura Sokolov <[email protected]> <[email protected]>
---
 src/backend/access/heap/README.XID64 | 128 +++++++++++++++++++++++++++
 1 file changed, 128 insertions(+)
 create mode 100644 src/backend/access/heap/README.XID64

diff --git a/src/backend/access/heap/README.XID64 b/src/backend/access/heap/README.XID64
new file mode 100644
index 00000000000..457ba9b9ef5
--- /dev/null
+++ b/src/backend/access/heap/README.XID64
@@ -0,0 +1,128 @@
+src/backend/access/heap/README.XID64
+
+64-bit Transaction ID's (XID)
+=============================
+
+A limited number (N = 2^32) of XID's required to do vacuum freeze to prevent
+wraparound every N/2 transactions. This causes performance degradation due
+to the need to exclusively lock tables while being vacuumed. In each
+wraparound cycle, SLRU buffers are also being cut.
+
+With 64-bit XID's wraparound is effectively postponed to a very distant
+future. Even in highly loaded systems that had 2^32 transactions per day
+it will take huge 2^31 days before the first enforced "vacuum to prevent
+wraparound"). Buffers cutting and routine vacuum are not enforced, and DBA
+can plan them independently at the time with the least system load and least
+critical for database performance. Also, it can be done less frequently
+(several times a year vs every several days) on systems with transaction rates
+similar to those mentioned above.
+
+On-disk tuple and page format
+-----------------------------
+
+On-disk tuple format remains unchanged. 32-bit t_xmin and t_xmax store the
+lower parts of 64-bit XMIN and XMAX values. Each heap page has additional
+64-bit pd_xid_base and pd_multi_base which are common for all tuples on a page.
+They are placed into a pd_special area - 16 bytes in the end of a heap page.
+Actual XMIN/XMAX for a tuple are calculated upon reading a tuple from a page
+as follows:
+
+XMIN = t_xmin + pd_xid_base. 					(1)
+XMAX = t_xmax + pd_xid_base/pd_multi_base.		(2)
+
+"Double XMAX" page format
+---------------------------------
+
+At first read of a heap page after pg_upgrade from 32-bit XID PostgreSQL
+version pd_special area with a size of 16 bytes should be added to a page.
+Though a page may not have space for this. Then it can be converted to a
+temporary format called "double XMAX".
+
+All tuples after pg-upgrade would necessarily have xmin = FrozenTransactionId.
+So we don't need tuple header t_xmin field and we reuse t_xmin to store higher
+32 bits of its XMAX.
+
+Double XMAX format is only for full pages that don't have 16 bytes for
+pd_special. So it neither has a�place for a single tuple. Insert and HOT update
+for double XMAX pages is impossible and not supported. We can only read or
+delete tuples from it.
+
+When we are able to prune page double XMAX it will be converted from it to
+general 64-bit XID page format with all operations on its tuples supported.
+
+In-memory tuple format
+----------------------
+
+In-memory tuple representation consists of two parts:
+- HeapTupleHeader from disk page (contains all heap tuple contents, not only
+header)
+- HeapTuple with additional in-memory fields
+
+HeapTuple for each tuple in memory stores t_xid_base/t_multi_base - a copies of
+page's pd_xid_base/pd_multi_base. With tuple's 32-bit t_xmin and t_xmax from
+HeapTupleHeader they are used to calculate actual 64-bit XMIN and XMAX:
+
+XMIN = t_xmin + t_xid_base. 					(3)
+XMAX = t_xmax + t_xid_base/t_multi_base.		(4)
+
+The downside of this is that we can not use tuple's XMIN and XMAX right away.
+We often need to re-read t_xmin and t_xmax - which could actually be pointers
+into a page in shared buffers and therefore they could be updated by any other
+backend.
+
+Update/delete with 64-bit XIDs and 32-bit t_xmin/t_xmax
+--------------------------------------------------------------
+
+When we try to delete/update a tuple, we check that XMAX for a page fits (2).
+I.e. that t_xmax will not be over MaxShortTransactionId relative to
+pd_xid_base/pd_multi_base of a its page.
+
+If the current XID doesn't fit a range
+(pd_xid_base, pd_xid_base + MaxShortTransactionId) (5):
+
+- heap_page_prepare_for_xid() will try to increase pd_xid_base/pd_multi_base on
+a page and update all t_xmin/t_xmax of the other tuples on the page to
+correspond new pd_xid_base/pd_multi_base.
+
+- If it was impossible, it will try to prune and freeze tuples on a page.
+
+- If this is unsuccessful it will throw an error. Normally this is very
+unlikely but if there is a very old living transaction with an age of around
+2^32 this can arise. Basically, this is a behavior similar to one during the
+vacuum to prevent wraparound when XID was 32-bit. Dba should take care and
+avoid very-long-living transactions with an age close to 2^32. So long-living
+transactions often they are most likely defunct.
+
+Insert with 64-bit XIDs and 32-bit t_xmin/t_xmax
+------------------------------------------------
+
+On insert we check if current XID fits a range (5). Otherwise:
+
+- heap_page_prepare_for_xid() will try to increase pd_xid_base for t_xmin will
+not be over MaxShortTransactionId.
+
+- If it is impossible, then it will try to prune and freeze tuples on a page.
+
+Known issue: if pd_xid_base could not be shifted to accommodate a tuple being
+inserted due to a very long-running transaction, we just throw an error. We
+neither try to insert a�tuple into another page nor mark the current page as
+full. So, in this (unlikely) case we will get regular insert errors on the next
+tries to insert to the page 'locked' by this very long-running transaction.
+
+Upgrade from 32-bit XID versions
+--------------------------------
+
+pg_upgrade doesn't change pages format itself. It is done lazily after.
+
+1. At first heap page read, tuples on a page are repacked to free 16 bytes
+at the end of a page, possibly freeing space from dead tuples.
+
+2A. 16 bytes of pd_special is added if there is a place for it
+
+2B. Page is converted to "Double XMAX" format if there is no place for
+pd_special
+
+3. If a page is in double XMAX format after its first read, and vacuum (or
+micro-vacuum at select query) could prune some tuples and free space for
+pd_special, prune_page will add pd_special and convert page from double XMAX
+to general 64-bit XID page format.
-- 
2.24.3 (Apple Git-128)


--cpok4wp6gsarlzvp--





^ permalink  raw  reply  [nested|flat] 265+ messages in thread

* [PATCH v6] README for 64bit xid
@ 2022-01-10 19:20 Pavel Borisov <[email protected]>
  0 siblings, 0 replies; 265+ messages in thread

From: Pavel Borisov @ 2022-01-10 19:20 UTC (permalink / raw)

Authors:
- Pavel Borisov <[email protected]>
- Maxim Orlov <[email protected]>
- Yura Sokolov <[email protected]> <[email protected]>
---
 src/backend/access/heap/README.XID64 | 128 +++++++++++++++++++++++++++
 1 file changed, 128 insertions(+)
 create mode 100644 src/backend/access/heap/README.XID64

diff --git a/src/backend/access/heap/README.XID64 b/src/backend/access/heap/README.XID64
new file mode 100644
index 00000000000..457ba9b9ef5
--- /dev/null
+++ b/src/backend/access/heap/README.XID64
@@ -0,0 +1,128 @@
+src/backend/access/heap/README.XID64
+
+64-bit Transaction ID's (XID)
+=============================
+
+A limited number (N = 2^32) of XID's required to do vacuum freeze to prevent
+wraparound every N/2 transactions. This causes performance degradation due
+to the need to exclusively lock tables while being vacuumed. In each
+wraparound cycle, SLRU buffers are also being cut.
+
+With 64-bit XID's wraparound is effectively postponed to a very distant
+future. Even in highly loaded systems that had 2^32 transactions per day
+it will take huge 2^31 days before the first enforced "vacuum to prevent
+wraparound"). Buffers cutting and routine vacuum are not enforced, and DBA
+can plan them independently at the time with the least system load and least
+critical for database performance. Also, it can be done less frequently
+(several times a year vs every several days) on systems with transaction rates
+similar to those mentioned above.
+
+On-disk tuple and page format
+-----------------------------
+
+On-disk tuple format remains unchanged. 32-bit t_xmin and t_xmax store the
+lower parts of 64-bit XMIN and XMAX values. Each heap page has additional
+64-bit pd_xid_base and pd_multi_base which are common for all tuples on a page.
+They are placed into a pd_special area - 16 bytes in the end of a heap page.
+Actual XMIN/XMAX for a tuple are calculated upon reading a tuple from a page
+as follows:
+
+XMIN = t_xmin + pd_xid_base. 					(1)
+XMAX = t_xmax + pd_xid_base/pd_multi_base.		(2)
+
+"Double XMAX" page format
+---------------------------------
+
+At first read of a heap page after pg_upgrade from 32-bit XID PostgreSQL
+version pd_special area with a size of 16 bytes should be added to a page.
+Though a page may not have space for this. Then it can be converted to a
+temporary format called "double XMAX".
+
+All tuples after pg-upgrade would necessarily have xmin = FrozenTransactionId.
+So we don't need tuple header t_xmin field and we reuse t_xmin to store higher
+32 bits of its XMAX.
+
+Double XMAX format is only for full pages that don't have 16 bytes for
+pd_special. So it neither has a�place for a single tuple. Insert and HOT update
+for double XMAX pages is impossible and not supported. We can only read or
+delete tuples from it.
+
+When we are able to prune page double XMAX it will be converted from it to
+general 64-bit XID page format with all operations on its tuples supported.
+
+In-memory tuple format
+----------------------
+
+In-memory tuple representation consists of two parts:
+- HeapTupleHeader from disk page (contains all heap tuple contents, not only
+header)
+- HeapTuple with additional in-memory fields
+
+HeapTuple for each tuple in memory stores t_xid_base/t_multi_base - a copies of
+page's pd_xid_base/pd_multi_base. With tuple's 32-bit t_xmin and t_xmax from
+HeapTupleHeader they are used to calculate actual 64-bit XMIN and XMAX:
+
+XMIN = t_xmin + t_xid_base. 					(3)
+XMAX = t_xmax + t_xid_base/t_multi_base.		(4)
+
+The downside of this is that we can not use tuple's XMIN and XMAX right away.
+We often need to re-read t_xmin and t_xmax - which could actually be pointers
+into a page in shared buffers and therefore they could be updated by any other
+backend.
+
+Update/delete with 64-bit XIDs and 32-bit t_xmin/t_xmax
+--------------------------------------------------------------
+
+When we try to delete/update a tuple, we check that XMAX for a page fits (2).
+I.e. that t_xmax will not be over MaxShortTransactionId relative to
+pd_xid_base/pd_multi_base of a its page.
+
+If the current XID doesn't fit a range
+(pd_xid_base, pd_xid_base + MaxShortTransactionId) (5):
+
+- heap_page_prepare_for_xid() will try to increase pd_xid_base/pd_multi_base on
+a page and update all t_xmin/t_xmax of the other tuples on the page to
+correspond new pd_xid_base/pd_multi_base.
+
+- If it was impossible, it will try to prune and freeze tuples on a page.
+
+- If this is unsuccessful it will throw an error. Normally this is very
+unlikely but if there is a very old living transaction with an age of around
+2^32 this can arise. Basically, this is a behavior similar to one during the
+vacuum to prevent wraparound when XID was 32-bit. Dba should take care and
+avoid very-long-living transactions with an age close to 2^32. So long-living
+transactions often they are most likely defunct.
+
+Insert with 64-bit XIDs and 32-bit t_xmin/t_xmax
+------------------------------------------------
+
+On insert we check if current XID fits a range (5). Otherwise:
+
+- heap_page_prepare_for_xid() will try to increase pd_xid_base for t_xmin will
+not be over MaxShortTransactionId.
+
+- If it is impossible, then it will try to prune and freeze tuples on a page.
+
+Known issue: if pd_xid_base could not be shifted to accommodate a tuple being
+inserted due to a very long-running transaction, we just throw an error. We
+neither try to insert a�tuple into another page nor mark the current page as
+full. So, in this (unlikely) case we will get regular insert errors on the next
+tries to insert to the page 'locked' by this very long-running transaction.
+
+Upgrade from 32-bit XID versions
+--------------------------------
+
+pg_upgrade doesn't change pages format itself. It is done lazily after.
+
+1. At first heap page read, tuples on a page are repacked to free 16 bytes
+at the end of a page, possibly freeing space from dead tuples.
+
+2A. 16 bytes of pd_special is added if there is a place for it
+
+2B. Page is converted to "Double XMAX" format if there is no place for
+pd_special
+
+3. If a page is in double XMAX format after its first read, and vacuum (or
+micro-vacuum at select query) could prune some tuples and free space for
+pd_special, prune_page will add pd_special and convert page from double XMAX
+to general 64-bit XID page format.
-- 
2.24.3 (Apple Git-128)


--cpok4wp6gsarlzvp--





^ permalink  raw  reply  [nested|flat] 265+ messages in thread

* [PATCH v6] README for 64bit xid
@ 2022-01-10 19:20 Pavel Borisov <[email protected]>
  0 siblings, 0 replies; 265+ messages in thread

From: Pavel Borisov @ 2022-01-10 19:20 UTC (permalink / raw)

Authors:
- Pavel Borisov <[email protected]>
- Maxim Orlov <[email protected]>
- Yura Sokolov <[email protected]> <[email protected]>
---
 src/backend/access/heap/README.XID64 | 128 +++++++++++++++++++++++++++
 1 file changed, 128 insertions(+)
 create mode 100644 src/backend/access/heap/README.XID64

diff --git a/src/backend/access/heap/README.XID64 b/src/backend/access/heap/README.XID64
new file mode 100644
index 00000000000..457ba9b9ef5
--- /dev/null
+++ b/src/backend/access/heap/README.XID64
@@ -0,0 +1,128 @@
+src/backend/access/heap/README.XID64
+
+64-bit Transaction ID's (XID)
+=============================
+
+A limited number (N = 2^32) of XID's required to do vacuum freeze to prevent
+wraparound every N/2 transactions. This causes performance degradation due
+to the need to exclusively lock tables while being vacuumed. In each
+wraparound cycle, SLRU buffers are also being cut.
+
+With 64-bit XID's wraparound is effectively postponed to a very distant
+future. Even in highly loaded systems that had 2^32 transactions per day
+it will take huge 2^31 days before the first enforced "vacuum to prevent
+wraparound"). Buffers cutting and routine vacuum are not enforced, and DBA
+can plan them independently at the time with the least system load and least
+critical for database performance. Also, it can be done less frequently
+(several times a year vs every several days) on systems with transaction rates
+similar to those mentioned above.
+
+On-disk tuple and page format
+-----------------------------
+
+On-disk tuple format remains unchanged. 32-bit t_xmin and t_xmax store the
+lower parts of 64-bit XMIN and XMAX values. Each heap page has additional
+64-bit pd_xid_base and pd_multi_base which are common for all tuples on a page.
+They are placed into a pd_special area - 16 bytes in the end of a heap page.
+Actual XMIN/XMAX for a tuple are calculated upon reading a tuple from a page
+as follows:
+
+XMIN = t_xmin + pd_xid_base. 					(1)
+XMAX = t_xmax + pd_xid_base/pd_multi_base.		(2)
+
+"Double XMAX" page format
+---------------------------------
+
+At first read of a heap page after pg_upgrade from 32-bit XID PostgreSQL
+version pd_special area with a size of 16 bytes should be added to a page.
+Though a page may not have space for this. Then it can be converted to a
+temporary format called "double XMAX".
+
+All tuples after pg-upgrade would necessarily have xmin = FrozenTransactionId.
+So we don't need tuple header t_xmin field and we reuse t_xmin to store higher
+32 bits of its XMAX.
+
+Double XMAX format is only for full pages that don't have 16 bytes for
+pd_special. So it neither has a�place for a single tuple. Insert and HOT update
+for double XMAX pages is impossible and not supported. We can only read or
+delete tuples from it.
+
+When we are able to prune page double XMAX it will be converted from it to
+general 64-bit XID page format with all operations on its tuples supported.
+
+In-memory tuple format
+----------------------
+
+In-memory tuple representation consists of two parts:
+- HeapTupleHeader from disk page (contains all heap tuple contents, not only
+header)
+- HeapTuple with additional in-memory fields
+
+HeapTuple for each tuple in memory stores t_xid_base/t_multi_base - a copies of
+page's pd_xid_base/pd_multi_base. With tuple's 32-bit t_xmin and t_xmax from
+HeapTupleHeader they are used to calculate actual 64-bit XMIN and XMAX:
+
+XMIN = t_xmin + t_xid_base. 					(3)
+XMAX = t_xmax + t_xid_base/t_multi_base.		(4)
+
+The downside of this is that we can not use tuple's XMIN and XMAX right away.
+We often need to re-read t_xmin and t_xmax - which could actually be pointers
+into a page in shared buffers and therefore they could be updated by any other
+backend.
+
+Update/delete with 64-bit XIDs and 32-bit t_xmin/t_xmax
+--------------------------------------------------------------
+
+When we try to delete/update a tuple, we check that XMAX for a page fits (2).
+I.e. that t_xmax will not be over MaxShortTransactionId relative to
+pd_xid_base/pd_multi_base of a its page.
+
+If the current XID doesn't fit a range
+(pd_xid_base, pd_xid_base + MaxShortTransactionId) (5):
+
+- heap_page_prepare_for_xid() will try to increase pd_xid_base/pd_multi_base on
+a page and update all t_xmin/t_xmax of the other tuples on the page to
+correspond new pd_xid_base/pd_multi_base.
+
+- If it was impossible, it will try to prune and freeze tuples on a page.
+
+- If this is unsuccessful it will throw an error. Normally this is very
+unlikely but if there is a very old living transaction with an age of around
+2^32 this can arise. Basically, this is a behavior similar to one during the
+vacuum to prevent wraparound when XID was 32-bit. Dba should take care and
+avoid very-long-living transactions with an age close to 2^32. So long-living
+transactions often they are most likely defunct.
+
+Insert with 64-bit XIDs and 32-bit t_xmin/t_xmax
+------------------------------------------------
+
+On insert we check if current XID fits a range (5). Otherwise:
+
+- heap_page_prepare_for_xid() will try to increase pd_xid_base for t_xmin will
+not be over MaxShortTransactionId.
+
+- If it is impossible, then it will try to prune and freeze tuples on a page.
+
+Known issue: if pd_xid_base could not be shifted to accommodate a tuple being
+inserted due to a very long-running transaction, we just throw an error. We
+neither try to insert a�tuple into another page nor mark the current page as
+full. So, in this (unlikely) case we will get regular insert errors on the next
+tries to insert to the page 'locked' by this very long-running transaction.
+
+Upgrade from 32-bit XID versions
+--------------------------------
+
+pg_upgrade doesn't change pages format itself. It is done lazily after.
+
+1. At first heap page read, tuples on a page are repacked to free 16 bytes
+at the end of a page, possibly freeing space from dead tuples.
+
+2A. 16 bytes of pd_special is added if there is a place for it
+
+2B. Page is converted to "Double XMAX" format if there is no place for
+pd_special
+
+3. If a page is in double XMAX format after its first read, and vacuum (or
+micro-vacuum at select query) could prune some tuples and free space for
+pd_special, prune_page will add pd_special and convert page from double XMAX
+to general 64-bit XID page format.
-- 
2.24.3 (Apple Git-128)


--cpok4wp6gsarlzvp--





^ permalink  raw  reply  [nested|flat] 265+ messages in thread

* [PATCH v6] README for 64bit xid
@ 2022-01-10 19:20 Pavel Borisov <[email protected]>
  0 siblings, 0 replies; 265+ messages in thread

From: Pavel Borisov @ 2022-01-10 19:20 UTC (permalink / raw)

Authors:
- Pavel Borisov <[email protected]>
- Maxim Orlov <[email protected]>
- Yura Sokolov <[email protected]> <[email protected]>
---
 src/backend/access/heap/README.XID64 | 128 +++++++++++++++++++++++++++
 1 file changed, 128 insertions(+)
 create mode 100644 src/backend/access/heap/README.XID64

diff --git a/src/backend/access/heap/README.XID64 b/src/backend/access/heap/README.XID64
new file mode 100644
index 00000000000..457ba9b9ef5
--- /dev/null
+++ b/src/backend/access/heap/README.XID64
@@ -0,0 +1,128 @@
+src/backend/access/heap/README.XID64
+
+64-bit Transaction ID's (XID)
+=============================
+
+A limited number (N = 2^32) of XID's required to do vacuum freeze to prevent
+wraparound every N/2 transactions. This causes performance degradation due
+to the need to exclusively lock tables while being vacuumed. In each
+wraparound cycle, SLRU buffers are also being cut.
+
+With 64-bit XID's wraparound is effectively postponed to a very distant
+future. Even in highly loaded systems that had 2^32 transactions per day
+it will take huge 2^31 days before the first enforced "vacuum to prevent
+wraparound"). Buffers cutting and routine vacuum are not enforced, and DBA
+can plan them independently at the time with the least system load and least
+critical for database performance. Also, it can be done less frequently
+(several times a year vs every several days) on systems with transaction rates
+similar to those mentioned above.
+
+On-disk tuple and page format
+-----------------------------
+
+On-disk tuple format remains unchanged. 32-bit t_xmin and t_xmax store the
+lower parts of 64-bit XMIN and XMAX values. Each heap page has additional
+64-bit pd_xid_base and pd_multi_base which are common for all tuples on a page.
+They are placed into a pd_special area - 16 bytes in the end of a heap page.
+Actual XMIN/XMAX for a tuple are calculated upon reading a tuple from a page
+as follows:
+
+XMIN = t_xmin + pd_xid_base. 					(1)
+XMAX = t_xmax + pd_xid_base/pd_multi_base.		(2)
+
+"Double XMAX" page format
+---------------------------------
+
+At first read of a heap page after pg_upgrade from 32-bit XID PostgreSQL
+version pd_special area with a size of 16 bytes should be added to a page.
+Though a page may not have space for this. Then it can be converted to a
+temporary format called "double XMAX".
+
+All tuples after pg-upgrade would necessarily have xmin = FrozenTransactionId.
+So we don't need tuple header t_xmin field and we reuse t_xmin to store higher
+32 bits of its XMAX.
+
+Double XMAX format is only for full pages that don't have 16 bytes for
+pd_special. So it neither has a�place for a single tuple. Insert and HOT update
+for double XMAX pages is impossible and not supported. We can only read or
+delete tuples from it.
+
+When we are able to prune page double XMAX it will be converted from it to
+general 64-bit XID page format with all operations on its tuples supported.
+
+In-memory tuple format
+----------------------
+
+In-memory tuple representation consists of two parts:
+- HeapTupleHeader from disk page (contains all heap tuple contents, not only
+header)
+- HeapTuple with additional in-memory fields
+
+HeapTuple for each tuple in memory stores t_xid_base/t_multi_base - a copies of
+page's pd_xid_base/pd_multi_base. With tuple's 32-bit t_xmin and t_xmax from
+HeapTupleHeader they are used to calculate actual 64-bit XMIN and XMAX:
+
+XMIN = t_xmin + t_xid_base. 					(3)
+XMAX = t_xmax + t_xid_base/t_multi_base.		(4)
+
+The downside of this is that we can not use tuple's XMIN and XMAX right away.
+We often need to re-read t_xmin and t_xmax - which could actually be pointers
+into a page in shared buffers and therefore they could be updated by any other
+backend.
+
+Update/delete with 64-bit XIDs and 32-bit t_xmin/t_xmax
+--------------------------------------------------------------
+
+When we try to delete/update a tuple, we check that XMAX for a page fits (2).
+I.e. that t_xmax will not be over MaxShortTransactionId relative to
+pd_xid_base/pd_multi_base of a its page.
+
+If the current XID doesn't fit a range
+(pd_xid_base, pd_xid_base + MaxShortTransactionId) (5):
+
+- heap_page_prepare_for_xid() will try to increase pd_xid_base/pd_multi_base on
+a page and update all t_xmin/t_xmax of the other tuples on the page to
+correspond new pd_xid_base/pd_multi_base.
+
+- If it was impossible, it will try to prune and freeze tuples on a page.
+
+- If this is unsuccessful it will throw an error. Normally this is very
+unlikely but if there is a very old living transaction with an age of around
+2^32 this can arise. Basically, this is a behavior similar to one during the
+vacuum to prevent wraparound when XID was 32-bit. Dba should take care and
+avoid very-long-living transactions with an age close to 2^32. So long-living
+transactions often they are most likely defunct.
+
+Insert with 64-bit XIDs and 32-bit t_xmin/t_xmax
+------------------------------------------------
+
+On insert we check if current XID fits a range (5). Otherwise:
+
+- heap_page_prepare_for_xid() will try to increase pd_xid_base for t_xmin will
+not be over MaxShortTransactionId.
+
+- If it is impossible, then it will try to prune and freeze tuples on a page.
+
+Known issue: if pd_xid_base could not be shifted to accommodate a tuple being
+inserted due to a very long-running transaction, we just throw an error. We
+neither try to insert a�tuple into another page nor mark the current page as
+full. So, in this (unlikely) case we will get regular insert errors on the next
+tries to insert to the page 'locked' by this very long-running transaction.
+
+Upgrade from 32-bit XID versions
+--------------------------------
+
+pg_upgrade doesn't change pages format itself. It is done lazily after.
+
+1. At first heap page read, tuples on a page are repacked to free 16 bytes
+at the end of a page, possibly freeing space from dead tuples.
+
+2A. 16 bytes of pd_special is added if there is a place for it
+
+2B. Page is converted to "Double XMAX" format if there is no place for
+pd_special
+
+3. If a page is in double XMAX format after its first read, and vacuum (or
+micro-vacuum at select query) could prune some tuples and free space for
+pd_special, prune_page will add pd_special and convert page from double XMAX
+to general 64-bit XID page format.
-- 
2.24.3 (Apple Git-128)


--cpok4wp6gsarlzvp--





^ permalink  raw  reply  [nested|flat] 265+ messages in thread

* [PATCH v6] README for 64bit xid
@ 2022-01-10 19:20 Pavel Borisov <[email protected]>
  0 siblings, 0 replies; 265+ messages in thread

From: Pavel Borisov @ 2022-01-10 19:20 UTC (permalink / raw)

Authors:
- Pavel Borisov <[email protected]>
- Maxim Orlov <[email protected]>
- Yura Sokolov <[email protected]> <[email protected]>
---
 src/backend/access/heap/README.XID64 | 128 +++++++++++++++++++++++++++
 1 file changed, 128 insertions(+)
 create mode 100644 src/backend/access/heap/README.XID64

diff --git a/src/backend/access/heap/README.XID64 b/src/backend/access/heap/README.XID64
new file mode 100644
index 00000000000..457ba9b9ef5
--- /dev/null
+++ b/src/backend/access/heap/README.XID64
@@ -0,0 +1,128 @@
+src/backend/access/heap/README.XID64
+
+64-bit Transaction ID's (XID)
+=============================
+
+A limited number (N = 2^32) of XID's required to do vacuum freeze to prevent
+wraparound every N/2 transactions. This causes performance degradation due
+to the need to exclusively lock tables while being vacuumed. In each
+wraparound cycle, SLRU buffers are also being cut.
+
+With 64-bit XID's wraparound is effectively postponed to a very distant
+future. Even in highly loaded systems that had 2^32 transactions per day
+it will take huge 2^31 days before the first enforced "vacuum to prevent
+wraparound"). Buffers cutting and routine vacuum are not enforced, and DBA
+can plan them independently at the time with the least system load and least
+critical for database performance. Also, it can be done less frequently
+(several times a year vs every several days) on systems with transaction rates
+similar to those mentioned above.
+
+On-disk tuple and page format
+-----------------------------
+
+On-disk tuple format remains unchanged. 32-bit t_xmin and t_xmax store the
+lower parts of 64-bit XMIN and XMAX values. Each heap page has additional
+64-bit pd_xid_base and pd_multi_base which are common for all tuples on a page.
+They are placed into a pd_special area - 16 bytes in the end of a heap page.
+Actual XMIN/XMAX for a tuple are calculated upon reading a tuple from a page
+as follows:
+
+XMIN = t_xmin + pd_xid_base. 					(1)
+XMAX = t_xmax + pd_xid_base/pd_multi_base.		(2)
+
+"Double XMAX" page format
+---------------------------------
+
+At first read of a heap page after pg_upgrade from 32-bit XID PostgreSQL
+version pd_special area with a size of 16 bytes should be added to a page.
+Though a page may not have space for this. Then it can be converted to a
+temporary format called "double XMAX".
+
+All tuples after pg-upgrade would necessarily have xmin = FrozenTransactionId.
+So we don't need tuple header t_xmin field and we reuse t_xmin to store higher
+32 bits of its XMAX.
+
+Double XMAX format is only for full pages that don't have 16 bytes for
+pd_special. So it neither has a�place for a single tuple. Insert and HOT update
+for double XMAX pages is impossible and not supported. We can only read or
+delete tuples from it.
+
+When we are able to prune page double XMAX it will be converted from it to
+general 64-bit XID page format with all operations on its tuples supported.
+
+In-memory tuple format
+----------------------
+
+In-memory tuple representation consists of two parts:
+- HeapTupleHeader from disk page (contains all heap tuple contents, not only
+header)
+- HeapTuple with additional in-memory fields
+
+HeapTuple for each tuple in memory stores t_xid_base/t_multi_base - a copies of
+page's pd_xid_base/pd_multi_base. With tuple's 32-bit t_xmin and t_xmax from
+HeapTupleHeader they are used to calculate actual 64-bit XMIN and XMAX:
+
+XMIN = t_xmin + t_xid_base. 					(3)
+XMAX = t_xmax + t_xid_base/t_multi_base.		(4)
+
+The downside of this is that we can not use tuple's XMIN and XMAX right away.
+We often need to re-read t_xmin and t_xmax - which could actually be pointers
+into a page in shared buffers and therefore they could be updated by any other
+backend.
+
+Update/delete with 64-bit XIDs and 32-bit t_xmin/t_xmax
+--------------------------------------------------------------
+
+When we try to delete/update a tuple, we check that XMAX for a page fits (2).
+I.e. that t_xmax will not be over MaxShortTransactionId relative to
+pd_xid_base/pd_multi_base of a its page.
+
+If the current XID doesn't fit a range
+(pd_xid_base, pd_xid_base + MaxShortTransactionId) (5):
+
+- heap_page_prepare_for_xid() will try to increase pd_xid_base/pd_multi_base on
+a page and update all t_xmin/t_xmax of the other tuples on the page to
+correspond new pd_xid_base/pd_multi_base.
+
+- If it was impossible, it will try to prune and freeze tuples on a page.
+
+- If this is unsuccessful it will throw an error. Normally this is very
+unlikely but if there is a very old living transaction with an age of around
+2^32 this can arise. Basically, this is a behavior similar to one during the
+vacuum to prevent wraparound when XID was 32-bit. Dba should take care and
+avoid very-long-living transactions with an age close to 2^32. So long-living
+transactions often they are most likely defunct.
+
+Insert with 64-bit XIDs and 32-bit t_xmin/t_xmax
+------------------------------------------------
+
+On insert we check if current XID fits a range (5). Otherwise:
+
+- heap_page_prepare_for_xid() will try to increase pd_xid_base for t_xmin will
+not be over MaxShortTransactionId.
+
+- If it is impossible, then it will try to prune and freeze tuples on a page.
+
+Known issue: if pd_xid_base could not be shifted to accommodate a tuple being
+inserted due to a very long-running transaction, we just throw an error. We
+neither try to insert a�tuple into another page nor mark the current page as
+full. So, in this (unlikely) case we will get regular insert errors on the next
+tries to insert to the page 'locked' by this very long-running transaction.
+
+Upgrade from 32-bit XID versions
+--------------------------------
+
+pg_upgrade doesn't change pages format itself. It is done lazily after.
+
+1. At first heap page read, tuples on a page are repacked to free 16 bytes
+at the end of a page, possibly freeing space from dead tuples.
+
+2A. 16 bytes of pd_special is added if there is a place for it
+
+2B. Page is converted to "Double XMAX" format if there is no place for
+pd_special
+
+3. If a page is in double XMAX format after its first read, and vacuum (or
+micro-vacuum at select query) could prune some tuples and free space for
+pd_special, prune_page will add pd_special and convert page from double XMAX
+to general 64-bit XID page format.
-- 
2.24.3 (Apple Git-128)


--cpok4wp6gsarlzvp--





^ permalink  raw  reply  [nested|flat] 265+ messages in thread

* [PATCH v6] README for 64bit xid
@ 2022-01-10 19:20 Pavel Borisov <[email protected]>
  0 siblings, 0 replies; 265+ messages in thread

From: Pavel Borisov @ 2022-01-10 19:20 UTC (permalink / raw)

Authors:
- Pavel Borisov <[email protected]>
- Maxim Orlov <[email protected]>
- Yura Sokolov <[email protected]> <[email protected]>
---
 src/backend/access/heap/README.XID64 | 128 +++++++++++++++++++++++++++
 1 file changed, 128 insertions(+)
 create mode 100644 src/backend/access/heap/README.XID64

diff --git a/src/backend/access/heap/README.XID64 b/src/backend/access/heap/README.XID64
new file mode 100644
index 00000000000..457ba9b9ef5
--- /dev/null
+++ b/src/backend/access/heap/README.XID64
@@ -0,0 +1,128 @@
+src/backend/access/heap/README.XID64
+
+64-bit Transaction ID's (XID)
+=============================
+
+A limited number (N = 2^32) of XID's required to do vacuum freeze to prevent
+wraparound every N/2 transactions. This causes performance degradation due
+to the need to exclusively lock tables while being vacuumed. In each
+wraparound cycle, SLRU buffers are also being cut.
+
+With 64-bit XID's wraparound is effectively postponed to a very distant
+future. Even in highly loaded systems that had 2^32 transactions per day
+it will take huge 2^31 days before the first enforced "vacuum to prevent
+wraparound"). Buffers cutting and routine vacuum are not enforced, and DBA
+can plan them independently at the time with the least system load and least
+critical for database performance. Also, it can be done less frequently
+(several times a year vs every several days) on systems with transaction rates
+similar to those mentioned above.
+
+On-disk tuple and page format
+-----------------------------
+
+On-disk tuple format remains unchanged. 32-bit t_xmin and t_xmax store the
+lower parts of 64-bit XMIN and XMAX values. Each heap page has additional
+64-bit pd_xid_base and pd_multi_base which are common for all tuples on a page.
+They are placed into a pd_special area - 16 bytes in the end of a heap page.
+Actual XMIN/XMAX for a tuple are calculated upon reading a tuple from a page
+as follows:
+
+XMIN = t_xmin + pd_xid_base. 					(1)
+XMAX = t_xmax + pd_xid_base/pd_multi_base.		(2)
+
+"Double XMAX" page format
+---------------------------------
+
+At first read of a heap page after pg_upgrade from 32-bit XID PostgreSQL
+version pd_special area with a size of 16 bytes should be added to a page.
+Though a page may not have space for this. Then it can be converted to a
+temporary format called "double XMAX".
+
+All tuples after pg-upgrade would necessarily have xmin = FrozenTransactionId.
+So we don't need tuple header t_xmin field and we reuse t_xmin to store higher
+32 bits of its XMAX.
+
+Double XMAX format is only for full pages that don't have 16 bytes for
+pd_special. So it neither has a�place for a single tuple. Insert and HOT update
+for double XMAX pages is impossible and not supported. We can only read or
+delete tuples from it.
+
+When we are able to prune page double XMAX it will be converted from it to
+general 64-bit XID page format with all operations on its tuples supported.
+
+In-memory tuple format
+----------------------
+
+In-memory tuple representation consists of two parts:
+- HeapTupleHeader from disk page (contains all heap tuple contents, not only
+header)
+- HeapTuple with additional in-memory fields
+
+HeapTuple for each tuple in memory stores t_xid_base/t_multi_base - a copies of
+page's pd_xid_base/pd_multi_base. With tuple's 32-bit t_xmin and t_xmax from
+HeapTupleHeader they are used to calculate actual 64-bit XMIN and XMAX:
+
+XMIN = t_xmin + t_xid_base. 					(3)
+XMAX = t_xmax + t_xid_base/t_multi_base.		(4)
+
+The downside of this is that we can not use tuple's XMIN and XMAX right away.
+We often need to re-read t_xmin and t_xmax - which could actually be pointers
+into a page in shared buffers and therefore they could be updated by any other
+backend.
+
+Update/delete with 64-bit XIDs and 32-bit t_xmin/t_xmax
+--------------------------------------------------------------
+
+When we try to delete/update a tuple, we check that XMAX for a page fits (2).
+I.e. that t_xmax will not be over MaxShortTransactionId relative to
+pd_xid_base/pd_multi_base of a its page.
+
+If the current XID doesn't fit a range
+(pd_xid_base, pd_xid_base + MaxShortTransactionId) (5):
+
+- heap_page_prepare_for_xid() will try to increase pd_xid_base/pd_multi_base on
+a page and update all t_xmin/t_xmax of the other tuples on the page to
+correspond new pd_xid_base/pd_multi_base.
+
+- If it was impossible, it will try to prune and freeze tuples on a page.
+
+- If this is unsuccessful it will throw an error. Normally this is very
+unlikely but if there is a very old living transaction with an age of around
+2^32 this can arise. Basically, this is a behavior similar to one during the
+vacuum to prevent wraparound when XID was 32-bit. Dba should take care and
+avoid very-long-living transactions with an age close to 2^32. So long-living
+transactions often they are most likely defunct.
+
+Insert with 64-bit XIDs and 32-bit t_xmin/t_xmax
+------------------------------------------------
+
+On insert we check if current XID fits a range (5). Otherwise:
+
+- heap_page_prepare_for_xid() will try to increase pd_xid_base for t_xmin will
+not be over MaxShortTransactionId.
+
+- If it is impossible, then it will try to prune and freeze tuples on a page.
+
+Known issue: if pd_xid_base could not be shifted to accommodate a tuple being
+inserted due to a very long-running transaction, we just throw an error. We
+neither try to insert a�tuple into another page nor mark the current page as
+full. So, in this (unlikely) case we will get regular insert errors on the next
+tries to insert to the page 'locked' by this very long-running transaction.
+
+Upgrade from 32-bit XID versions
+--------------------------------
+
+pg_upgrade doesn't change pages format itself. It is done lazily after.
+
+1. At first heap page read, tuples on a page are repacked to free 16 bytes
+at the end of a page, possibly freeing space from dead tuples.
+
+2A. 16 bytes of pd_special is added if there is a place for it
+
+2B. Page is converted to "Double XMAX" format if there is no place for
+pd_special
+
+3. If a page is in double XMAX format after its first read, and vacuum (or
+micro-vacuum at select query) could prune some tuples and free space for
+pd_special, prune_page will add pd_special and convert page from double XMAX
+to general 64-bit XID page format.
-- 
2.24.3 (Apple Git-128)


--cpok4wp6gsarlzvp--





^ permalink  raw  reply  [nested|flat] 265+ messages in thread

* [PATCH v6] README for 64bit xid
@ 2022-01-10 19:20 Pavel Borisov <[email protected]>
  0 siblings, 0 replies; 265+ messages in thread

From: Pavel Borisov @ 2022-01-10 19:20 UTC (permalink / raw)

Authors:
- Pavel Borisov <[email protected]>
- Maxim Orlov <[email protected]>
- Yura Sokolov <[email protected]> <[email protected]>
---
 src/backend/access/heap/README.XID64 | 128 +++++++++++++++++++++++++++
 1 file changed, 128 insertions(+)
 create mode 100644 src/backend/access/heap/README.XID64

diff --git a/src/backend/access/heap/README.XID64 b/src/backend/access/heap/README.XID64
new file mode 100644
index 00000000000..457ba9b9ef5
--- /dev/null
+++ b/src/backend/access/heap/README.XID64
@@ -0,0 +1,128 @@
+src/backend/access/heap/README.XID64
+
+64-bit Transaction ID's (XID)
+=============================
+
+A limited number (N = 2^32) of XID's required to do vacuum freeze to prevent
+wraparound every N/2 transactions. This causes performance degradation due
+to the need to exclusively lock tables while being vacuumed. In each
+wraparound cycle, SLRU buffers are also being cut.
+
+With 64-bit XID's wraparound is effectively postponed to a very distant
+future. Even in highly loaded systems that had 2^32 transactions per day
+it will take huge 2^31 days before the first enforced "vacuum to prevent
+wraparound"). Buffers cutting and routine vacuum are not enforced, and DBA
+can plan them independently at the time with the least system load and least
+critical for database performance. Also, it can be done less frequently
+(several times a year vs every several days) on systems with transaction rates
+similar to those mentioned above.
+
+On-disk tuple and page format
+-----------------------------
+
+On-disk tuple format remains unchanged. 32-bit t_xmin and t_xmax store the
+lower parts of 64-bit XMIN and XMAX values. Each heap page has additional
+64-bit pd_xid_base and pd_multi_base which are common for all tuples on a page.
+They are placed into a pd_special area - 16 bytes in the end of a heap page.
+Actual XMIN/XMAX for a tuple are calculated upon reading a tuple from a page
+as follows:
+
+XMIN = t_xmin + pd_xid_base. 					(1)
+XMAX = t_xmax + pd_xid_base/pd_multi_base.		(2)
+
+"Double XMAX" page format
+---------------------------------
+
+At first read of a heap page after pg_upgrade from 32-bit XID PostgreSQL
+version pd_special area with a size of 16 bytes should be added to a page.
+Though a page may not have space for this. Then it can be converted to a
+temporary format called "double XMAX".
+
+All tuples after pg-upgrade would necessarily have xmin = FrozenTransactionId.
+So we don't need tuple header t_xmin field and we reuse t_xmin to store higher
+32 bits of its XMAX.
+
+Double XMAX format is only for full pages that don't have 16 bytes for
+pd_special. So it neither has a�place for a single tuple. Insert and HOT update
+for double XMAX pages is impossible and not supported. We can only read or
+delete tuples from it.
+
+When we are able to prune page double XMAX it will be converted from it to
+general 64-bit XID page format with all operations on its tuples supported.
+
+In-memory tuple format
+----------------------
+
+In-memory tuple representation consists of two parts:
+- HeapTupleHeader from disk page (contains all heap tuple contents, not only
+header)
+- HeapTuple with additional in-memory fields
+
+HeapTuple for each tuple in memory stores t_xid_base/t_multi_base - a copies of
+page's pd_xid_base/pd_multi_base. With tuple's 32-bit t_xmin and t_xmax from
+HeapTupleHeader they are used to calculate actual 64-bit XMIN and XMAX:
+
+XMIN = t_xmin + t_xid_base. 					(3)
+XMAX = t_xmax + t_xid_base/t_multi_base.		(4)
+
+The downside of this is that we can not use tuple's XMIN and XMAX right away.
+We often need to re-read t_xmin and t_xmax - which could actually be pointers
+into a page in shared buffers and therefore they could be updated by any other
+backend.
+
+Update/delete with 64-bit XIDs and 32-bit t_xmin/t_xmax
+--------------------------------------------------------------
+
+When we try to delete/update a tuple, we check that XMAX for a page fits (2).
+I.e. that t_xmax will not be over MaxShortTransactionId relative to
+pd_xid_base/pd_multi_base of a its page.
+
+If the current XID doesn't fit a range
+(pd_xid_base, pd_xid_base + MaxShortTransactionId) (5):
+
+- heap_page_prepare_for_xid() will try to increase pd_xid_base/pd_multi_base on
+a page and update all t_xmin/t_xmax of the other tuples on the page to
+correspond new pd_xid_base/pd_multi_base.
+
+- If it was impossible, it will try to prune and freeze tuples on a page.
+
+- If this is unsuccessful it will throw an error. Normally this is very
+unlikely but if there is a very old living transaction with an age of around
+2^32 this can arise. Basically, this is a behavior similar to one during the
+vacuum to prevent wraparound when XID was 32-bit. Dba should take care and
+avoid very-long-living transactions with an age close to 2^32. So long-living
+transactions often they are most likely defunct.
+
+Insert with 64-bit XIDs and 32-bit t_xmin/t_xmax
+------------------------------------------------
+
+On insert we check if current XID fits a range (5). Otherwise:
+
+- heap_page_prepare_for_xid() will try to increase pd_xid_base for t_xmin will
+not be over MaxShortTransactionId.
+
+- If it is impossible, then it will try to prune and freeze tuples on a page.
+
+Known issue: if pd_xid_base could not be shifted to accommodate a tuple being
+inserted due to a very long-running transaction, we just throw an error. We
+neither try to insert a�tuple into another page nor mark the current page as
+full. So, in this (unlikely) case we will get regular insert errors on the next
+tries to insert to the page 'locked' by this very long-running transaction.
+
+Upgrade from 32-bit XID versions
+--------------------------------
+
+pg_upgrade doesn't change pages format itself. It is done lazily after.
+
+1. At first heap page read, tuples on a page are repacked to free 16 bytes
+at the end of a page, possibly freeing space from dead tuples.
+
+2A. 16 bytes of pd_special is added if there is a place for it
+
+2B. Page is converted to "Double XMAX" format if there is no place for
+pd_special
+
+3. If a page is in double XMAX format after its first read, and vacuum (or
+micro-vacuum at select query) could prune some tuples and free space for
+pd_special, prune_page will add pd_special and convert page from double XMAX
+to general 64-bit XID page format.
-- 
2.24.3 (Apple Git-128)


--cpok4wp6gsarlzvp--





^ permalink  raw  reply  [nested|flat] 265+ messages in thread

* [PATCH v6] README for 64bit xid
@ 2022-01-10 19:20 Pavel Borisov <[email protected]>
  0 siblings, 0 replies; 265+ messages in thread

From: Pavel Borisov @ 2022-01-10 19:20 UTC (permalink / raw)

Authors:
- Pavel Borisov <[email protected]>
- Maxim Orlov <[email protected]>
- Yura Sokolov <[email protected]> <[email protected]>
---
 src/backend/access/heap/README.XID64 | 128 +++++++++++++++++++++++++++
 1 file changed, 128 insertions(+)
 create mode 100644 src/backend/access/heap/README.XID64

diff --git a/src/backend/access/heap/README.XID64 b/src/backend/access/heap/README.XID64
new file mode 100644
index 00000000000..457ba9b9ef5
--- /dev/null
+++ b/src/backend/access/heap/README.XID64
@@ -0,0 +1,128 @@
+src/backend/access/heap/README.XID64
+
+64-bit Transaction ID's (XID)
+=============================
+
+A limited number (N = 2^32) of XID's required to do vacuum freeze to prevent
+wraparound every N/2 transactions. This causes performance degradation due
+to the need to exclusively lock tables while being vacuumed. In each
+wraparound cycle, SLRU buffers are also being cut.
+
+With 64-bit XID's wraparound is effectively postponed to a very distant
+future. Even in highly loaded systems that had 2^32 transactions per day
+it will take huge 2^31 days before the first enforced "vacuum to prevent
+wraparound"). Buffers cutting and routine vacuum are not enforced, and DBA
+can plan them independently at the time with the least system load and least
+critical for database performance. Also, it can be done less frequently
+(several times a year vs every several days) on systems with transaction rates
+similar to those mentioned above.
+
+On-disk tuple and page format
+-----------------------------
+
+On-disk tuple format remains unchanged. 32-bit t_xmin and t_xmax store the
+lower parts of 64-bit XMIN and XMAX values. Each heap page has additional
+64-bit pd_xid_base and pd_multi_base which are common for all tuples on a page.
+They are placed into a pd_special area - 16 bytes in the end of a heap page.
+Actual XMIN/XMAX for a tuple are calculated upon reading a tuple from a page
+as follows:
+
+XMIN = t_xmin + pd_xid_base. 					(1)
+XMAX = t_xmax + pd_xid_base/pd_multi_base.		(2)
+
+"Double XMAX" page format
+---------------------------------
+
+At first read of a heap page after pg_upgrade from 32-bit XID PostgreSQL
+version pd_special area with a size of 16 bytes should be added to a page.
+Though a page may not have space for this. Then it can be converted to a
+temporary format called "double XMAX".
+
+All tuples after pg-upgrade would necessarily have xmin = FrozenTransactionId.
+So we don't need tuple header t_xmin field and we reuse t_xmin to store higher
+32 bits of its XMAX.
+
+Double XMAX format is only for full pages that don't have 16 bytes for
+pd_special. So it neither has a�place for a single tuple. Insert and HOT update
+for double XMAX pages is impossible and not supported. We can only read or
+delete tuples from it.
+
+When we are able to prune page double XMAX it will be converted from it to
+general 64-bit XID page format with all operations on its tuples supported.
+
+In-memory tuple format
+----------------------
+
+In-memory tuple representation consists of two parts:
+- HeapTupleHeader from disk page (contains all heap tuple contents, not only
+header)
+- HeapTuple with additional in-memory fields
+
+HeapTuple for each tuple in memory stores t_xid_base/t_multi_base - a copies of
+page's pd_xid_base/pd_multi_base. With tuple's 32-bit t_xmin and t_xmax from
+HeapTupleHeader they are used to calculate actual 64-bit XMIN and XMAX:
+
+XMIN = t_xmin + t_xid_base. 					(3)
+XMAX = t_xmax + t_xid_base/t_multi_base.		(4)
+
+The downside of this is that we can not use tuple's XMIN and XMAX right away.
+We often need to re-read t_xmin and t_xmax - which could actually be pointers
+into a page in shared buffers and therefore they could be updated by any other
+backend.
+
+Update/delete with 64-bit XIDs and 32-bit t_xmin/t_xmax
+--------------------------------------------------------------
+
+When we try to delete/update a tuple, we check that XMAX for a page fits (2).
+I.e. that t_xmax will not be over MaxShortTransactionId relative to
+pd_xid_base/pd_multi_base of a its page.
+
+If the current XID doesn't fit a range
+(pd_xid_base, pd_xid_base + MaxShortTransactionId) (5):
+
+- heap_page_prepare_for_xid() will try to increase pd_xid_base/pd_multi_base on
+a page and update all t_xmin/t_xmax of the other tuples on the page to
+correspond new pd_xid_base/pd_multi_base.
+
+- If it was impossible, it will try to prune and freeze tuples on a page.
+
+- If this is unsuccessful it will throw an error. Normally this is very
+unlikely but if there is a very old living transaction with an age of around
+2^32 this can arise. Basically, this is a behavior similar to one during the
+vacuum to prevent wraparound when XID was 32-bit. Dba should take care and
+avoid very-long-living transactions with an age close to 2^32. So long-living
+transactions often they are most likely defunct.
+
+Insert with 64-bit XIDs and 32-bit t_xmin/t_xmax
+------------------------------------------------
+
+On insert we check if current XID fits a range (5). Otherwise:
+
+- heap_page_prepare_for_xid() will try to increase pd_xid_base for t_xmin will
+not be over MaxShortTransactionId.
+
+- If it is impossible, then it will try to prune and freeze tuples on a page.
+
+Known issue: if pd_xid_base could not be shifted to accommodate a tuple being
+inserted due to a very long-running transaction, we just throw an error. We
+neither try to insert a�tuple into another page nor mark the current page as
+full. So, in this (unlikely) case we will get regular insert errors on the next
+tries to insert to the page 'locked' by this very long-running transaction.
+
+Upgrade from 32-bit XID versions
+--------------------------------
+
+pg_upgrade doesn't change pages format itself. It is done lazily after.
+
+1. At first heap page read, tuples on a page are repacked to free 16 bytes
+at the end of a page, possibly freeing space from dead tuples.
+
+2A. 16 bytes of pd_special is added if there is a place for it
+
+2B. Page is converted to "Double XMAX" format if there is no place for
+pd_special
+
+3. If a page is in double XMAX format after its first read, and vacuum (or
+micro-vacuum at select query) could prune some tuples and free space for
+pd_special, prune_page will add pd_special and convert page from double XMAX
+to general 64-bit XID page format.
-- 
2.24.3 (Apple Git-128)


--cpok4wp6gsarlzvp--





^ permalink  raw  reply  [nested|flat] 265+ messages in thread

* [PATCH v6] README for 64bit xid
@ 2022-01-10 19:20 Pavel Borisov <[email protected]>
  0 siblings, 0 replies; 265+ messages in thread

From: Pavel Borisov @ 2022-01-10 19:20 UTC (permalink / raw)

Authors:
- Pavel Borisov <[email protected]>
- Maxim Orlov <[email protected]>
- Yura Sokolov <[email protected]> <[email protected]>
---
 src/backend/access/heap/README.XID64 | 128 +++++++++++++++++++++++++++
 1 file changed, 128 insertions(+)
 create mode 100644 src/backend/access/heap/README.XID64

diff --git a/src/backend/access/heap/README.XID64 b/src/backend/access/heap/README.XID64
new file mode 100644
index 00000000000..457ba9b9ef5
--- /dev/null
+++ b/src/backend/access/heap/README.XID64
@@ -0,0 +1,128 @@
+src/backend/access/heap/README.XID64
+
+64-bit Transaction ID's (XID)
+=============================
+
+A limited number (N = 2^32) of XID's required to do vacuum freeze to prevent
+wraparound every N/2 transactions. This causes performance degradation due
+to the need to exclusively lock tables while being vacuumed. In each
+wraparound cycle, SLRU buffers are also being cut.
+
+With 64-bit XID's wraparound is effectively postponed to a very distant
+future. Even in highly loaded systems that had 2^32 transactions per day
+it will take huge 2^31 days before the first enforced "vacuum to prevent
+wraparound"). Buffers cutting and routine vacuum are not enforced, and DBA
+can plan them independently at the time with the least system load and least
+critical for database performance. Also, it can be done less frequently
+(several times a year vs every several days) on systems with transaction rates
+similar to those mentioned above.
+
+On-disk tuple and page format
+-----------------------------
+
+On-disk tuple format remains unchanged. 32-bit t_xmin and t_xmax store the
+lower parts of 64-bit XMIN and XMAX values. Each heap page has additional
+64-bit pd_xid_base and pd_multi_base which are common for all tuples on a page.
+They are placed into a pd_special area - 16 bytes in the end of a heap page.
+Actual XMIN/XMAX for a tuple are calculated upon reading a tuple from a page
+as follows:
+
+XMIN = t_xmin + pd_xid_base. 					(1)
+XMAX = t_xmax + pd_xid_base/pd_multi_base.		(2)
+
+"Double XMAX" page format
+---------------------------------
+
+At first read of a heap page after pg_upgrade from 32-bit XID PostgreSQL
+version pd_special area with a size of 16 bytes should be added to a page.
+Though a page may not have space for this. Then it can be converted to a
+temporary format called "double XMAX".
+
+All tuples after pg-upgrade would necessarily have xmin = FrozenTransactionId.
+So we don't need tuple header t_xmin field and we reuse t_xmin to store higher
+32 bits of its XMAX.
+
+Double XMAX format is only for full pages that don't have 16 bytes for
+pd_special. So it neither has a�place for a single tuple. Insert and HOT update
+for double XMAX pages is impossible and not supported. We can only read or
+delete tuples from it.
+
+When we are able to prune page double XMAX it will be converted from it to
+general 64-bit XID page format with all operations on its tuples supported.
+
+In-memory tuple format
+----------------------
+
+In-memory tuple representation consists of two parts:
+- HeapTupleHeader from disk page (contains all heap tuple contents, not only
+header)
+- HeapTuple with additional in-memory fields
+
+HeapTuple for each tuple in memory stores t_xid_base/t_multi_base - a copies of
+page's pd_xid_base/pd_multi_base. With tuple's 32-bit t_xmin and t_xmax from
+HeapTupleHeader they are used to calculate actual 64-bit XMIN and XMAX:
+
+XMIN = t_xmin + t_xid_base. 					(3)
+XMAX = t_xmax + t_xid_base/t_multi_base.		(4)
+
+The downside of this is that we can not use tuple's XMIN and XMAX right away.
+We often need to re-read t_xmin and t_xmax - which could actually be pointers
+into a page in shared buffers and therefore they could be updated by any other
+backend.
+
+Update/delete with 64-bit XIDs and 32-bit t_xmin/t_xmax
+--------------------------------------------------------------
+
+When we try to delete/update a tuple, we check that XMAX for a page fits (2).
+I.e. that t_xmax will not be over MaxShortTransactionId relative to
+pd_xid_base/pd_multi_base of a its page.
+
+If the current XID doesn't fit a range
+(pd_xid_base, pd_xid_base + MaxShortTransactionId) (5):
+
+- heap_page_prepare_for_xid() will try to increase pd_xid_base/pd_multi_base on
+a page and update all t_xmin/t_xmax of the other tuples on the page to
+correspond new pd_xid_base/pd_multi_base.
+
+- If it was impossible, it will try to prune and freeze tuples on a page.
+
+- If this is unsuccessful it will throw an error. Normally this is very
+unlikely but if there is a very old living transaction with an age of around
+2^32 this can arise. Basically, this is a behavior similar to one during the
+vacuum to prevent wraparound when XID was 32-bit. Dba should take care and
+avoid very-long-living transactions with an age close to 2^32. So long-living
+transactions often they are most likely defunct.
+
+Insert with 64-bit XIDs and 32-bit t_xmin/t_xmax
+------------------------------------------------
+
+On insert we check if current XID fits a range (5). Otherwise:
+
+- heap_page_prepare_for_xid() will try to increase pd_xid_base for t_xmin will
+not be over MaxShortTransactionId.
+
+- If it is impossible, then it will try to prune and freeze tuples on a page.
+
+Known issue: if pd_xid_base could not be shifted to accommodate a tuple being
+inserted due to a very long-running transaction, we just throw an error. We
+neither try to insert a�tuple into another page nor mark the current page as
+full. So, in this (unlikely) case we will get regular insert errors on the next
+tries to insert to the page 'locked' by this very long-running transaction.
+
+Upgrade from 32-bit XID versions
+--------------------------------
+
+pg_upgrade doesn't change pages format itself. It is done lazily after.
+
+1. At first heap page read, tuples on a page are repacked to free 16 bytes
+at the end of a page, possibly freeing space from dead tuples.
+
+2A. 16 bytes of pd_special is added if there is a place for it
+
+2B. Page is converted to "Double XMAX" format if there is no place for
+pd_special
+
+3. If a page is in double XMAX format after its first read, and vacuum (or
+micro-vacuum at select query) could prune some tuples and free space for
+pd_special, prune_page will add pd_special and convert page from double XMAX
+to general 64-bit XID page format.
-- 
2.24.3 (Apple Git-128)


--cpok4wp6gsarlzvp--





^ permalink  raw  reply  [nested|flat] 265+ messages in thread

* [PATCH v6] README for 64bit xid
@ 2022-01-10 19:20 Pavel Borisov <[email protected]>
  0 siblings, 0 replies; 265+ messages in thread

From: Pavel Borisov @ 2022-01-10 19:20 UTC (permalink / raw)

Authors:
- Pavel Borisov <[email protected]>
- Maxim Orlov <[email protected]>
- Yura Sokolov <[email protected]> <[email protected]>
---
 src/backend/access/heap/README.XID64 | 128 +++++++++++++++++++++++++++
 1 file changed, 128 insertions(+)
 create mode 100644 src/backend/access/heap/README.XID64

diff --git a/src/backend/access/heap/README.XID64 b/src/backend/access/heap/README.XID64
new file mode 100644
index 00000000000..457ba9b9ef5
--- /dev/null
+++ b/src/backend/access/heap/README.XID64
@@ -0,0 +1,128 @@
+src/backend/access/heap/README.XID64
+
+64-bit Transaction ID's (XID)
+=============================
+
+A limited number (N = 2^32) of XID's required to do vacuum freeze to prevent
+wraparound every N/2 transactions. This causes performance degradation due
+to the need to exclusively lock tables while being vacuumed. In each
+wraparound cycle, SLRU buffers are also being cut.
+
+With 64-bit XID's wraparound is effectively postponed to a very distant
+future. Even in highly loaded systems that had 2^32 transactions per day
+it will take huge 2^31 days before the first enforced "vacuum to prevent
+wraparound"). Buffers cutting and routine vacuum are not enforced, and DBA
+can plan them independently at the time with the least system load and least
+critical for database performance. Also, it can be done less frequently
+(several times a year vs every several days) on systems with transaction rates
+similar to those mentioned above.
+
+On-disk tuple and page format
+-----------------------------
+
+On-disk tuple format remains unchanged. 32-bit t_xmin and t_xmax store the
+lower parts of 64-bit XMIN and XMAX values. Each heap page has additional
+64-bit pd_xid_base and pd_multi_base which are common for all tuples on a page.
+They are placed into a pd_special area - 16 bytes in the end of a heap page.
+Actual XMIN/XMAX for a tuple are calculated upon reading a tuple from a page
+as follows:
+
+XMIN = t_xmin + pd_xid_base. 					(1)
+XMAX = t_xmax + pd_xid_base/pd_multi_base.		(2)
+
+"Double XMAX" page format
+---------------------------------
+
+At first read of a heap page after pg_upgrade from 32-bit XID PostgreSQL
+version pd_special area with a size of 16 bytes should be added to a page.
+Though a page may not have space for this. Then it can be converted to a
+temporary format called "double XMAX".
+
+All tuples after pg-upgrade would necessarily have xmin = FrozenTransactionId.
+So we don't need tuple header t_xmin field and we reuse t_xmin to store higher
+32 bits of its XMAX.
+
+Double XMAX format is only for full pages that don't have 16 bytes for
+pd_special. So it neither has a�place for a single tuple. Insert and HOT update
+for double XMAX pages is impossible and not supported. We can only read or
+delete tuples from it.
+
+When we are able to prune page double XMAX it will be converted from it to
+general 64-bit XID page format with all operations on its tuples supported.
+
+In-memory tuple format
+----------------------
+
+In-memory tuple representation consists of two parts:
+- HeapTupleHeader from disk page (contains all heap tuple contents, not only
+header)
+- HeapTuple with additional in-memory fields
+
+HeapTuple for each tuple in memory stores t_xid_base/t_multi_base - a copies of
+page's pd_xid_base/pd_multi_base. With tuple's 32-bit t_xmin and t_xmax from
+HeapTupleHeader they are used to calculate actual 64-bit XMIN and XMAX:
+
+XMIN = t_xmin + t_xid_base. 					(3)
+XMAX = t_xmax + t_xid_base/t_multi_base.		(4)
+
+The downside of this is that we can not use tuple's XMIN and XMAX right away.
+We often need to re-read t_xmin and t_xmax - which could actually be pointers
+into a page in shared buffers and therefore they could be updated by any other
+backend.
+
+Update/delete with 64-bit XIDs and 32-bit t_xmin/t_xmax
+--------------------------------------------------------------
+
+When we try to delete/update a tuple, we check that XMAX for a page fits (2).
+I.e. that t_xmax will not be over MaxShortTransactionId relative to
+pd_xid_base/pd_multi_base of a its page.
+
+If the current XID doesn't fit a range
+(pd_xid_base, pd_xid_base + MaxShortTransactionId) (5):
+
+- heap_page_prepare_for_xid() will try to increase pd_xid_base/pd_multi_base on
+a page and update all t_xmin/t_xmax of the other tuples on the page to
+correspond new pd_xid_base/pd_multi_base.
+
+- If it was impossible, it will try to prune and freeze tuples on a page.
+
+- If this is unsuccessful it will throw an error. Normally this is very
+unlikely but if there is a very old living transaction with an age of around
+2^32 this can arise. Basically, this is a behavior similar to one during the
+vacuum to prevent wraparound when XID was 32-bit. Dba should take care and
+avoid very-long-living transactions with an age close to 2^32. So long-living
+transactions often they are most likely defunct.
+
+Insert with 64-bit XIDs and 32-bit t_xmin/t_xmax
+------------------------------------------------
+
+On insert we check if current XID fits a range (5). Otherwise:
+
+- heap_page_prepare_for_xid() will try to increase pd_xid_base for t_xmin will
+not be over MaxShortTransactionId.
+
+- If it is impossible, then it will try to prune and freeze tuples on a page.
+
+Known issue: if pd_xid_base could not be shifted to accommodate a tuple being
+inserted due to a very long-running transaction, we just throw an error. We
+neither try to insert a�tuple into another page nor mark the current page as
+full. So, in this (unlikely) case we will get regular insert errors on the next
+tries to insert to the page 'locked' by this very long-running transaction.
+
+Upgrade from 32-bit XID versions
+--------------------------------
+
+pg_upgrade doesn't change pages format itself. It is done lazily after.
+
+1. At first heap page read, tuples on a page are repacked to free 16 bytes
+at the end of a page, possibly freeing space from dead tuples.
+
+2A. 16 bytes of pd_special is added if there is a place for it
+
+2B. Page is converted to "Double XMAX" format if there is no place for
+pd_special
+
+3. If a page is in double XMAX format after its first read, and vacuum (or
+micro-vacuum at select query) could prune some tuples and free space for
+pd_special, prune_page will add pd_special and convert page from double XMAX
+to general 64-bit XID page format.
-- 
2.24.3 (Apple Git-128)


--cpok4wp6gsarlzvp--





^ permalink  raw  reply  [nested|flat] 265+ messages in thread

* [PATCH v6] README for 64bit xid
@ 2022-01-10 19:20 Pavel Borisov <[email protected]>
  0 siblings, 0 replies; 265+ messages in thread

From: Pavel Borisov @ 2022-01-10 19:20 UTC (permalink / raw)

Authors:
- Pavel Borisov <[email protected]>
- Maxim Orlov <[email protected]>
- Yura Sokolov <[email protected]> <[email protected]>
---
 src/backend/access/heap/README.XID64 | 128 +++++++++++++++++++++++++++
 1 file changed, 128 insertions(+)
 create mode 100644 src/backend/access/heap/README.XID64

diff --git a/src/backend/access/heap/README.XID64 b/src/backend/access/heap/README.XID64
new file mode 100644
index 00000000000..457ba9b9ef5
--- /dev/null
+++ b/src/backend/access/heap/README.XID64
@@ -0,0 +1,128 @@
+src/backend/access/heap/README.XID64
+
+64-bit Transaction ID's (XID)
+=============================
+
+A limited number (N = 2^32) of XID's required to do vacuum freeze to prevent
+wraparound every N/2 transactions. This causes performance degradation due
+to the need to exclusively lock tables while being vacuumed. In each
+wraparound cycle, SLRU buffers are also being cut.
+
+With 64-bit XID's wraparound is effectively postponed to a very distant
+future. Even in highly loaded systems that had 2^32 transactions per day
+it will take huge 2^31 days before the first enforced "vacuum to prevent
+wraparound"). Buffers cutting and routine vacuum are not enforced, and DBA
+can plan them independently at the time with the least system load and least
+critical for database performance. Also, it can be done less frequently
+(several times a year vs every several days) on systems with transaction rates
+similar to those mentioned above.
+
+On-disk tuple and page format
+-----------------------------
+
+On-disk tuple format remains unchanged. 32-bit t_xmin and t_xmax store the
+lower parts of 64-bit XMIN and XMAX values. Each heap page has additional
+64-bit pd_xid_base and pd_multi_base which are common for all tuples on a page.
+They are placed into a pd_special area - 16 bytes in the end of a heap page.
+Actual XMIN/XMAX for a tuple are calculated upon reading a tuple from a page
+as follows:
+
+XMIN = t_xmin + pd_xid_base. 					(1)
+XMAX = t_xmax + pd_xid_base/pd_multi_base.		(2)
+
+"Double XMAX" page format
+---------------------------------
+
+At first read of a heap page after pg_upgrade from 32-bit XID PostgreSQL
+version pd_special area with a size of 16 bytes should be added to a page.
+Though a page may not have space for this. Then it can be converted to a
+temporary format called "double XMAX".
+
+All tuples after pg-upgrade would necessarily have xmin = FrozenTransactionId.
+So we don't need tuple header t_xmin field and we reuse t_xmin to store higher
+32 bits of its XMAX.
+
+Double XMAX format is only for full pages that don't have 16 bytes for
+pd_special. So it neither has a�place for a single tuple. Insert and HOT update
+for double XMAX pages is impossible and not supported. We can only read or
+delete tuples from it.
+
+When we are able to prune page double XMAX it will be converted from it to
+general 64-bit XID page format with all operations on its tuples supported.
+
+In-memory tuple format
+----------------------
+
+In-memory tuple representation consists of two parts:
+- HeapTupleHeader from disk page (contains all heap tuple contents, not only
+header)
+- HeapTuple with additional in-memory fields
+
+HeapTuple for each tuple in memory stores t_xid_base/t_multi_base - a copies of
+page's pd_xid_base/pd_multi_base. With tuple's 32-bit t_xmin and t_xmax from
+HeapTupleHeader they are used to calculate actual 64-bit XMIN and XMAX:
+
+XMIN = t_xmin + t_xid_base. 					(3)
+XMAX = t_xmax + t_xid_base/t_multi_base.		(4)
+
+The downside of this is that we can not use tuple's XMIN and XMAX right away.
+We often need to re-read t_xmin and t_xmax - which could actually be pointers
+into a page in shared buffers and therefore they could be updated by any other
+backend.
+
+Update/delete with 64-bit XIDs and 32-bit t_xmin/t_xmax
+--------------------------------------------------------------
+
+When we try to delete/update a tuple, we check that XMAX for a page fits (2).
+I.e. that t_xmax will not be over MaxShortTransactionId relative to
+pd_xid_base/pd_multi_base of a its page.
+
+If the current XID doesn't fit a range
+(pd_xid_base, pd_xid_base + MaxShortTransactionId) (5):
+
+- heap_page_prepare_for_xid() will try to increase pd_xid_base/pd_multi_base on
+a page and update all t_xmin/t_xmax of the other tuples on the page to
+correspond new pd_xid_base/pd_multi_base.
+
+- If it was impossible, it will try to prune and freeze tuples on a page.
+
+- If this is unsuccessful it will throw an error. Normally this is very
+unlikely but if there is a very old living transaction with an age of around
+2^32 this can arise. Basically, this is a behavior similar to one during the
+vacuum to prevent wraparound when XID was 32-bit. Dba should take care and
+avoid very-long-living transactions with an age close to 2^32. So long-living
+transactions often they are most likely defunct.
+
+Insert with 64-bit XIDs and 32-bit t_xmin/t_xmax
+------------------------------------------------
+
+On insert we check if current XID fits a range (5). Otherwise:
+
+- heap_page_prepare_for_xid() will try to increase pd_xid_base for t_xmin will
+not be over MaxShortTransactionId.
+
+- If it is impossible, then it will try to prune and freeze tuples on a page.
+
+Known issue: if pd_xid_base could not be shifted to accommodate a tuple being
+inserted due to a very long-running transaction, we just throw an error. We
+neither try to insert a�tuple into another page nor mark the current page as
+full. So, in this (unlikely) case we will get regular insert errors on the next
+tries to insert to the page 'locked' by this very long-running transaction.
+
+Upgrade from 32-bit XID versions
+--------------------------------
+
+pg_upgrade doesn't change pages format itself. It is done lazily after.
+
+1. At first heap page read, tuples on a page are repacked to free 16 bytes
+at the end of a page, possibly freeing space from dead tuples.
+
+2A. 16 bytes of pd_special is added if there is a place for it
+
+2B. Page is converted to "Double XMAX" format if there is no place for
+pd_special
+
+3. If a page is in double XMAX format after its first read, and vacuum (or
+micro-vacuum at select query) could prune some tuples and free space for
+pd_special, prune_page will add pd_special and convert page from double XMAX
+to general 64-bit XID page format.
-- 
2.24.3 (Apple Git-128)


--cpok4wp6gsarlzvp--





^ permalink  raw  reply  [nested|flat] 265+ messages in thread

* [PATCH v6] README for 64bit xid
@ 2022-01-10 19:20 Pavel Borisov <[email protected]>
  0 siblings, 0 replies; 265+ messages in thread

From: Pavel Borisov @ 2022-01-10 19:20 UTC (permalink / raw)

Authors:
- Pavel Borisov <[email protected]>
- Maxim Orlov <[email protected]>
- Yura Sokolov <[email protected]> <[email protected]>
---
 src/backend/access/heap/README.XID64 | 128 +++++++++++++++++++++++++++
 1 file changed, 128 insertions(+)
 create mode 100644 src/backend/access/heap/README.XID64

diff --git a/src/backend/access/heap/README.XID64 b/src/backend/access/heap/README.XID64
new file mode 100644
index 00000000000..457ba9b9ef5
--- /dev/null
+++ b/src/backend/access/heap/README.XID64
@@ -0,0 +1,128 @@
+src/backend/access/heap/README.XID64
+
+64-bit Transaction ID's (XID)
+=============================
+
+A limited number (N = 2^32) of XID's required to do vacuum freeze to prevent
+wraparound every N/2 transactions. This causes performance degradation due
+to the need to exclusively lock tables while being vacuumed. In each
+wraparound cycle, SLRU buffers are also being cut.
+
+With 64-bit XID's wraparound is effectively postponed to a very distant
+future. Even in highly loaded systems that had 2^32 transactions per day
+it will take huge 2^31 days before the first enforced "vacuum to prevent
+wraparound"). Buffers cutting and routine vacuum are not enforced, and DBA
+can plan them independently at the time with the least system load and least
+critical for database performance. Also, it can be done less frequently
+(several times a year vs every several days) on systems with transaction rates
+similar to those mentioned above.
+
+On-disk tuple and page format
+-----------------------------
+
+On-disk tuple format remains unchanged. 32-bit t_xmin and t_xmax store the
+lower parts of 64-bit XMIN and XMAX values. Each heap page has additional
+64-bit pd_xid_base and pd_multi_base which are common for all tuples on a page.
+They are placed into a pd_special area - 16 bytes in the end of a heap page.
+Actual XMIN/XMAX for a tuple are calculated upon reading a tuple from a page
+as follows:
+
+XMIN = t_xmin + pd_xid_base. 					(1)
+XMAX = t_xmax + pd_xid_base/pd_multi_base.		(2)
+
+"Double XMAX" page format
+---------------------------------
+
+At first read of a heap page after pg_upgrade from 32-bit XID PostgreSQL
+version pd_special area with a size of 16 bytes should be added to a page.
+Though a page may not have space for this. Then it can be converted to a
+temporary format called "double XMAX".
+
+All tuples after pg-upgrade would necessarily have xmin = FrozenTransactionId.
+So we don't need tuple header t_xmin field and we reuse t_xmin to store higher
+32 bits of its XMAX.
+
+Double XMAX format is only for full pages that don't have 16 bytes for
+pd_special. So it neither has a�place for a single tuple. Insert and HOT update
+for double XMAX pages is impossible and not supported. We can only read or
+delete tuples from it.
+
+When we are able to prune page double XMAX it will be converted from it to
+general 64-bit XID page format with all operations on its tuples supported.
+
+In-memory tuple format
+----------------------
+
+In-memory tuple representation consists of two parts:
+- HeapTupleHeader from disk page (contains all heap tuple contents, not only
+header)
+- HeapTuple with additional in-memory fields
+
+HeapTuple for each tuple in memory stores t_xid_base/t_multi_base - a copies of
+page's pd_xid_base/pd_multi_base. With tuple's 32-bit t_xmin and t_xmax from
+HeapTupleHeader they are used to calculate actual 64-bit XMIN and XMAX:
+
+XMIN = t_xmin + t_xid_base. 					(3)
+XMAX = t_xmax + t_xid_base/t_multi_base.		(4)
+
+The downside of this is that we can not use tuple's XMIN and XMAX right away.
+We often need to re-read t_xmin and t_xmax - which could actually be pointers
+into a page in shared buffers and therefore they could be updated by any other
+backend.
+
+Update/delete with 64-bit XIDs and 32-bit t_xmin/t_xmax
+--------------------------------------------------------------
+
+When we try to delete/update a tuple, we check that XMAX for a page fits (2).
+I.e. that t_xmax will not be over MaxShortTransactionId relative to
+pd_xid_base/pd_multi_base of a its page.
+
+If the current XID doesn't fit a range
+(pd_xid_base, pd_xid_base + MaxShortTransactionId) (5):
+
+- heap_page_prepare_for_xid() will try to increase pd_xid_base/pd_multi_base on
+a page and update all t_xmin/t_xmax of the other tuples on the page to
+correspond new pd_xid_base/pd_multi_base.
+
+- If it was impossible, it will try to prune and freeze tuples on a page.
+
+- If this is unsuccessful it will throw an error. Normally this is very
+unlikely but if there is a very old living transaction with an age of around
+2^32 this can arise. Basically, this is a behavior similar to one during the
+vacuum to prevent wraparound when XID was 32-bit. Dba should take care and
+avoid very-long-living transactions with an age close to 2^32. So long-living
+transactions often they are most likely defunct.
+
+Insert with 64-bit XIDs and 32-bit t_xmin/t_xmax
+------------------------------------------------
+
+On insert we check if current XID fits a range (5). Otherwise:
+
+- heap_page_prepare_for_xid() will try to increase pd_xid_base for t_xmin will
+not be over MaxShortTransactionId.
+
+- If it is impossible, then it will try to prune and freeze tuples on a page.
+
+Known issue: if pd_xid_base could not be shifted to accommodate a tuple being
+inserted due to a very long-running transaction, we just throw an error. We
+neither try to insert a�tuple into another page nor mark the current page as
+full. So, in this (unlikely) case we will get regular insert errors on the next
+tries to insert to the page 'locked' by this very long-running transaction.
+
+Upgrade from 32-bit XID versions
+--------------------------------
+
+pg_upgrade doesn't change pages format itself. It is done lazily after.
+
+1. At first heap page read, tuples on a page are repacked to free 16 bytes
+at the end of a page, possibly freeing space from dead tuples.
+
+2A. 16 bytes of pd_special is added if there is a place for it
+
+2B. Page is converted to "Double XMAX" format if there is no place for
+pd_special
+
+3. If a page is in double XMAX format after its first read, and vacuum (or
+micro-vacuum at select query) could prune some tuples and free space for
+pd_special, prune_page will add pd_special and convert page from double XMAX
+to general 64-bit XID page format.
-- 
2.24.3 (Apple Git-128)


--cpok4wp6gsarlzvp--





^ permalink  raw  reply  [nested|flat] 265+ messages in thread

* [PATCH v6] README for 64bit xid
@ 2022-01-10 19:20 Pavel Borisov <[email protected]>
  0 siblings, 0 replies; 265+ messages in thread

From: Pavel Borisov @ 2022-01-10 19:20 UTC (permalink / raw)

Authors:
- Pavel Borisov <[email protected]>
- Maxim Orlov <[email protected]>
- Yura Sokolov <[email protected]> <[email protected]>
---
 src/backend/access/heap/README.XID64 | 128 +++++++++++++++++++++++++++
 1 file changed, 128 insertions(+)
 create mode 100644 src/backend/access/heap/README.XID64

diff --git a/src/backend/access/heap/README.XID64 b/src/backend/access/heap/README.XID64
new file mode 100644
index 00000000000..457ba9b9ef5
--- /dev/null
+++ b/src/backend/access/heap/README.XID64
@@ -0,0 +1,128 @@
+src/backend/access/heap/README.XID64
+
+64-bit Transaction ID's (XID)
+=============================
+
+A limited number (N = 2^32) of XID's required to do vacuum freeze to prevent
+wraparound every N/2 transactions. This causes performance degradation due
+to the need to exclusively lock tables while being vacuumed. In each
+wraparound cycle, SLRU buffers are also being cut.
+
+With 64-bit XID's wraparound is effectively postponed to a very distant
+future. Even in highly loaded systems that had 2^32 transactions per day
+it will take huge 2^31 days before the first enforced "vacuum to prevent
+wraparound"). Buffers cutting and routine vacuum are not enforced, and DBA
+can plan them independently at the time with the least system load and least
+critical for database performance. Also, it can be done less frequently
+(several times a year vs every several days) on systems with transaction rates
+similar to those mentioned above.
+
+On-disk tuple and page format
+-----------------------------
+
+On-disk tuple format remains unchanged. 32-bit t_xmin and t_xmax store the
+lower parts of 64-bit XMIN and XMAX values. Each heap page has additional
+64-bit pd_xid_base and pd_multi_base which are common for all tuples on a page.
+They are placed into a pd_special area - 16 bytes in the end of a heap page.
+Actual XMIN/XMAX for a tuple are calculated upon reading a tuple from a page
+as follows:
+
+XMIN = t_xmin + pd_xid_base. 					(1)
+XMAX = t_xmax + pd_xid_base/pd_multi_base.		(2)
+
+"Double XMAX" page format
+---------------------------------
+
+At first read of a heap page after pg_upgrade from 32-bit XID PostgreSQL
+version pd_special area with a size of 16 bytes should be added to a page.
+Though a page may not have space for this. Then it can be converted to a
+temporary format called "double XMAX".
+
+All tuples after pg-upgrade would necessarily have xmin = FrozenTransactionId.
+So we don't need tuple header t_xmin field and we reuse t_xmin to store higher
+32 bits of its XMAX.
+
+Double XMAX format is only for full pages that don't have 16 bytes for
+pd_special. So it neither has a�place for a single tuple. Insert and HOT update
+for double XMAX pages is impossible and not supported. We can only read or
+delete tuples from it.
+
+When we are able to prune page double XMAX it will be converted from it to
+general 64-bit XID page format with all operations on its tuples supported.
+
+In-memory tuple format
+----------------------
+
+In-memory tuple representation consists of two parts:
+- HeapTupleHeader from disk page (contains all heap tuple contents, not only
+header)
+- HeapTuple with additional in-memory fields
+
+HeapTuple for each tuple in memory stores t_xid_base/t_multi_base - a copies of
+page's pd_xid_base/pd_multi_base. With tuple's 32-bit t_xmin and t_xmax from
+HeapTupleHeader they are used to calculate actual 64-bit XMIN and XMAX:
+
+XMIN = t_xmin + t_xid_base. 					(3)
+XMAX = t_xmax + t_xid_base/t_multi_base.		(4)
+
+The downside of this is that we can not use tuple's XMIN and XMAX right away.
+We often need to re-read t_xmin and t_xmax - which could actually be pointers
+into a page in shared buffers and therefore they could be updated by any other
+backend.
+
+Update/delete with 64-bit XIDs and 32-bit t_xmin/t_xmax
+--------------------------------------------------------------
+
+When we try to delete/update a tuple, we check that XMAX for a page fits (2).
+I.e. that t_xmax will not be over MaxShortTransactionId relative to
+pd_xid_base/pd_multi_base of a its page.
+
+If the current XID doesn't fit a range
+(pd_xid_base, pd_xid_base + MaxShortTransactionId) (5):
+
+- heap_page_prepare_for_xid() will try to increase pd_xid_base/pd_multi_base on
+a page and update all t_xmin/t_xmax of the other tuples on the page to
+correspond new pd_xid_base/pd_multi_base.
+
+- If it was impossible, it will try to prune and freeze tuples on a page.
+
+- If this is unsuccessful it will throw an error. Normally this is very
+unlikely but if there is a very old living transaction with an age of around
+2^32 this can arise. Basically, this is a behavior similar to one during the
+vacuum to prevent wraparound when XID was 32-bit. Dba should take care and
+avoid very-long-living transactions with an age close to 2^32. So long-living
+transactions often they are most likely defunct.
+
+Insert with 64-bit XIDs and 32-bit t_xmin/t_xmax
+------------------------------------------------
+
+On insert we check if current XID fits a range (5). Otherwise:
+
+- heap_page_prepare_for_xid() will try to increase pd_xid_base for t_xmin will
+not be over MaxShortTransactionId.
+
+- If it is impossible, then it will try to prune and freeze tuples on a page.
+
+Known issue: if pd_xid_base could not be shifted to accommodate a tuple being
+inserted due to a very long-running transaction, we just throw an error. We
+neither try to insert a�tuple into another page nor mark the current page as
+full. So, in this (unlikely) case we will get regular insert errors on the next
+tries to insert to the page 'locked' by this very long-running transaction.
+
+Upgrade from 32-bit XID versions
+--------------------------------
+
+pg_upgrade doesn't change pages format itself. It is done lazily after.
+
+1. At first heap page read, tuples on a page are repacked to free 16 bytes
+at the end of a page, possibly freeing space from dead tuples.
+
+2A. 16 bytes of pd_special is added if there is a place for it
+
+2B. Page is converted to "Double XMAX" format if there is no place for
+pd_special
+
+3. If a page is in double XMAX format after its first read, and vacuum (or
+micro-vacuum at select query) could prune some tuples and free space for
+pd_special, prune_page will add pd_special and convert page from double XMAX
+to general 64-bit XID page format.
-- 
2.24.3 (Apple Git-128)


--cpok4wp6gsarlzvp--





^ permalink  raw  reply  [nested|flat] 265+ messages in thread

* [PATCH v6] README for 64bit xid
@ 2022-01-10 19:20 Pavel Borisov <[email protected]>
  0 siblings, 0 replies; 265+ messages in thread

From: Pavel Borisov @ 2022-01-10 19:20 UTC (permalink / raw)

Authors:
- Pavel Borisov <[email protected]>
- Maxim Orlov <[email protected]>
- Yura Sokolov <[email protected]> <[email protected]>
---
 src/backend/access/heap/README.XID64 | 128 +++++++++++++++++++++++++++
 1 file changed, 128 insertions(+)
 create mode 100644 src/backend/access/heap/README.XID64

diff --git a/src/backend/access/heap/README.XID64 b/src/backend/access/heap/README.XID64
new file mode 100644
index 00000000000..457ba9b9ef5
--- /dev/null
+++ b/src/backend/access/heap/README.XID64
@@ -0,0 +1,128 @@
+src/backend/access/heap/README.XID64
+
+64-bit Transaction ID's (XID)
+=============================
+
+A limited number (N = 2^32) of XID's required to do vacuum freeze to prevent
+wraparound every N/2 transactions. This causes performance degradation due
+to the need to exclusively lock tables while being vacuumed. In each
+wraparound cycle, SLRU buffers are also being cut.
+
+With 64-bit XID's wraparound is effectively postponed to a very distant
+future. Even in highly loaded systems that had 2^32 transactions per day
+it will take huge 2^31 days before the first enforced "vacuum to prevent
+wraparound"). Buffers cutting and routine vacuum are not enforced, and DBA
+can plan them independently at the time with the least system load and least
+critical for database performance. Also, it can be done less frequently
+(several times a year vs every several days) on systems with transaction rates
+similar to those mentioned above.
+
+On-disk tuple and page format
+-----------------------------
+
+On-disk tuple format remains unchanged. 32-bit t_xmin and t_xmax store the
+lower parts of 64-bit XMIN and XMAX values. Each heap page has additional
+64-bit pd_xid_base and pd_multi_base which are common for all tuples on a page.
+They are placed into a pd_special area - 16 bytes in the end of a heap page.
+Actual XMIN/XMAX for a tuple are calculated upon reading a tuple from a page
+as follows:
+
+XMIN = t_xmin + pd_xid_base. 					(1)
+XMAX = t_xmax + pd_xid_base/pd_multi_base.		(2)
+
+"Double XMAX" page format
+---------------------------------
+
+At first read of a heap page after pg_upgrade from 32-bit XID PostgreSQL
+version pd_special area with a size of 16 bytes should be added to a page.
+Though a page may not have space for this. Then it can be converted to a
+temporary format called "double XMAX".
+
+All tuples after pg-upgrade would necessarily have xmin = FrozenTransactionId.
+So we don't need tuple header t_xmin field and we reuse t_xmin to store higher
+32 bits of its XMAX.
+
+Double XMAX format is only for full pages that don't have 16 bytes for
+pd_special. So it neither has a�place for a single tuple. Insert and HOT update
+for double XMAX pages is impossible and not supported. We can only read or
+delete tuples from it.
+
+When we are able to prune page double XMAX it will be converted from it to
+general 64-bit XID page format with all operations on its tuples supported.
+
+In-memory tuple format
+----------------------
+
+In-memory tuple representation consists of two parts:
+- HeapTupleHeader from disk page (contains all heap tuple contents, not only
+header)
+- HeapTuple with additional in-memory fields
+
+HeapTuple for each tuple in memory stores t_xid_base/t_multi_base - a copies of
+page's pd_xid_base/pd_multi_base. With tuple's 32-bit t_xmin and t_xmax from
+HeapTupleHeader they are used to calculate actual 64-bit XMIN and XMAX:
+
+XMIN = t_xmin + t_xid_base. 					(3)
+XMAX = t_xmax + t_xid_base/t_multi_base.		(4)
+
+The downside of this is that we can not use tuple's XMIN and XMAX right away.
+We often need to re-read t_xmin and t_xmax - which could actually be pointers
+into a page in shared buffers and therefore they could be updated by any other
+backend.
+
+Update/delete with 64-bit XIDs and 32-bit t_xmin/t_xmax
+--------------------------------------------------------------
+
+When we try to delete/update a tuple, we check that XMAX for a page fits (2).
+I.e. that t_xmax will not be over MaxShortTransactionId relative to
+pd_xid_base/pd_multi_base of a its page.
+
+If the current XID doesn't fit a range
+(pd_xid_base, pd_xid_base + MaxShortTransactionId) (5):
+
+- heap_page_prepare_for_xid() will try to increase pd_xid_base/pd_multi_base on
+a page and update all t_xmin/t_xmax of the other tuples on the page to
+correspond new pd_xid_base/pd_multi_base.
+
+- If it was impossible, it will try to prune and freeze tuples on a page.
+
+- If this is unsuccessful it will throw an error. Normally this is very
+unlikely but if there is a very old living transaction with an age of around
+2^32 this can arise. Basically, this is a behavior similar to one during the
+vacuum to prevent wraparound when XID was 32-bit. Dba should take care and
+avoid very-long-living transactions with an age close to 2^32. So long-living
+transactions often they are most likely defunct.
+
+Insert with 64-bit XIDs and 32-bit t_xmin/t_xmax
+------------------------------------------------
+
+On insert we check if current XID fits a range (5). Otherwise:
+
+- heap_page_prepare_for_xid() will try to increase pd_xid_base for t_xmin will
+not be over MaxShortTransactionId.
+
+- If it is impossible, then it will try to prune and freeze tuples on a page.
+
+Known issue: if pd_xid_base could not be shifted to accommodate a tuple being
+inserted due to a very long-running transaction, we just throw an error. We
+neither try to insert a�tuple into another page nor mark the current page as
+full. So, in this (unlikely) case we will get regular insert errors on the next
+tries to insert to the page 'locked' by this very long-running transaction.
+
+Upgrade from 32-bit XID versions
+--------------------------------
+
+pg_upgrade doesn't change pages format itself. It is done lazily after.
+
+1. At first heap page read, tuples on a page are repacked to free 16 bytes
+at the end of a page, possibly freeing space from dead tuples.
+
+2A. 16 bytes of pd_special is added if there is a place for it
+
+2B. Page is converted to "Double XMAX" format if there is no place for
+pd_special
+
+3. If a page is in double XMAX format after its first read, and vacuum (or
+micro-vacuum at select query) could prune some tuples and free space for
+pd_special, prune_page will add pd_special and convert page from double XMAX
+to general 64-bit XID page format.
-- 
2.24.3 (Apple Git-128)


--cpok4wp6gsarlzvp--





^ permalink  raw  reply  [nested|flat] 265+ messages in thread

* [PATCH v6] README for 64bit xid
@ 2022-01-10 19:20 Pavel Borisov <[email protected]>
  0 siblings, 0 replies; 265+ messages in thread

From: Pavel Borisov @ 2022-01-10 19:20 UTC (permalink / raw)

Authors:
- Pavel Borisov <[email protected]>
- Maxim Orlov <[email protected]>
- Yura Sokolov <[email protected]> <[email protected]>
---
 src/backend/access/heap/README.XID64 | 128 +++++++++++++++++++++++++++
 1 file changed, 128 insertions(+)
 create mode 100644 src/backend/access/heap/README.XID64

diff --git a/src/backend/access/heap/README.XID64 b/src/backend/access/heap/README.XID64
new file mode 100644
index 00000000000..457ba9b9ef5
--- /dev/null
+++ b/src/backend/access/heap/README.XID64
@@ -0,0 +1,128 @@
+src/backend/access/heap/README.XID64
+
+64-bit Transaction ID's (XID)
+=============================
+
+A limited number (N = 2^32) of XID's required to do vacuum freeze to prevent
+wraparound every N/2 transactions. This causes performance degradation due
+to the need to exclusively lock tables while being vacuumed. In each
+wraparound cycle, SLRU buffers are also being cut.
+
+With 64-bit XID's wraparound is effectively postponed to a very distant
+future. Even in highly loaded systems that had 2^32 transactions per day
+it will take huge 2^31 days before the first enforced "vacuum to prevent
+wraparound"). Buffers cutting and routine vacuum are not enforced, and DBA
+can plan them independently at the time with the least system load and least
+critical for database performance. Also, it can be done less frequently
+(several times a year vs every several days) on systems with transaction rates
+similar to those mentioned above.
+
+On-disk tuple and page format
+-----------------------------
+
+On-disk tuple format remains unchanged. 32-bit t_xmin and t_xmax store the
+lower parts of 64-bit XMIN and XMAX values. Each heap page has additional
+64-bit pd_xid_base and pd_multi_base which are common for all tuples on a page.
+They are placed into a pd_special area - 16 bytes in the end of a heap page.
+Actual XMIN/XMAX for a tuple are calculated upon reading a tuple from a page
+as follows:
+
+XMIN = t_xmin + pd_xid_base. 					(1)
+XMAX = t_xmax + pd_xid_base/pd_multi_base.		(2)
+
+"Double XMAX" page format
+---------------------------------
+
+At first read of a heap page after pg_upgrade from 32-bit XID PostgreSQL
+version pd_special area with a size of 16 bytes should be added to a page.
+Though a page may not have space for this. Then it can be converted to a
+temporary format called "double XMAX".
+
+All tuples after pg-upgrade would necessarily have xmin = FrozenTransactionId.
+So we don't need tuple header t_xmin field and we reuse t_xmin to store higher
+32 bits of its XMAX.
+
+Double XMAX format is only for full pages that don't have 16 bytes for
+pd_special. So it neither has a�place for a single tuple. Insert and HOT update
+for double XMAX pages is impossible and not supported. We can only read or
+delete tuples from it.
+
+When we are able to prune page double XMAX it will be converted from it to
+general 64-bit XID page format with all operations on its tuples supported.
+
+In-memory tuple format
+----------------------
+
+In-memory tuple representation consists of two parts:
+- HeapTupleHeader from disk page (contains all heap tuple contents, not only
+header)
+- HeapTuple with additional in-memory fields
+
+HeapTuple for each tuple in memory stores t_xid_base/t_multi_base - a copies of
+page's pd_xid_base/pd_multi_base. With tuple's 32-bit t_xmin and t_xmax from
+HeapTupleHeader they are used to calculate actual 64-bit XMIN and XMAX:
+
+XMIN = t_xmin + t_xid_base. 					(3)
+XMAX = t_xmax + t_xid_base/t_multi_base.		(4)
+
+The downside of this is that we can not use tuple's XMIN and XMAX right away.
+We often need to re-read t_xmin and t_xmax - which could actually be pointers
+into a page in shared buffers and therefore they could be updated by any other
+backend.
+
+Update/delete with 64-bit XIDs and 32-bit t_xmin/t_xmax
+--------------------------------------------------------------
+
+When we try to delete/update a tuple, we check that XMAX for a page fits (2).
+I.e. that t_xmax will not be over MaxShortTransactionId relative to
+pd_xid_base/pd_multi_base of a its page.
+
+If the current XID doesn't fit a range
+(pd_xid_base, pd_xid_base + MaxShortTransactionId) (5):
+
+- heap_page_prepare_for_xid() will try to increase pd_xid_base/pd_multi_base on
+a page and update all t_xmin/t_xmax of the other tuples on the page to
+correspond new pd_xid_base/pd_multi_base.
+
+- If it was impossible, it will try to prune and freeze tuples on a page.
+
+- If this is unsuccessful it will throw an error. Normally this is very
+unlikely but if there is a very old living transaction with an age of around
+2^32 this can arise. Basically, this is a behavior similar to one during the
+vacuum to prevent wraparound when XID was 32-bit. Dba should take care and
+avoid very-long-living transactions with an age close to 2^32. So long-living
+transactions often they are most likely defunct.
+
+Insert with 64-bit XIDs and 32-bit t_xmin/t_xmax
+------------------------------------------------
+
+On insert we check if current XID fits a range (5). Otherwise:
+
+- heap_page_prepare_for_xid() will try to increase pd_xid_base for t_xmin will
+not be over MaxShortTransactionId.
+
+- If it is impossible, then it will try to prune and freeze tuples on a page.
+
+Known issue: if pd_xid_base could not be shifted to accommodate a tuple being
+inserted due to a very long-running transaction, we just throw an error. We
+neither try to insert a�tuple into another page nor mark the current page as
+full. So, in this (unlikely) case we will get regular insert errors on the next
+tries to insert to the page 'locked' by this very long-running transaction.
+
+Upgrade from 32-bit XID versions
+--------------------------------
+
+pg_upgrade doesn't change pages format itself. It is done lazily after.
+
+1. At first heap page read, tuples on a page are repacked to free 16 bytes
+at the end of a page, possibly freeing space from dead tuples.
+
+2A. 16 bytes of pd_special is added if there is a place for it
+
+2B. Page is converted to "Double XMAX" format if there is no place for
+pd_special
+
+3. If a page is in double XMAX format after its first read, and vacuum (or
+micro-vacuum at select query) could prune some tuples and free space for
+pd_special, prune_page will add pd_special and convert page from double XMAX
+to general 64-bit XID page format.
-- 
2.24.3 (Apple Git-128)


--cpok4wp6gsarlzvp--





^ permalink  raw  reply  [nested|flat] 265+ messages in thread

* [PATCH v6] README for 64bit xid
@ 2022-01-10 19:20 Pavel Borisov <[email protected]>
  0 siblings, 0 replies; 265+ messages in thread

From: Pavel Borisov @ 2022-01-10 19:20 UTC (permalink / raw)

Authors:
- Pavel Borisov <[email protected]>
- Maxim Orlov <[email protected]>
- Yura Sokolov <[email protected]> <[email protected]>
---
 src/backend/access/heap/README.XID64 | 128 +++++++++++++++++++++++++++
 1 file changed, 128 insertions(+)
 create mode 100644 src/backend/access/heap/README.XID64

diff --git a/src/backend/access/heap/README.XID64 b/src/backend/access/heap/README.XID64
new file mode 100644
index 00000000000..457ba9b9ef5
--- /dev/null
+++ b/src/backend/access/heap/README.XID64
@@ -0,0 +1,128 @@
+src/backend/access/heap/README.XID64
+
+64-bit Transaction ID's (XID)
+=============================
+
+A limited number (N = 2^32) of XID's required to do vacuum freeze to prevent
+wraparound every N/2 transactions. This causes performance degradation due
+to the need to exclusively lock tables while being vacuumed. In each
+wraparound cycle, SLRU buffers are also being cut.
+
+With 64-bit XID's wraparound is effectively postponed to a very distant
+future. Even in highly loaded systems that had 2^32 transactions per day
+it will take huge 2^31 days before the first enforced "vacuum to prevent
+wraparound"). Buffers cutting and routine vacuum are not enforced, and DBA
+can plan them independently at the time with the least system load and least
+critical for database performance. Also, it can be done less frequently
+(several times a year vs every several days) on systems with transaction rates
+similar to those mentioned above.
+
+On-disk tuple and page format
+-----------------------------
+
+On-disk tuple format remains unchanged. 32-bit t_xmin and t_xmax store the
+lower parts of 64-bit XMIN and XMAX values. Each heap page has additional
+64-bit pd_xid_base and pd_multi_base which are common for all tuples on a page.
+They are placed into a pd_special area - 16 bytes in the end of a heap page.
+Actual XMIN/XMAX for a tuple are calculated upon reading a tuple from a page
+as follows:
+
+XMIN = t_xmin + pd_xid_base. 					(1)
+XMAX = t_xmax + pd_xid_base/pd_multi_base.		(2)
+
+"Double XMAX" page format
+---------------------------------
+
+At first read of a heap page after pg_upgrade from 32-bit XID PostgreSQL
+version pd_special area with a size of 16 bytes should be added to a page.
+Though a page may not have space for this. Then it can be converted to a
+temporary format called "double XMAX".
+
+All tuples after pg-upgrade would necessarily have xmin = FrozenTransactionId.
+So we don't need tuple header t_xmin field and we reuse t_xmin to store higher
+32 bits of its XMAX.
+
+Double XMAX format is only for full pages that don't have 16 bytes for
+pd_special. So it neither has a�place for a single tuple. Insert and HOT update
+for double XMAX pages is impossible and not supported. We can only read or
+delete tuples from it.
+
+When we are able to prune page double XMAX it will be converted from it to
+general 64-bit XID page format with all operations on its tuples supported.
+
+In-memory tuple format
+----------------------
+
+In-memory tuple representation consists of two parts:
+- HeapTupleHeader from disk page (contains all heap tuple contents, not only
+header)
+- HeapTuple with additional in-memory fields
+
+HeapTuple for each tuple in memory stores t_xid_base/t_multi_base - a copies of
+page's pd_xid_base/pd_multi_base. With tuple's 32-bit t_xmin and t_xmax from
+HeapTupleHeader they are used to calculate actual 64-bit XMIN and XMAX:
+
+XMIN = t_xmin + t_xid_base. 					(3)
+XMAX = t_xmax + t_xid_base/t_multi_base.		(4)
+
+The downside of this is that we can not use tuple's XMIN and XMAX right away.
+We often need to re-read t_xmin and t_xmax - which could actually be pointers
+into a page in shared buffers and therefore they could be updated by any other
+backend.
+
+Update/delete with 64-bit XIDs and 32-bit t_xmin/t_xmax
+--------------------------------------------------------------
+
+When we try to delete/update a tuple, we check that XMAX for a page fits (2).
+I.e. that t_xmax will not be over MaxShortTransactionId relative to
+pd_xid_base/pd_multi_base of a its page.
+
+If the current XID doesn't fit a range
+(pd_xid_base, pd_xid_base + MaxShortTransactionId) (5):
+
+- heap_page_prepare_for_xid() will try to increase pd_xid_base/pd_multi_base on
+a page and update all t_xmin/t_xmax of the other tuples on the page to
+correspond new pd_xid_base/pd_multi_base.
+
+- If it was impossible, it will try to prune and freeze tuples on a page.
+
+- If this is unsuccessful it will throw an error. Normally this is very
+unlikely but if there is a very old living transaction with an age of around
+2^32 this can arise. Basically, this is a behavior similar to one during the
+vacuum to prevent wraparound when XID was 32-bit. Dba should take care and
+avoid very-long-living transactions with an age close to 2^32. So long-living
+transactions often they are most likely defunct.
+
+Insert with 64-bit XIDs and 32-bit t_xmin/t_xmax
+------------------------------------------------
+
+On insert we check if current XID fits a range (5). Otherwise:
+
+- heap_page_prepare_for_xid() will try to increase pd_xid_base for t_xmin will
+not be over MaxShortTransactionId.
+
+- If it is impossible, then it will try to prune and freeze tuples on a page.
+
+Known issue: if pd_xid_base could not be shifted to accommodate a tuple being
+inserted due to a very long-running transaction, we just throw an error. We
+neither try to insert a�tuple into another page nor mark the current page as
+full. So, in this (unlikely) case we will get regular insert errors on the next
+tries to insert to the page 'locked' by this very long-running transaction.
+
+Upgrade from 32-bit XID versions
+--------------------------------
+
+pg_upgrade doesn't change pages format itself. It is done lazily after.
+
+1. At first heap page read, tuples on a page are repacked to free 16 bytes
+at the end of a page, possibly freeing space from dead tuples.
+
+2A. 16 bytes of pd_special is added if there is a place for it
+
+2B. Page is converted to "Double XMAX" format if there is no place for
+pd_special
+
+3. If a page is in double XMAX format after its first read, and vacuum (or
+micro-vacuum at select query) could prune some tuples and free space for
+pd_special, prune_page will add pd_special and convert page from double XMAX
+to general 64-bit XID page format.
-- 
2.24.3 (Apple Git-128)


--cpok4wp6gsarlzvp--





^ permalink  raw  reply  [nested|flat] 265+ messages in thread

* [PATCH v6] README for 64bit xid
@ 2022-01-10 19:20 Pavel Borisov <[email protected]>
  0 siblings, 0 replies; 265+ messages in thread

From: Pavel Borisov @ 2022-01-10 19:20 UTC (permalink / raw)

Authors:
- Pavel Borisov <[email protected]>
- Maxim Orlov <[email protected]>
- Yura Sokolov <[email protected]> <[email protected]>
---
 src/backend/access/heap/README.XID64 | 128 +++++++++++++++++++++++++++
 1 file changed, 128 insertions(+)
 create mode 100644 src/backend/access/heap/README.XID64

diff --git a/src/backend/access/heap/README.XID64 b/src/backend/access/heap/README.XID64
new file mode 100644
index 00000000000..457ba9b9ef5
--- /dev/null
+++ b/src/backend/access/heap/README.XID64
@@ -0,0 +1,128 @@
+src/backend/access/heap/README.XID64
+
+64-bit Transaction ID's (XID)
+=============================
+
+A limited number (N = 2^32) of XID's required to do vacuum freeze to prevent
+wraparound every N/2 transactions. This causes performance degradation due
+to the need to exclusively lock tables while being vacuumed. In each
+wraparound cycle, SLRU buffers are also being cut.
+
+With 64-bit XID's wraparound is effectively postponed to a very distant
+future. Even in highly loaded systems that had 2^32 transactions per day
+it will take huge 2^31 days before the first enforced "vacuum to prevent
+wraparound"). Buffers cutting and routine vacuum are not enforced, and DBA
+can plan them independently at the time with the least system load and least
+critical for database performance. Also, it can be done less frequently
+(several times a year vs every several days) on systems with transaction rates
+similar to those mentioned above.
+
+On-disk tuple and page format
+-----------------------------
+
+On-disk tuple format remains unchanged. 32-bit t_xmin and t_xmax store the
+lower parts of 64-bit XMIN and XMAX values. Each heap page has additional
+64-bit pd_xid_base and pd_multi_base which are common for all tuples on a page.
+They are placed into a pd_special area - 16 bytes in the end of a heap page.
+Actual XMIN/XMAX for a tuple are calculated upon reading a tuple from a page
+as follows:
+
+XMIN = t_xmin + pd_xid_base. 					(1)
+XMAX = t_xmax + pd_xid_base/pd_multi_base.		(2)
+
+"Double XMAX" page format
+---------------------------------
+
+At first read of a heap page after pg_upgrade from 32-bit XID PostgreSQL
+version pd_special area with a size of 16 bytes should be added to a page.
+Though a page may not have space for this. Then it can be converted to a
+temporary format called "double XMAX".
+
+All tuples after pg-upgrade would necessarily have xmin = FrozenTransactionId.
+So we don't need tuple header t_xmin field and we reuse t_xmin to store higher
+32 bits of its XMAX.
+
+Double XMAX format is only for full pages that don't have 16 bytes for
+pd_special. So it neither has a�place for a single tuple. Insert and HOT update
+for double XMAX pages is impossible and not supported. We can only read or
+delete tuples from it.
+
+When we are able to prune page double XMAX it will be converted from it to
+general 64-bit XID page format with all operations on its tuples supported.
+
+In-memory tuple format
+----------------------
+
+In-memory tuple representation consists of two parts:
+- HeapTupleHeader from disk page (contains all heap tuple contents, not only
+header)
+- HeapTuple with additional in-memory fields
+
+HeapTuple for each tuple in memory stores t_xid_base/t_multi_base - a copies of
+page's pd_xid_base/pd_multi_base. With tuple's 32-bit t_xmin and t_xmax from
+HeapTupleHeader they are used to calculate actual 64-bit XMIN and XMAX:
+
+XMIN = t_xmin + t_xid_base. 					(3)
+XMAX = t_xmax + t_xid_base/t_multi_base.		(4)
+
+The downside of this is that we can not use tuple's XMIN and XMAX right away.
+We often need to re-read t_xmin and t_xmax - which could actually be pointers
+into a page in shared buffers and therefore they could be updated by any other
+backend.
+
+Update/delete with 64-bit XIDs and 32-bit t_xmin/t_xmax
+--------------------------------------------------------------
+
+When we try to delete/update a tuple, we check that XMAX for a page fits (2).
+I.e. that t_xmax will not be over MaxShortTransactionId relative to
+pd_xid_base/pd_multi_base of a its page.
+
+If the current XID doesn't fit a range
+(pd_xid_base, pd_xid_base + MaxShortTransactionId) (5):
+
+- heap_page_prepare_for_xid() will try to increase pd_xid_base/pd_multi_base on
+a page and update all t_xmin/t_xmax of the other tuples on the page to
+correspond new pd_xid_base/pd_multi_base.
+
+- If it was impossible, it will try to prune and freeze tuples on a page.
+
+- If this is unsuccessful it will throw an error. Normally this is very
+unlikely but if there is a very old living transaction with an age of around
+2^32 this can arise. Basically, this is a behavior similar to one during the
+vacuum to prevent wraparound when XID was 32-bit. Dba should take care and
+avoid very-long-living transactions with an age close to 2^32. So long-living
+transactions often they are most likely defunct.
+
+Insert with 64-bit XIDs and 32-bit t_xmin/t_xmax
+------------------------------------------------
+
+On insert we check if current XID fits a range (5). Otherwise:
+
+- heap_page_prepare_for_xid() will try to increase pd_xid_base for t_xmin will
+not be over MaxShortTransactionId.
+
+- If it is impossible, then it will try to prune and freeze tuples on a page.
+
+Known issue: if pd_xid_base could not be shifted to accommodate a tuple being
+inserted due to a very long-running transaction, we just throw an error. We
+neither try to insert a�tuple into another page nor mark the current page as
+full. So, in this (unlikely) case we will get regular insert errors on the next
+tries to insert to the page 'locked' by this very long-running transaction.
+
+Upgrade from 32-bit XID versions
+--------------------------------
+
+pg_upgrade doesn't change pages format itself. It is done lazily after.
+
+1. At first heap page read, tuples on a page are repacked to free 16 bytes
+at the end of a page, possibly freeing space from dead tuples.
+
+2A. 16 bytes of pd_special is added if there is a place for it
+
+2B. Page is converted to "Double XMAX" format if there is no place for
+pd_special
+
+3. If a page is in double XMAX format after its first read, and vacuum (or
+micro-vacuum at select query) could prune some tuples and free space for
+pd_special, prune_page will add pd_special and convert page from double XMAX
+to general 64-bit XID page format.
-- 
2.24.3 (Apple Git-128)


--cpok4wp6gsarlzvp--





^ permalink  raw  reply  [nested|flat] 265+ messages in thread

* [PATCH v6] README for 64bit xid
@ 2022-01-10 19:20 Pavel Borisov <[email protected]>
  0 siblings, 0 replies; 265+ messages in thread

From: Pavel Borisov @ 2022-01-10 19:20 UTC (permalink / raw)

Authors:
- Pavel Borisov <[email protected]>
- Maxim Orlov <[email protected]>
- Yura Sokolov <[email protected]> <[email protected]>
---
 src/backend/access/heap/README.XID64 | 128 +++++++++++++++++++++++++++
 1 file changed, 128 insertions(+)
 create mode 100644 src/backend/access/heap/README.XID64

diff --git a/src/backend/access/heap/README.XID64 b/src/backend/access/heap/README.XID64
new file mode 100644
index 00000000000..457ba9b9ef5
--- /dev/null
+++ b/src/backend/access/heap/README.XID64
@@ -0,0 +1,128 @@
+src/backend/access/heap/README.XID64
+
+64-bit Transaction ID's (XID)
+=============================
+
+A limited number (N = 2^32) of XID's required to do vacuum freeze to prevent
+wraparound every N/2 transactions. This causes performance degradation due
+to the need to exclusively lock tables while being vacuumed. In each
+wraparound cycle, SLRU buffers are also being cut.
+
+With 64-bit XID's wraparound is effectively postponed to a very distant
+future. Even in highly loaded systems that had 2^32 transactions per day
+it will take huge 2^31 days before the first enforced "vacuum to prevent
+wraparound"). Buffers cutting and routine vacuum are not enforced, and DBA
+can plan them independently at the time with the least system load and least
+critical for database performance. Also, it can be done less frequently
+(several times a year vs every several days) on systems with transaction rates
+similar to those mentioned above.
+
+On-disk tuple and page format
+-----------------------------
+
+On-disk tuple format remains unchanged. 32-bit t_xmin and t_xmax store the
+lower parts of 64-bit XMIN and XMAX values. Each heap page has additional
+64-bit pd_xid_base and pd_multi_base which are common for all tuples on a page.
+They are placed into a pd_special area - 16 bytes in the end of a heap page.
+Actual XMIN/XMAX for a tuple are calculated upon reading a tuple from a page
+as follows:
+
+XMIN = t_xmin + pd_xid_base. 					(1)
+XMAX = t_xmax + pd_xid_base/pd_multi_base.		(2)
+
+"Double XMAX" page format
+---------------------------------
+
+At first read of a heap page after pg_upgrade from 32-bit XID PostgreSQL
+version pd_special area with a size of 16 bytes should be added to a page.
+Though a page may not have space for this. Then it can be converted to a
+temporary format called "double XMAX".
+
+All tuples after pg-upgrade would necessarily have xmin = FrozenTransactionId.
+So we don't need tuple header t_xmin field and we reuse t_xmin to store higher
+32 bits of its XMAX.
+
+Double XMAX format is only for full pages that don't have 16 bytes for
+pd_special. So it neither has a�place for a single tuple. Insert and HOT update
+for double XMAX pages is impossible and not supported. We can only read or
+delete tuples from it.
+
+When we are able to prune page double XMAX it will be converted from it to
+general 64-bit XID page format with all operations on its tuples supported.
+
+In-memory tuple format
+----------------------
+
+In-memory tuple representation consists of two parts:
+- HeapTupleHeader from disk page (contains all heap tuple contents, not only
+header)
+- HeapTuple with additional in-memory fields
+
+HeapTuple for each tuple in memory stores t_xid_base/t_multi_base - a copies of
+page's pd_xid_base/pd_multi_base. With tuple's 32-bit t_xmin and t_xmax from
+HeapTupleHeader they are used to calculate actual 64-bit XMIN and XMAX:
+
+XMIN = t_xmin + t_xid_base. 					(3)
+XMAX = t_xmax + t_xid_base/t_multi_base.		(4)
+
+The downside of this is that we can not use tuple's XMIN and XMAX right away.
+We often need to re-read t_xmin and t_xmax - which could actually be pointers
+into a page in shared buffers and therefore they could be updated by any other
+backend.
+
+Update/delete with 64-bit XIDs and 32-bit t_xmin/t_xmax
+--------------------------------------------------------------
+
+When we try to delete/update a tuple, we check that XMAX for a page fits (2).
+I.e. that t_xmax will not be over MaxShortTransactionId relative to
+pd_xid_base/pd_multi_base of a its page.
+
+If the current XID doesn't fit a range
+(pd_xid_base, pd_xid_base + MaxShortTransactionId) (5):
+
+- heap_page_prepare_for_xid() will try to increase pd_xid_base/pd_multi_base on
+a page and update all t_xmin/t_xmax of the other tuples on the page to
+correspond new pd_xid_base/pd_multi_base.
+
+- If it was impossible, it will try to prune and freeze tuples on a page.
+
+- If this is unsuccessful it will throw an error. Normally this is very
+unlikely but if there is a very old living transaction with an age of around
+2^32 this can arise. Basically, this is a behavior similar to one during the
+vacuum to prevent wraparound when XID was 32-bit. Dba should take care and
+avoid very-long-living transactions with an age close to 2^32. So long-living
+transactions often they are most likely defunct.
+
+Insert with 64-bit XIDs and 32-bit t_xmin/t_xmax
+------------------------------------------------
+
+On insert we check if current XID fits a range (5). Otherwise:
+
+- heap_page_prepare_for_xid() will try to increase pd_xid_base for t_xmin will
+not be over MaxShortTransactionId.
+
+- If it is impossible, then it will try to prune and freeze tuples on a page.
+
+Known issue: if pd_xid_base could not be shifted to accommodate a tuple being
+inserted due to a very long-running transaction, we just throw an error. We
+neither try to insert a�tuple into another page nor mark the current page as
+full. So, in this (unlikely) case we will get regular insert errors on the next
+tries to insert to the page 'locked' by this very long-running transaction.
+
+Upgrade from 32-bit XID versions
+--------------------------------
+
+pg_upgrade doesn't change pages format itself. It is done lazily after.
+
+1. At first heap page read, tuples on a page are repacked to free 16 bytes
+at the end of a page, possibly freeing space from dead tuples.
+
+2A. 16 bytes of pd_special is added if there is a place for it
+
+2B. Page is converted to "Double XMAX" format if there is no place for
+pd_special
+
+3. If a page is in double XMAX format after its first read, and vacuum (or
+micro-vacuum at select query) could prune some tuples and free space for
+pd_special, prune_page will add pd_special and convert page from double XMAX
+to general 64-bit XID page format.
-- 
2.24.3 (Apple Git-128)


--cpok4wp6gsarlzvp--





^ permalink  raw  reply  [nested|flat] 265+ messages in thread

* [PATCH v6] README for 64bit xid
@ 2022-01-10 19:20 Pavel Borisov <[email protected]>
  0 siblings, 0 replies; 265+ messages in thread

From: Pavel Borisov @ 2022-01-10 19:20 UTC (permalink / raw)

Authors:
- Pavel Borisov <[email protected]>
- Maxim Orlov <[email protected]>
- Yura Sokolov <[email protected]> <[email protected]>
---
 src/backend/access/heap/README.XID64 | 128 +++++++++++++++++++++++++++
 1 file changed, 128 insertions(+)
 create mode 100644 src/backend/access/heap/README.XID64

diff --git a/src/backend/access/heap/README.XID64 b/src/backend/access/heap/README.XID64
new file mode 100644
index 00000000000..457ba9b9ef5
--- /dev/null
+++ b/src/backend/access/heap/README.XID64
@@ -0,0 +1,128 @@
+src/backend/access/heap/README.XID64
+
+64-bit Transaction ID's (XID)
+=============================
+
+A limited number (N = 2^32) of XID's required to do vacuum freeze to prevent
+wraparound every N/2 transactions. This causes performance degradation due
+to the need to exclusively lock tables while being vacuumed. In each
+wraparound cycle, SLRU buffers are also being cut.
+
+With 64-bit XID's wraparound is effectively postponed to a very distant
+future. Even in highly loaded systems that had 2^32 transactions per day
+it will take huge 2^31 days before the first enforced "vacuum to prevent
+wraparound"). Buffers cutting and routine vacuum are not enforced, and DBA
+can plan them independently at the time with the least system load and least
+critical for database performance. Also, it can be done less frequently
+(several times a year vs every several days) on systems with transaction rates
+similar to those mentioned above.
+
+On-disk tuple and page format
+-----------------------------
+
+On-disk tuple format remains unchanged. 32-bit t_xmin and t_xmax store the
+lower parts of 64-bit XMIN and XMAX values. Each heap page has additional
+64-bit pd_xid_base and pd_multi_base which are common for all tuples on a page.
+They are placed into a pd_special area - 16 bytes in the end of a heap page.
+Actual XMIN/XMAX for a tuple are calculated upon reading a tuple from a page
+as follows:
+
+XMIN = t_xmin + pd_xid_base. 					(1)
+XMAX = t_xmax + pd_xid_base/pd_multi_base.		(2)
+
+"Double XMAX" page format
+---------------------------------
+
+At first read of a heap page after pg_upgrade from 32-bit XID PostgreSQL
+version pd_special area with a size of 16 bytes should be added to a page.
+Though a page may not have space for this. Then it can be converted to a
+temporary format called "double XMAX".
+
+All tuples after pg-upgrade would necessarily have xmin = FrozenTransactionId.
+So we don't need tuple header t_xmin field and we reuse t_xmin to store higher
+32 bits of its XMAX.
+
+Double XMAX format is only for full pages that don't have 16 bytes for
+pd_special. So it neither has a�place for a single tuple. Insert and HOT update
+for double XMAX pages is impossible and not supported. We can only read or
+delete tuples from it.
+
+When we are able to prune page double XMAX it will be converted from it to
+general 64-bit XID page format with all operations on its tuples supported.
+
+In-memory tuple format
+----------------------
+
+In-memory tuple representation consists of two parts:
+- HeapTupleHeader from disk page (contains all heap tuple contents, not only
+header)
+- HeapTuple with additional in-memory fields
+
+HeapTuple for each tuple in memory stores t_xid_base/t_multi_base - a copies of
+page's pd_xid_base/pd_multi_base. With tuple's 32-bit t_xmin and t_xmax from
+HeapTupleHeader they are used to calculate actual 64-bit XMIN and XMAX:
+
+XMIN = t_xmin + t_xid_base. 					(3)
+XMAX = t_xmax + t_xid_base/t_multi_base.		(4)
+
+The downside of this is that we can not use tuple's XMIN and XMAX right away.
+We often need to re-read t_xmin and t_xmax - which could actually be pointers
+into a page in shared buffers and therefore they could be updated by any other
+backend.
+
+Update/delete with 64-bit XIDs and 32-bit t_xmin/t_xmax
+--------------------------------------------------------------
+
+When we try to delete/update a tuple, we check that XMAX for a page fits (2).
+I.e. that t_xmax will not be over MaxShortTransactionId relative to
+pd_xid_base/pd_multi_base of a its page.
+
+If the current XID doesn't fit a range
+(pd_xid_base, pd_xid_base + MaxShortTransactionId) (5):
+
+- heap_page_prepare_for_xid() will try to increase pd_xid_base/pd_multi_base on
+a page and update all t_xmin/t_xmax of the other tuples on the page to
+correspond new pd_xid_base/pd_multi_base.
+
+- If it was impossible, it will try to prune and freeze tuples on a page.
+
+- If this is unsuccessful it will throw an error. Normally this is very
+unlikely but if there is a very old living transaction with an age of around
+2^32 this can arise. Basically, this is a behavior similar to one during the
+vacuum to prevent wraparound when XID was 32-bit. Dba should take care and
+avoid very-long-living transactions with an age close to 2^32. So long-living
+transactions often they are most likely defunct.
+
+Insert with 64-bit XIDs and 32-bit t_xmin/t_xmax
+------------------------------------------------
+
+On insert we check if current XID fits a range (5). Otherwise:
+
+- heap_page_prepare_for_xid() will try to increase pd_xid_base for t_xmin will
+not be over MaxShortTransactionId.
+
+- If it is impossible, then it will try to prune and freeze tuples on a page.
+
+Known issue: if pd_xid_base could not be shifted to accommodate a tuple being
+inserted due to a very long-running transaction, we just throw an error. We
+neither try to insert a�tuple into another page nor mark the current page as
+full. So, in this (unlikely) case we will get regular insert errors on the next
+tries to insert to the page 'locked' by this very long-running transaction.
+
+Upgrade from 32-bit XID versions
+--------------------------------
+
+pg_upgrade doesn't change pages format itself. It is done lazily after.
+
+1. At first heap page read, tuples on a page are repacked to free 16 bytes
+at the end of a page, possibly freeing space from dead tuples.
+
+2A. 16 bytes of pd_special is added if there is a place for it
+
+2B. Page is converted to "Double XMAX" format if there is no place for
+pd_special
+
+3. If a page is in double XMAX format after its first read, and vacuum (or
+micro-vacuum at select query) could prune some tuples and free space for
+pd_special, prune_page will add pd_special and convert page from double XMAX
+to general 64-bit XID page format.
-- 
2.24.3 (Apple Git-128)


--cpok4wp6gsarlzvp--





^ permalink  raw  reply  [nested|flat] 265+ messages in thread

* [PATCH v6] README for 64bit xid
@ 2022-01-10 19:20 Pavel Borisov <[email protected]>
  0 siblings, 0 replies; 265+ messages in thread

From: Pavel Borisov @ 2022-01-10 19:20 UTC (permalink / raw)

Authors:
- Pavel Borisov <[email protected]>
- Maxim Orlov <[email protected]>
- Yura Sokolov <[email protected]> <[email protected]>
---
 src/backend/access/heap/README.XID64 | 128 +++++++++++++++++++++++++++
 1 file changed, 128 insertions(+)
 create mode 100644 src/backend/access/heap/README.XID64

diff --git a/src/backend/access/heap/README.XID64 b/src/backend/access/heap/README.XID64
new file mode 100644
index 00000000000..457ba9b9ef5
--- /dev/null
+++ b/src/backend/access/heap/README.XID64
@@ -0,0 +1,128 @@
+src/backend/access/heap/README.XID64
+
+64-bit Transaction ID's (XID)
+=============================
+
+A limited number (N = 2^32) of XID's required to do vacuum freeze to prevent
+wraparound every N/2 transactions. This causes performance degradation due
+to the need to exclusively lock tables while being vacuumed. In each
+wraparound cycle, SLRU buffers are also being cut.
+
+With 64-bit XID's wraparound is effectively postponed to a very distant
+future. Even in highly loaded systems that had 2^32 transactions per day
+it will take huge 2^31 days before the first enforced "vacuum to prevent
+wraparound"). Buffers cutting and routine vacuum are not enforced, and DBA
+can plan them independently at the time with the least system load and least
+critical for database performance. Also, it can be done less frequently
+(several times a year vs every several days) on systems with transaction rates
+similar to those mentioned above.
+
+On-disk tuple and page format
+-----------------------------
+
+On-disk tuple format remains unchanged. 32-bit t_xmin and t_xmax store the
+lower parts of 64-bit XMIN and XMAX values. Each heap page has additional
+64-bit pd_xid_base and pd_multi_base which are common for all tuples on a page.
+They are placed into a pd_special area - 16 bytes in the end of a heap page.
+Actual XMIN/XMAX for a tuple are calculated upon reading a tuple from a page
+as follows:
+
+XMIN = t_xmin + pd_xid_base. 					(1)
+XMAX = t_xmax + pd_xid_base/pd_multi_base.		(2)
+
+"Double XMAX" page format
+---------------------------------
+
+At first read of a heap page after pg_upgrade from 32-bit XID PostgreSQL
+version pd_special area with a size of 16 bytes should be added to a page.
+Though a page may not have space for this. Then it can be converted to a
+temporary format called "double XMAX".
+
+All tuples after pg-upgrade would necessarily have xmin = FrozenTransactionId.
+So we don't need tuple header t_xmin field and we reuse t_xmin to store higher
+32 bits of its XMAX.
+
+Double XMAX format is only for full pages that don't have 16 bytes for
+pd_special. So it neither has a�place for a single tuple. Insert and HOT update
+for double XMAX pages is impossible and not supported. We can only read or
+delete tuples from it.
+
+When we are able to prune page double XMAX it will be converted from it to
+general 64-bit XID page format with all operations on its tuples supported.
+
+In-memory tuple format
+----------------------
+
+In-memory tuple representation consists of two parts:
+- HeapTupleHeader from disk page (contains all heap tuple contents, not only
+header)
+- HeapTuple with additional in-memory fields
+
+HeapTuple for each tuple in memory stores t_xid_base/t_multi_base - a copies of
+page's pd_xid_base/pd_multi_base. With tuple's 32-bit t_xmin and t_xmax from
+HeapTupleHeader they are used to calculate actual 64-bit XMIN and XMAX:
+
+XMIN = t_xmin + t_xid_base. 					(3)
+XMAX = t_xmax + t_xid_base/t_multi_base.		(4)
+
+The downside of this is that we can not use tuple's XMIN and XMAX right away.
+We often need to re-read t_xmin and t_xmax - which could actually be pointers
+into a page in shared buffers and therefore they could be updated by any other
+backend.
+
+Update/delete with 64-bit XIDs and 32-bit t_xmin/t_xmax
+--------------------------------------------------------------
+
+When we try to delete/update a tuple, we check that XMAX for a page fits (2).
+I.e. that t_xmax will not be over MaxShortTransactionId relative to
+pd_xid_base/pd_multi_base of a its page.
+
+If the current XID doesn't fit a range
+(pd_xid_base, pd_xid_base + MaxShortTransactionId) (5):
+
+- heap_page_prepare_for_xid() will try to increase pd_xid_base/pd_multi_base on
+a page and update all t_xmin/t_xmax of the other tuples on the page to
+correspond new pd_xid_base/pd_multi_base.
+
+- If it was impossible, it will try to prune and freeze tuples on a page.
+
+- If this is unsuccessful it will throw an error. Normally this is very
+unlikely but if there is a very old living transaction with an age of around
+2^32 this can arise. Basically, this is a behavior similar to one during the
+vacuum to prevent wraparound when XID was 32-bit. Dba should take care and
+avoid very-long-living transactions with an age close to 2^32. So long-living
+transactions often they are most likely defunct.
+
+Insert with 64-bit XIDs and 32-bit t_xmin/t_xmax
+------------------------------------------------
+
+On insert we check if current XID fits a range (5). Otherwise:
+
+- heap_page_prepare_for_xid() will try to increase pd_xid_base for t_xmin will
+not be over MaxShortTransactionId.
+
+- If it is impossible, then it will try to prune and freeze tuples on a page.
+
+Known issue: if pd_xid_base could not be shifted to accommodate a tuple being
+inserted due to a very long-running transaction, we just throw an error. We
+neither try to insert a�tuple into another page nor mark the current page as
+full. So, in this (unlikely) case we will get regular insert errors on the next
+tries to insert to the page 'locked' by this very long-running transaction.
+
+Upgrade from 32-bit XID versions
+--------------------------------
+
+pg_upgrade doesn't change pages format itself. It is done lazily after.
+
+1. At first heap page read, tuples on a page are repacked to free 16 bytes
+at the end of a page, possibly freeing space from dead tuples.
+
+2A. 16 bytes of pd_special is added if there is a place for it
+
+2B. Page is converted to "Double XMAX" format if there is no place for
+pd_special
+
+3. If a page is in double XMAX format after its first read, and vacuum (or
+micro-vacuum at select query) could prune some tuples and free space for
+pd_special, prune_page will add pd_special and convert page from double XMAX
+to general 64-bit XID page format.
-- 
2.24.3 (Apple Git-128)


--cpok4wp6gsarlzvp--





^ permalink  raw  reply  [nested|flat] 265+ messages in thread

* [PATCH v6] README for 64bit xid
@ 2022-01-10 19:20 Pavel Borisov <[email protected]>
  0 siblings, 0 replies; 265+ messages in thread

From: Pavel Borisov @ 2022-01-10 19:20 UTC (permalink / raw)

Authors:
- Pavel Borisov <[email protected]>
- Maxim Orlov <[email protected]>
- Yura Sokolov <[email protected]> <[email protected]>
---
 src/backend/access/heap/README.XID64 | 128 +++++++++++++++++++++++++++
 1 file changed, 128 insertions(+)
 create mode 100644 src/backend/access/heap/README.XID64

diff --git a/src/backend/access/heap/README.XID64 b/src/backend/access/heap/README.XID64
new file mode 100644
index 00000000000..457ba9b9ef5
--- /dev/null
+++ b/src/backend/access/heap/README.XID64
@@ -0,0 +1,128 @@
+src/backend/access/heap/README.XID64
+
+64-bit Transaction ID's (XID)
+=============================
+
+A limited number (N = 2^32) of XID's required to do vacuum freeze to prevent
+wraparound every N/2 transactions. This causes performance degradation due
+to the need to exclusively lock tables while being vacuumed. In each
+wraparound cycle, SLRU buffers are also being cut.
+
+With 64-bit XID's wraparound is effectively postponed to a very distant
+future. Even in highly loaded systems that had 2^32 transactions per day
+it will take huge 2^31 days before the first enforced "vacuum to prevent
+wraparound"). Buffers cutting and routine vacuum are not enforced, and DBA
+can plan them independently at the time with the least system load and least
+critical for database performance. Also, it can be done less frequently
+(several times a year vs every several days) on systems with transaction rates
+similar to those mentioned above.
+
+On-disk tuple and page format
+-----------------------------
+
+On-disk tuple format remains unchanged. 32-bit t_xmin and t_xmax store the
+lower parts of 64-bit XMIN and XMAX values. Each heap page has additional
+64-bit pd_xid_base and pd_multi_base which are common for all tuples on a page.
+They are placed into a pd_special area - 16 bytes in the end of a heap page.
+Actual XMIN/XMAX for a tuple are calculated upon reading a tuple from a page
+as follows:
+
+XMIN = t_xmin + pd_xid_base. 					(1)
+XMAX = t_xmax + pd_xid_base/pd_multi_base.		(2)
+
+"Double XMAX" page format
+---------------------------------
+
+At first read of a heap page after pg_upgrade from 32-bit XID PostgreSQL
+version pd_special area with a size of 16 bytes should be added to a page.
+Though a page may not have space for this. Then it can be converted to a
+temporary format called "double XMAX".
+
+All tuples after pg-upgrade would necessarily have xmin = FrozenTransactionId.
+So we don't need tuple header t_xmin field and we reuse t_xmin to store higher
+32 bits of its XMAX.
+
+Double XMAX format is only for full pages that don't have 16 bytes for
+pd_special. So it neither has a�place for a single tuple. Insert and HOT update
+for double XMAX pages is impossible and not supported. We can only read or
+delete tuples from it.
+
+When we are able to prune page double XMAX it will be converted from it to
+general 64-bit XID page format with all operations on its tuples supported.
+
+In-memory tuple format
+----------------------
+
+In-memory tuple representation consists of two parts:
+- HeapTupleHeader from disk page (contains all heap tuple contents, not only
+header)
+- HeapTuple with additional in-memory fields
+
+HeapTuple for each tuple in memory stores t_xid_base/t_multi_base - a copies of
+page's pd_xid_base/pd_multi_base. With tuple's 32-bit t_xmin and t_xmax from
+HeapTupleHeader they are used to calculate actual 64-bit XMIN and XMAX:
+
+XMIN = t_xmin + t_xid_base. 					(3)
+XMAX = t_xmax + t_xid_base/t_multi_base.		(4)
+
+The downside of this is that we can not use tuple's XMIN and XMAX right away.
+We often need to re-read t_xmin and t_xmax - which could actually be pointers
+into a page in shared buffers and therefore they could be updated by any other
+backend.
+
+Update/delete with 64-bit XIDs and 32-bit t_xmin/t_xmax
+--------------------------------------------------------------
+
+When we try to delete/update a tuple, we check that XMAX for a page fits (2).
+I.e. that t_xmax will not be over MaxShortTransactionId relative to
+pd_xid_base/pd_multi_base of a its page.
+
+If the current XID doesn't fit a range
+(pd_xid_base, pd_xid_base + MaxShortTransactionId) (5):
+
+- heap_page_prepare_for_xid() will try to increase pd_xid_base/pd_multi_base on
+a page and update all t_xmin/t_xmax of the other tuples on the page to
+correspond new pd_xid_base/pd_multi_base.
+
+- If it was impossible, it will try to prune and freeze tuples on a page.
+
+- If this is unsuccessful it will throw an error. Normally this is very
+unlikely but if there is a very old living transaction with an age of around
+2^32 this can arise. Basically, this is a behavior similar to one during the
+vacuum to prevent wraparound when XID was 32-bit. Dba should take care and
+avoid very-long-living transactions with an age close to 2^32. So long-living
+transactions often they are most likely defunct.
+
+Insert with 64-bit XIDs and 32-bit t_xmin/t_xmax
+------------------------------------------------
+
+On insert we check if current XID fits a range (5). Otherwise:
+
+- heap_page_prepare_for_xid() will try to increase pd_xid_base for t_xmin will
+not be over MaxShortTransactionId.
+
+- If it is impossible, then it will try to prune and freeze tuples on a page.
+
+Known issue: if pd_xid_base could not be shifted to accommodate a tuple being
+inserted due to a very long-running transaction, we just throw an error. We
+neither try to insert a�tuple into another page nor mark the current page as
+full. So, in this (unlikely) case we will get regular insert errors on the next
+tries to insert to the page 'locked' by this very long-running transaction.
+
+Upgrade from 32-bit XID versions
+--------------------------------
+
+pg_upgrade doesn't change pages format itself. It is done lazily after.
+
+1. At first heap page read, tuples on a page are repacked to free 16 bytes
+at the end of a page, possibly freeing space from dead tuples.
+
+2A. 16 bytes of pd_special is added if there is a place for it
+
+2B. Page is converted to "Double XMAX" format if there is no place for
+pd_special
+
+3. If a page is in double XMAX format after its first read, and vacuum (or
+micro-vacuum at select query) could prune some tuples and free space for
+pd_special, prune_page will add pd_special and convert page from double XMAX
+to general 64-bit XID page format.
-- 
2.24.3 (Apple Git-128)


--cpok4wp6gsarlzvp--





^ permalink  raw  reply  [nested|flat] 265+ messages in thread

* [PATCH v6] README for 64bit xid
@ 2022-01-10 19:20 Pavel Borisov <[email protected]>
  0 siblings, 0 replies; 265+ messages in thread

From: Pavel Borisov @ 2022-01-10 19:20 UTC (permalink / raw)

Authors:
- Pavel Borisov <[email protected]>
- Maxim Orlov <[email protected]>
- Yura Sokolov <[email protected]> <[email protected]>
---
 src/backend/access/heap/README.XID64 | 128 +++++++++++++++++++++++++++
 1 file changed, 128 insertions(+)
 create mode 100644 src/backend/access/heap/README.XID64

diff --git a/src/backend/access/heap/README.XID64 b/src/backend/access/heap/README.XID64
new file mode 100644
index 00000000000..457ba9b9ef5
--- /dev/null
+++ b/src/backend/access/heap/README.XID64
@@ -0,0 +1,128 @@
+src/backend/access/heap/README.XID64
+
+64-bit Transaction ID's (XID)
+=============================
+
+A limited number (N = 2^32) of XID's required to do vacuum freeze to prevent
+wraparound every N/2 transactions. This causes performance degradation due
+to the need to exclusively lock tables while being vacuumed. In each
+wraparound cycle, SLRU buffers are also being cut.
+
+With 64-bit XID's wraparound is effectively postponed to a very distant
+future. Even in highly loaded systems that had 2^32 transactions per day
+it will take huge 2^31 days before the first enforced "vacuum to prevent
+wraparound"). Buffers cutting and routine vacuum are not enforced, and DBA
+can plan them independently at the time with the least system load and least
+critical for database performance. Also, it can be done less frequently
+(several times a year vs every several days) on systems with transaction rates
+similar to those mentioned above.
+
+On-disk tuple and page format
+-----------------------------
+
+On-disk tuple format remains unchanged. 32-bit t_xmin and t_xmax store the
+lower parts of 64-bit XMIN and XMAX values. Each heap page has additional
+64-bit pd_xid_base and pd_multi_base which are common for all tuples on a page.
+They are placed into a pd_special area - 16 bytes in the end of a heap page.
+Actual XMIN/XMAX for a tuple are calculated upon reading a tuple from a page
+as follows:
+
+XMIN = t_xmin + pd_xid_base. 					(1)
+XMAX = t_xmax + pd_xid_base/pd_multi_base.		(2)
+
+"Double XMAX" page format
+---------------------------------
+
+At first read of a heap page after pg_upgrade from 32-bit XID PostgreSQL
+version pd_special area with a size of 16 bytes should be added to a page.
+Though a page may not have space for this. Then it can be converted to a
+temporary format called "double XMAX".
+
+All tuples after pg-upgrade would necessarily have xmin = FrozenTransactionId.
+So we don't need tuple header t_xmin field and we reuse t_xmin to store higher
+32 bits of its XMAX.
+
+Double XMAX format is only for full pages that don't have 16 bytes for
+pd_special. So it neither has a�place for a single tuple. Insert and HOT update
+for double XMAX pages is impossible and not supported. We can only read or
+delete tuples from it.
+
+When we are able to prune page double XMAX it will be converted from it to
+general 64-bit XID page format with all operations on its tuples supported.
+
+In-memory tuple format
+----------------------
+
+In-memory tuple representation consists of two parts:
+- HeapTupleHeader from disk page (contains all heap tuple contents, not only
+header)
+- HeapTuple with additional in-memory fields
+
+HeapTuple for each tuple in memory stores t_xid_base/t_multi_base - a copies of
+page's pd_xid_base/pd_multi_base. With tuple's 32-bit t_xmin and t_xmax from
+HeapTupleHeader they are used to calculate actual 64-bit XMIN and XMAX:
+
+XMIN = t_xmin + t_xid_base. 					(3)
+XMAX = t_xmax + t_xid_base/t_multi_base.		(4)
+
+The downside of this is that we can not use tuple's XMIN and XMAX right away.
+We often need to re-read t_xmin and t_xmax - which could actually be pointers
+into a page in shared buffers and therefore they could be updated by any other
+backend.
+
+Update/delete with 64-bit XIDs and 32-bit t_xmin/t_xmax
+--------------------------------------------------------------
+
+When we try to delete/update a tuple, we check that XMAX for a page fits (2).
+I.e. that t_xmax will not be over MaxShortTransactionId relative to
+pd_xid_base/pd_multi_base of a its page.
+
+If the current XID doesn't fit a range
+(pd_xid_base, pd_xid_base + MaxShortTransactionId) (5):
+
+- heap_page_prepare_for_xid() will try to increase pd_xid_base/pd_multi_base on
+a page and update all t_xmin/t_xmax of the other tuples on the page to
+correspond new pd_xid_base/pd_multi_base.
+
+- If it was impossible, it will try to prune and freeze tuples on a page.
+
+- If this is unsuccessful it will throw an error. Normally this is very
+unlikely but if there is a very old living transaction with an age of around
+2^32 this can arise. Basically, this is a behavior similar to one during the
+vacuum to prevent wraparound when XID was 32-bit. Dba should take care and
+avoid very-long-living transactions with an age close to 2^32. So long-living
+transactions often they are most likely defunct.
+
+Insert with 64-bit XIDs and 32-bit t_xmin/t_xmax
+------------------------------------------------
+
+On insert we check if current XID fits a range (5). Otherwise:
+
+- heap_page_prepare_for_xid() will try to increase pd_xid_base for t_xmin will
+not be over MaxShortTransactionId.
+
+- If it is impossible, then it will try to prune and freeze tuples on a page.
+
+Known issue: if pd_xid_base could not be shifted to accommodate a tuple being
+inserted due to a very long-running transaction, we just throw an error. We
+neither try to insert a�tuple into another page nor mark the current page as
+full. So, in this (unlikely) case we will get regular insert errors on the next
+tries to insert to the page 'locked' by this very long-running transaction.
+
+Upgrade from 32-bit XID versions
+--------------------------------
+
+pg_upgrade doesn't change pages format itself. It is done lazily after.
+
+1. At first heap page read, tuples on a page are repacked to free 16 bytes
+at the end of a page, possibly freeing space from dead tuples.
+
+2A. 16 bytes of pd_special is added if there is a place for it
+
+2B. Page is converted to "Double XMAX" format if there is no place for
+pd_special
+
+3. If a page is in double XMAX format after its first read, and vacuum (or
+micro-vacuum at select query) could prune some tuples and free space for
+pd_special, prune_page will add pd_special and convert page from double XMAX
+to general 64-bit XID page format.
-- 
2.24.3 (Apple Git-128)


--cpok4wp6gsarlzvp--





^ permalink  raw  reply  [nested|flat] 265+ messages in thread

* [PATCH v6] README for 64bit xid
@ 2022-01-10 19:20 Pavel Borisov <[email protected]>
  0 siblings, 0 replies; 265+ messages in thread

From: Pavel Borisov @ 2022-01-10 19:20 UTC (permalink / raw)

Authors:
- Pavel Borisov <[email protected]>
- Maxim Orlov <[email protected]>
- Yura Sokolov <[email protected]> <[email protected]>
---
 src/backend/access/heap/README.XID64 | 128 +++++++++++++++++++++++++++
 1 file changed, 128 insertions(+)
 create mode 100644 src/backend/access/heap/README.XID64

diff --git a/src/backend/access/heap/README.XID64 b/src/backend/access/heap/README.XID64
new file mode 100644
index 00000000000..457ba9b9ef5
--- /dev/null
+++ b/src/backend/access/heap/README.XID64
@@ -0,0 +1,128 @@
+src/backend/access/heap/README.XID64
+
+64-bit Transaction ID's (XID)
+=============================
+
+A limited number (N = 2^32) of XID's required to do vacuum freeze to prevent
+wraparound every N/2 transactions. This causes performance degradation due
+to the need to exclusively lock tables while being vacuumed. In each
+wraparound cycle, SLRU buffers are also being cut.
+
+With 64-bit XID's wraparound is effectively postponed to a very distant
+future. Even in highly loaded systems that had 2^32 transactions per day
+it will take huge 2^31 days before the first enforced "vacuum to prevent
+wraparound"). Buffers cutting and routine vacuum are not enforced, and DBA
+can plan them independently at the time with the least system load and least
+critical for database performance. Also, it can be done less frequently
+(several times a year vs every several days) on systems with transaction rates
+similar to those mentioned above.
+
+On-disk tuple and page format
+-----------------------------
+
+On-disk tuple format remains unchanged. 32-bit t_xmin and t_xmax store the
+lower parts of 64-bit XMIN and XMAX values. Each heap page has additional
+64-bit pd_xid_base and pd_multi_base which are common for all tuples on a page.
+They are placed into a pd_special area - 16 bytes in the end of a heap page.
+Actual XMIN/XMAX for a tuple are calculated upon reading a tuple from a page
+as follows:
+
+XMIN = t_xmin + pd_xid_base. 					(1)
+XMAX = t_xmax + pd_xid_base/pd_multi_base.		(2)
+
+"Double XMAX" page format
+---------------------------------
+
+At first read of a heap page after pg_upgrade from 32-bit XID PostgreSQL
+version pd_special area with a size of 16 bytes should be added to a page.
+Though a page may not have space for this. Then it can be converted to a
+temporary format called "double XMAX".
+
+All tuples after pg-upgrade would necessarily have xmin = FrozenTransactionId.
+So we don't need tuple header t_xmin field and we reuse t_xmin to store higher
+32 bits of its XMAX.
+
+Double XMAX format is only for full pages that don't have 16 bytes for
+pd_special. So it neither has a�place for a single tuple. Insert and HOT update
+for double XMAX pages is impossible and not supported. We can only read or
+delete tuples from it.
+
+When we are able to prune page double XMAX it will be converted from it to
+general 64-bit XID page format with all operations on its tuples supported.
+
+In-memory tuple format
+----------------------
+
+In-memory tuple representation consists of two parts:
+- HeapTupleHeader from disk page (contains all heap tuple contents, not only
+header)
+- HeapTuple with additional in-memory fields
+
+HeapTuple for each tuple in memory stores t_xid_base/t_multi_base - a copies of
+page's pd_xid_base/pd_multi_base. With tuple's 32-bit t_xmin and t_xmax from
+HeapTupleHeader they are used to calculate actual 64-bit XMIN and XMAX:
+
+XMIN = t_xmin + t_xid_base. 					(3)
+XMAX = t_xmax + t_xid_base/t_multi_base.		(4)
+
+The downside of this is that we can not use tuple's XMIN and XMAX right away.
+We often need to re-read t_xmin and t_xmax - which could actually be pointers
+into a page in shared buffers and therefore they could be updated by any other
+backend.
+
+Update/delete with 64-bit XIDs and 32-bit t_xmin/t_xmax
+--------------------------------------------------------------
+
+When we try to delete/update a tuple, we check that XMAX for a page fits (2).
+I.e. that t_xmax will not be over MaxShortTransactionId relative to
+pd_xid_base/pd_multi_base of a its page.
+
+If the current XID doesn't fit a range
+(pd_xid_base, pd_xid_base + MaxShortTransactionId) (5):
+
+- heap_page_prepare_for_xid() will try to increase pd_xid_base/pd_multi_base on
+a page and update all t_xmin/t_xmax of the other tuples on the page to
+correspond new pd_xid_base/pd_multi_base.
+
+- If it was impossible, it will try to prune and freeze tuples on a page.
+
+- If this is unsuccessful it will throw an error. Normally this is very
+unlikely but if there is a very old living transaction with an age of around
+2^32 this can arise. Basically, this is a behavior similar to one during the
+vacuum to prevent wraparound when XID was 32-bit. Dba should take care and
+avoid very-long-living transactions with an age close to 2^32. So long-living
+transactions often they are most likely defunct.
+
+Insert with 64-bit XIDs and 32-bit t_xmin/t_xmax
+------------------------------------------------
+
+On insert we check if current XID fits a range (5). Otherwise:
+
+- heap_page_prepare_for_xid() will try to increase pd_xid_base for t_xmin will
+not be over MaxShortTransactionId.
+
+- If it is impossible, then it will try to prune and freeze tuples on a page.
+
+Known issue: if pd_xid_base could not be shifted to accommodate a tuple being
+inserted due to a very long-running transaction, we just throw an error. We
+neither try to insert a�tuple into another page nor mark the current page as
+full. So, in this (unlikely) case we will get regular insert errors on the next
+tries to insert to the page 'locked' by this very long-running transaction.
+
+Upgrade from 32-bit XID versions
+--------------------------------
+
+pg_upgrade doesn't change pages format itself. It is done lazily after.
+
+1. At first heap page read, tuples on a page are repacked to free 16 bytes
+at the end of a page, possibly freeing space from dead tuples.
+
+2A. 16 bytes of pd_special is added if there is a place for it
+
+2B. Page is converted to "Double XMAX" format if there is no place for
+pd_special
+
+3. If a page is in double XMAX format after its first read, and vacuum (or
+micro-vacuum at select query) could prune some tuples and free space for
+pd_special, prune_page will add pd_special and convert page from double XMAX
+to general 64-bit XID page format.
-- 
2.24.3 (Apple Git-128)


--cpok4wp6gsarlzvp--





^ permalink  raw  reply  [nested|flat] 265+ messages in thread

* [PATCH v6] README for 64bit xid
@ 2022-01-10 19:20 Pavel Borisov <[email protected]>
  0 siblings, 0 replies; 265+ messages in thread

From: Pavel Borisov @ 2022-01-10 19:20 UTC (permalink / raw)

Authors:
- Pavel Borisov <[email protected]>
- Maxim Orlov <[email protected]>
- Yura Sokolov <[email protected]> <[email protected]>
---
 src/backend/access/heap/README.XID64 | 128 +++++++++++++++++++++++++++
 1 file changed, 128 insertions(+)
 create mode 100644 src/backend/access/heap/README.XID64

diff --git a/src/backend/access/heap/README.XID64 b/src/backend/access/heap/README.XID64
new file mode 100644
index 00000000000..457ba9b9ef5
--- /dev/null
+++ b/src/backend/access/heap/README.XID64
@@ -0,0 +1,128 @@
+src/backend/access/heap/README.XID64
+
+64-bit Transaction ID's (XID)
+=============================
+
+A limited number (N = 2^32) of XID's required to do vacuum freeze to prevent
+wraparound every N/2 transactions. This causes performance degradation due
+to the need to exclusively lock tables while being vacuumed. In each
+wraparound cycle, SLRU buffers are also being cut.
+
+With 64-bit XID's wraparound is effectively postponed to a very distant
+future. Even in highly loaded systems that had 2^32 transactions per day
+it will take huge 2^31 days before the first enforced "vacuum to prevent
+wraparound"). Buffers cutting and routine vacuum are not enforced, and DBA
+can plan them independently at the time with the least system load and least
+critical for database performance. Also, it can be done less frequently
+(several times a year vs every several days) on systems with transaction rates
+similar to those mentioned above.
+
+On-disk tuple and page format
+-----------------------------
+
+On-disk tuple format remains unchanged. 32-bit t_xmin and t_xmax store the
+lower parts of 64-bit XMIN and XMAX values. Each heap page has additional
+64-bit pd_xid_base and pd_multi_base which are common for all tuples on a page.
+They are placed into a pd_special area - 16 bytes in the end of a heap page.
+Actual XMIN/XMAX for a tuple are calculated upon reading a tuple from a page
+as follows:
+
+XMIN = t_xmin + pd_xid_base. 					(1)
+XMAX = t_xmax + pd_xid_base/pd_multi_base.		(2)
+
+"Double XMAX" page format
+---------------------------------
+
+At first read of a heap page after pg_upgrade from 32-bit XID PostgreSQL
+version pd_special area with a size of 16 bytes should be added to a page.
+Though a page may not have space for this. Then it can be converted to a
+temporary format called "double XMAX".
+
+All tuples after pg-upgrade would necessarily have xmin = FrozenTransactionId.
+So we don't need tuple header t_xmin field and we reuse t_xmin to store higher
+32 bits of its XMAX.
+
+Double XMAX format is only for full pages that don't have 16 bytes for
+pd_special. So it neither has a�place for a single tuple. Insert and HOT update
+for double XMAX pages is impossible and not supported. We can only read or
+delete tuples from it.
+
+When we are able to prune page double XMAX it will be converted from it to
+general 64-bit XID page format with all operations on its tuples supported.
+
+In-memory tuple format
+----------------------
+
+In-memory tuple representation consists of two parts:
+- HeapTupleHeader from disk page (contains all heap tuple contents, not only
+header)
+- HeapTuple with additional in-memory fields
+
+HeapTuple for each tuple in memory stores t_xid_base/t_multi_base - a copies of
+page's pd_xid_base/pd_multi_base. With tuple's 32-bit t_xmin and t_xmax from
+HeapTupleHeader they are used to calculate actual 64-bit XMIN and XMAX:
+
+XMIN = t_xmin + t_xid_base. 					(3)
+XMAX = t_xmax + t_xid_base/t_multi_base.		(4)
+
+The downside of this is that we can not use tuple's XMIN and XMAX right away.
+We often need to re-read t_xmin and t_xmax - which could actually be pointers
+into a page in shared buffers and therefore they could be updated by any other
+backend.
+
+Update/delete with 64-bit XIDs and 32-bit t_xmin/t_xmax
+--------------------------------------------------------------
+
+When we try to delete/update a tuple, we check that XMAX for a page fits (2).
+I.e. that t_xmax will not be over MaxShortTransactionId relative to
+pd_xid_base/pd_multi_base of a its page.
+
+If the current XID doesn't fit a range
+(pd_xid_base, pd_xid_base + MaxShortTransactionId) (5):
+
+- heap_page_prepare_for_xid() will try to increase pd_xid_base/pd_multi_base on
+a page and update all t_xmin/t_xmax of the other tuples on the page to
+correspond new pd_xid_base/pd_multi_base.
+
+- If it was impossible, it will try to prune and freeze tuples on a page.
+
+- If this is unsuccessful it will throw an error. Normally this is very
+unlikely but if there is a very old living transaction with an age of around
+2^32 this can arise. Basically, this is a behavior similar to one during the
+vacuum to prevent wraparound when XID was 32-bit. Dba should take care and
+avoid very-long-living transactions with an age close to 2^32. So long-living
+transactions often they are most likely defunct.
+
+Insert with 64-bit XIDs and 32-bit t_xmin/t_xmax
+------------------------------------------------
+
+On insert we check if current XID fits a range (5). Otherwise:
+
+- heap_page_prepare_for_xid() will try to increase pd_xid_base for t_xmin will
+not be over MaxShortTransactionId.
+
+- If it is impossible, then it will try to prune and freeze tuples on a page.
+
+Known issue: if pd_xid_base could not be shifted to accommodate a tuple being
+inserted due to a very long-running transaction, we just throw an error. We
+neither try to insert a�tuple into another page nor mark the current page as
+full. So, in this (unlikely) case we will get regular insert errors on the next
+tries to insert to the page 'locked' by this very long-running transaction.
+
+Upgrade from 32-bit XID versions
+--------------------------------
+
+pg_upgrade doesn't change pages format itself. It is done lazily after.
+
+1. At first heap page read, tuples on a page are repacked to free 16 bytes
+at the end of a page, possibly freeing space from dead tuples.
+
+2A. 16 bytes of pd_special is added if there is a place for it
+
+2B. Page is converted to "Double XMAX" format if there is no place for
+pd_special
+
+3. If a page is in double XMAX format after its first read, and vacuum (or
+micro-vacuum at select query) could prune some tuples and free space for
+pd_special, prune_page will add pd_special and convert page from double XMAX
+to general 64-bit XID page format.
-- 
2.24.3 (Apple Git-128)


--cpok4wp6gsarlzvp--





^ permalink  raw  reply  [nested|flat] 265+ messages in thread

* [PATCH v6] README for 64bit xid
@ 2022-01-10 19:20 Pavel Borisov <[email protected]>
  0 siblings, 0 replies; 265+ messages in thread

From: Pavel Borisov @ 2022-01-10 19:20 UTC (permalink / raw)

Authors:
- Pavel Borisov <[email protected]>
- Maxim Orlov <[email protected]>
- Yura Sokolov <[email protected]> <[email protected]>
---
 src/backend/access/heap/README.XID64 | 128 +++++++++++++++++++++++++++
 1 file changed, 128 insertions(+)
 create mode 100644 src/backend/access/heap/README.XID64

diff --git a/src/backend/access/heap/README.XID64 b/src/backend/access/heap/README.XID64
new file mode 100644
index 00000000000..457ba9b9ef5
--- /dev/null
+++ b/src/backend/access/heap/README.XID64
@@ -0,0 +1,128 @@
+src/backend/access/heap/README.XID64
+
+64-bit Transaction ID's (XID)
+=============================
+
+A limited number (N = 2^32) of XID's required to do vacuum freeze to prevent
+wraparound every N/2 transactions. This causes performance degradation due
+to the need to exclusively lock tables while being vacuumed. In each
+wraparound cycle, SLRU buffers are also being cut.
+
+With 64-bit XID's wraparound is effectively postponed to a very distant
+future. Even in highly loaded systems that had 2^32 transactions per day
+it will take huge 2^31 days before the first enforced "vacuum to prevent
+wraparound"). Buffers cutting and routine vacuum are not enforced, and DBA
+can plan them independently at the time with the least system load and least
+critical for database performance. Also, it can be done less frequently
+(several times a year vs every several days) on systems with transaction rates
+similar to those mentioned above.
+
+On-disk tuple and page format
+-----------------------------
+
+On-disk tuple format remains unchanged. 32-bit t_xmin and t_xmax store the
+lower parts of 64-bit XMIN and XMAX values. Each heap page has additional
+64-bit pd_xid_base and pd_multi_base which are common for all tuples on a page.
+They are placed into a pd_special area - 16 bytes in the end of a heap page.
+Actual XMIN/XMAX for a tuple are calculated upon reading a tuple from a page
+as follows:
+
+XMIN = t_xmin + pd_xid_base. 					(1)
+XMAX = t_xmax + pd_xid_base/pd_multi_base.		(2)
+
+"Double XMAX" page format
+---------------------------------
+
+At first read of a heap page after pg_upgrade from 32-bit XID PostgreSQL
+version pd_special area with a size of 16 bytes should be added to a page.
+Though a page may not have space for this. Then it can be converted to a
+temporary format called "double XMAX".
+
+All tuples after pg-upgrade would necessarily have xmin = FrozenTransactionId.
+So we don't need tuple header t_xmin field and we reuse t_xmin to store higher
+32 bits of its XMAX.
+
+Double XMAX format is only for full pages that don't have 16 bytes for
+pd_special. So it neither has a�place for a single tuple. Insert and HOT update
+for double XMAX pages is impossible and not supported. We can only read or
+delete tuples from it.
+
+When we are able to prune page double XMAX it will be converted from it to
+general 64-bit XID page format with all operations on its tuples supported.
+
+In-memory tuple format
+----------------------
+
+In-memory tuple representation consists of two parts:
+- HeapTupleHeader from disk page (contains all heap tuple contents, not only
+header)
+- HeapTuple with additional in-memory fields
+
+HeapTuple for each tuple in memory stores t_xid_base/t_multi_base - a copies of
+page's pd_xid_base/pd_multi_base. With tuple's 32-bit t_xmin and t_xmax from
+HeapTupleHeader they are used to calculate actual 64-bit XMIN and XMAX:
+
+XMIN = t_xmin + t_xid_base. 					(3)
+XMAX = t_xmax + t_xid_base/t_multi_base.		(4)
+
+The downside of this is that we can not use tuple's XMIN and XMAX right away.
+We often need to re-read t_xmin and t_xmax - which could actually be pointers
+into a page in shared buffers and therefore they could be updated by any other
+backend.
+
+Update/delete with 64-bit XIDs and 32-bit t_xmin/t_xmax
+--------------------------------------------------------------
+
+When we try to delete/update a tuple, we check that XMAX for a page fits (2).
+I.e. that t_xmax will not be over MaxShortTransactionId relative to
+pd_xid_base/pd_multi_base of a its page.
+
+If the current XID doesn't fit a range
+(pd_xid_base, pd_xid_base + MaxShortTransactionId) (5):
+
+- heap_page_prepare_for_xid() will try to increase pd_xid_base/pd_multi_base on
+a page and update all t_xmin/t_xmax of the other tuples on the page to
+correspond new pd_xid_base/pd_multi_base.
+
+- If it was impossible, it will try to prune and freeze tuples on a page.
+
+- If this is unsuccessful it will throw an error. Normally this is very
+unlikely but if there is a very old living transaction with an age of around
+2^32 this can arise. Basically, this is a behavior similar to one during the
+vacuum to prevent wraparound when XID was 32-bit. Dba should take care and
+avoid very-long-living transactions with an age close to 2^32. So long-living
+transactions often they are most likely defunct.
+
+Insert with 64-bit XIDs and 32-bit t_xmin/t_xmax
+------------------------------------------------
+
+On insert we check if current XID fits a range (5). Otherwise:
+
+- heap_page_prepare_for_xid() will try to increase pd_xid_base for t_xmin will
+not be over MaxShortTransactionId.
+
+- If it is impossible, then it will try to prune and freeze tuples on a page.
+
+Known issue: if pd_xid_base could not be shifted to accommodate a tuple being
+inserted due to a very long-running transaction, we just throw an error. We
+neither try to insert a�tuple into another page nor mark the current page as
+full. So, in this (unlikely) case we will get regular insert errors on the next
+tries to insert to the page 'locked' by this very long-running transaction.
+
+Upgrade from 32-bit XID versions
+--------------------------------
+
+pg_upgrade doesn't change pages format itself. It is done lazily after.
+
+1. At first heap page read, tuples on a page are repacked to free 16 bytes
+at the end of a page, possibly freeing space from dead tuples.
+
+2A. 16 bytes of pd_special is added if there is a place for it
+
+2B. Page is converted to "Double XMAX" format if there is no place for
+pd_special
+
+3. If a page is in double XMAX format after its first read, and vacuum (or
+micro-vacuum at select query) could prune some tuples and free space for
+pd_special, prune_page will add pd_special and convert page from double XMAX
+to general 64-bit XID page format.
-- 
2.24.3 (Apple Git-128)


--cpok4wp6gsarlzvp--





^ permalink  raw  reply  [nested|flat] 265+ messages in thread

* [PATCH v6] README for 64bit xid
@ 2022-01-10 19:20 Pavel Borisov <[email protected]>
  0 siblings, 0 replies; 265+ messages in thread

From: Pavel Borisov @ 2022-01-10 19:20 UTC (permalink / raw)

Authors:
- Pavel Borisov <[email protected]>
- Maxim Orlov <[email protected]>
- Yura Sokolov <[email protected]> <[email protected]>
---
 src/backend/access/heap/README.XID64 | 128 +++++++++++++++++++++++++++
 1 file changed, 128 insertions(+)
 create mode 100644 src/backend/access/heap/README.XID64

diff --git a/src/backend/access/heap/README.XID64 b/src/backend/access/heap/README.XID64
new file mode 100644
index 00000000000..457ba9b9ef5
--- /dev/null
+++ b/src/backend/access/heap/README.XID64
@@ -0,0 +1,128 @@
+src/backend/access/heap/README.XID64
+
+64-bit Transaction ID's (XID)
+=============================
+
+A limited number (N = 2^32) of XID's required to do vacuum freeze to prevent
+wraparound every N/2 transactions. This causes performance degradation due
+to the need to exclusively lock tables while being vacuumed. In each
+wraparound cycle, SLRU buffers are also being cut.
+
+With 64-bit XID's wraparound is effectively postponed to a very distant
+future. Even in highly loaded systems that had 2^32 transactions per day
+it will take huge 2^31 days before the first enforced "vacuum to prevent
+wraparound"). Buffers cutting and routine vacuum are not enforced, and DBA
+can plan them independently at the time with the least system load and least
+critical for database performance. Also, it can be done less frequently
+(several times a year vs every several days) on systems with transaction rates
+similar to those mentioned above.
+
+On-disk tuple and page format
+-----------------------------
+
+On-disk tuple format remains unchanged. 32-bit t_xmin and t_xmax store the
+lower parts of 64-bit XMIN and XMAX values. Each heap page has additional
+64-bit pd_xid_base and pd_multi_base which are common for all tuples on a page.
+They are placed into a pd_special area - 16 bytes in the end of a heap page.
+Actual XMIN/XMAX for a tuple are calculated upon reading a tuple from a page
+as follows:
+
+XMIN = t_xmin + pd_xid_base. 					(1)
+XMAX = t_xmax + pd_xid_base/pd_multi_base.		(2)
+
+"Double XMAX" page format
+---------------------------------
+
+At first read of a heap page after pg_upgrade from 32-bit XID PostgreSQL
+version pd_special area with a size of 16 bytes should be added to a page.
+Though a page may not have space for this. Then it can be converted to a
+temporary format called "double XMAX".
+
+All tuples after pg-upgrade would necessarily have xmin = FrozenTransactionId.
+So we don't need tuple header t_xmin field and we reuse t_xmin to store higher
+32 bits of its XMAX.
+
+Double XMAX format is only for full pages that don't have 16 bytes for
+pd_special. So it neither has a�place for a single tuple. Insert and HOT update
+for double XMAX pages is impossible and not supported. We can only read or
+delete tuples from it.
+
+When we are able to prune page double XMAX it will be converted from it to
+general 64-bit XID page format with all operations on its tuples supported.
+
+In-memory tuple format
+----------------------
+
+In-memory tuple representation consists of two parts:
+- HeapTupleHeader from disk page (contains all heap tuple contents, not only
+header)
+- HeapTuple with additional in-memory fields
+
+HeapTuple for each tuple in memory stores t_xid_base/t_multi_base - a copies of
+page's pd_xid_base/pd_multi_base. With tuple's 32-bit t_xmin and t_xmax from
+HeapTupleHeader they are used to calculate actual 64-bit XMIN and XMAX:
+
+XMIN = t_xmin + t_xid_base. 					(3)
+XMAX = t_xmax + t_xid_base/t_multi_base.		(4)
+
+The downside of this is that we can not use tuple's XMIN and XMAX right away.
+We often need to re-read t_xmin and t_xmax - which could actually be pointers
+into a page in shared buffers and therefore they could be updated by any other
+backend.
+
+Update/delete with 64-bit XIDs and 32-bit t_xmin/t_xmax
+--------------------------------------------------------------
+
+When we try to delete/update a tuple, we check that XMAX for a page fits (2).
+I.e. that t_xmax will not be over MaxShortTransactionId relative to
+pd_xid_base/pd_multi_base of a its page.
+
+If the current XID doesn't fit a range
+(pd_xid_base, pd_xid_base + MaxShortTransactionId) (5):
+
+- heap_page_prepare_for_xid() will try to increase pd_xid_base/pd_multi_base on
+a page and update all t_xmin/t_xmax of the other tuples on the page to
+correspond new pd_xid_base/pd_multi_base.
+
+- If it was impossible, it will try to prune and freeze tuples on a page.
+
+- If this is unsuccessful it will throw an error. Normally this is very
+unlikely but if there is a very old living transaction with an age of around
+2^32 this can arise. Basically, this is a behavior similar to one during the
+vacuum to prevent wraparound when XID was 32-bit. Dba should take care and
+avoid very-long-living transactions with an age close to 2^32. So long-living
+transactions often they are most likely defunct.
+
+Insert with 64-bit XIDs and 32-bit t_xmin/t_xmax
+------------------------------------------------
+
+On insert we check if current XID fits a range (5). Otherwise:
+
+- heap_page_prepare_for_xid() will try to increase pd_xid_base for t_xmin will
+not be over MaxShortTransactionId.
+
+- If it is impossible, then it will try to prune and freeze tuples on a page.
+
+Known issue: if pd_xid_base could not be shifted to accommodate a tuple being
+inserted due to a very long-running transaction, we just throw an error. We
+neither try to insert a�tuple into another page nor mark the current page as
+full. So, in this (unlikely) case we will get regular insert errors on the next
+tries to insert to the page 'locked' by this very long-running transaction.
+
+Upgrade from 32-bit XID versions
+--------------------------------
+
+pg_upgrade doesn't change pages format itself. It is done lazily after.
+
+1. At first heap page read, tuples on a page are repacked to free 16 bytes
+at the end of a page, possibly freeing space from dead tuples.
+
+2A. 16 bytes of pd_special is added if there is a place for it
+
+2B. Page is converted to "Double XMAX" format if there is no place for
+pd_special
+
+3. If a page is in double XMAX format after its first read, and vacuum (or
+micro-vacuum at select query) could prune some tuples and free space for
+pd_special, prune_page will add pd_special and convert page from double XMAX
+to general 64-bit XID page format.
-- 
2.24.3 (Apple Git-128)


--cpok4wp6gsarlzvp--





^ permalink  raw  reply  [nested|flat] 265+ messages in thread

* [PATCH v6] README for 64bit xid
@ 2022-01-10 19:20 Pavel Borisov <[email protected]>
  0 siblings, 0 replies; 265+ messages in thread

From: Pavel Borisov @ 2022-01-10 19:20 UTC (permalink / raw)

Authors:
- Pavel Borisov <[email protected]>
- Maxim Orlov <[email protected]>
- Yura Sokolov <[email protected]> <[email protected]>
---
 src/backend/access/heap/README.XID64 | 128 +++++++++++++++++++++++++++
 1 file changed, 128 insertions(+)
 create mode 100644 src/backend/access/heap/README.XID64

diff --git a/src/backend/access/heap/README.XID64 b/src/backend/access/heap/README.XID64
new file mode 100644
index 00000000000..457ba9b9ef5
--- /dev/null
+++ b/src/backend/access/heap/README.XID64
@@ -0,0 +1,128 @@
+src/backend/access/heap/README.XID64
+
+64-bit Transaction ID's (XID)
+=============================
+
+A limited number (N = 2^32) of XID's required to do vacuum freeze to prevent
+wraparound every N/2 transactions. This causes performance degradation due
+to the need to exclusively lock tables while being vacuumed. In each
+wraparound cycle, SLRU buffers are also being cut.
+
+With 64-bit XID's wraparound is effectively postponed to a very distant
+future. Even in highly loaded systems that had 2^32 transactions per day
+it will take huge 2^31 days before the first enforced "vacuum to prevent
+wraparound"). Buffers cutting and routine vacuum are not enforced, and DBA
+can plan them independently at the time with the least system load and least
+critical for database performance. Also, it can be done less frequently
+(several times a year vs every several days) on systems with transaction rates
+similar to those mentioned above.
+
+On-disk tuple and page format
+-----------------------------
+
+On-disk tuple format remains unchanged. 32-bit t_xmin and t_xmax store the
+lower parts of 64-bit XMIN and XMAX values. Each heap page has additional
+64-bit pd_xid_base and pd_multi_base which are common for all tuples on a page.
+They are placed into a pd_special area - 16 bytes in the end of a heap page.
+Actual XMIN/XMAX for a tuple are calculated upon reading a tuple from a page
+as follows:
+
+XMIN = t_xmin + pd_xid_base. 					(1)
+XMAX = t_xmax + pd_xid_base/pd_multi_base.		(2)
+
+"Double XMAX" page format
+---------------------------------
+
+At first read of a heap page after pg_upgrade from 32-bit XID PostgreSQL
+version pd_special area with a size of 16 bytes should be added to a page.
+Though a page may not have space for this. Then it can be converted to a
+temporary format called "double XMAX".
+
+All tuples after pg-upgrade would necessarily have xmin = FrozenTransactionId.
+So we don't need tuple header t_xmin field and we reuse t_xmin to store higher
+32 bits of its XMAX.
+
+Double XMAX format is only for full pages that don't have 16 bytes for
+pd_special. So it neither has a�place for a single tuple. Insert and HOT update
+for double XMAX pages is impossible and not supported. We can only read or
+delete tuples from it.
+
+When we are able to prune page double XMAX it will be converted from it to
+general 64-bit XID page format with all operations on its tuples supported.
+
+In-memory tuple format
+----------------------
+
+In-memory tuple representation consists of two parts:
+- HeapTupleHeader from disk page (contains all heap tuple contents, not only
+header)
+- HeapTuple with additional in-memory fields
+
+HeapTuple for each tuple in memory stores t_xid_base/t_multi_base - a copies of
+page's pd_xid_base/pd_multi_base. With tuple's 32-bit t_xmin and t_xmax from
+HeapTupleHeader they are used to calculate actual 64-bit XMIN and XMAX:
+
+XMIN = t_xmin + t_xid_base. 					(3)
+XMAX = t_xmax + t_xid_base/t_multi_base.		(4)
+
+The downside of this is that we can not use tuple's XMIN and XMAX right away.
+We often need to re-read t_xmin and t_xmax - which could actually be pointers
+into a page in shared buffers and therefore they could be updated by any other
+backend.
+
+Update/delete with 64-bit XIDs and 32-bit t_xmin/t_xmax
+--------------------------------------------------------------
+
+When we try to delete/update a tuple, we check that XMAX for a page fits (2).
+I.e. that t_xmax will not be over MaxShortTransactionId relative to
+pd_xid_base/pd_multi_base of a its page.
+
+If the current XID doesn't fit a range
+(pd_xid_base, pd_xid_base + MaxShortTransactionId) (5):
+
+- heap_page_prepare_for_xid() will try to increase pd_xid_base/pd_multi_base on
+a page and update all t_xmin/t_xmax of the other tuples on the page to
+correspond new pd_xid_base/pd_multi_base.
+
+- If it was impossible, it will try to prune and freeze tuples on a page.
+
+- If this is unsuccessful it will throw an error. Normally this is very
+unlikely but if there is a very old living transaction with an age of around
+2^32 this can arise. Basically, this is a behavior similar to one during the
+vacuum to prevent wraparound when XID was 32-bit. Dba should take care and
+avoid very-long-living transactions with an age close to 2^32. So long-living
+transactions often they are most likely defunct.
+
+Insert with 64-bit XIDs and 32-bit t_xmin/t_xmax
+------------------------------------------------
+
+On insert we check if current XID fits a range (5). Otherwise:
+
+- heap_page_prepare_for_xid() will try to increase pd_xid_base for t_xmin will
+not be over MaxShortTransactionId.
+
+- If it is impossible, then it will try to prune and freeze tuples on a page.
+
+Known issue: if pd_xid_base could not be shifted to accommodate a tuple being
+inserted due to a very long-running transaction, we just throw an error. We
+neither try to insert a�tuple into another page nor mark the current page as
+full. So, in this (unlikely) case we will get regular insert errors on the next
+tries to insert to the page 'locked' by this very long-running transaction.
+
+Upgrade from 32-bit XID versions
+--------------------------------
+
+pg_upgrade doesn't change pages format itself. It is done lazily after.
+
+1. At first heap page read, tuples on a page are repacked to free 16 bytes
+at the end of a page, possibly freeing space from dead tuples.
+
+2A. 16 bytes of pd_special is added if there is a place for it
+
+2B. Page is converted to "Double XMAX" format if there is no place for
+pd_special
+
+3. If a page is in double XMAX format after its first read, and vacuum (or
+micro-vacuum at select query) could prune some tuples and free space for
+pd_special, prune_page will add pd_special and convert page from double XMAX
+to general 64-bit XID page format.
-- 
2.24.3 (Apple Git-128)


--cpok4wp6gsarlzvp--





^ permalink  raw  reply  [nested|flat] 265+ messages in thread

* [PATCH v6] README for 64bit xid
@ 2022-01-10 19:20 Pavel Borisov <[email protected]>
  0 siblings, 0 replies; 265+ messages in thread

From: Pavel Borisov @ 2022-01-10 19:20 UTC (permalink / raw)

Authors:
- Pavel Borisov <[email protected]>
- Maxim Orlov <[email protected]>
- Yura Sokolov <[email protected]> <[email protected]>
---
 src/backend/access/heap/README.XID64 | 128 +++++++++++++++++++++++++++
 1 file changed, 128 insertions(+)
 create mode 100644 src/backend/access/heap/README.XID64

diff --git a/src/backend/access/heap/README.XID64 b/src/backend/access/heap/README.XID64
new file mode 100644
index 00000000000..457ba9b9ef5
--- /dev/null
+++ b/src/backend/access/heap/README.XID64
@@ -0,0 +1,128 @@
+src/backend/access/heap/README.XID64
+
+64-bit Transaction ID's (XID)
+=============================
+
+A limited number (N = 2^32) of XID's required to do vacuum freeze to prevent
+wraparound every N/2 transactions. This causes performance degradation due
+to the need to exclusively lock tables while being vacuumed. In each
+wraparound cycle, SLRU buffers are also being cut.
+
+With 64-bit XID's wraparound is effectively postponed to a very distant
+future. Even in highly loaded systems that had 2^32 transactions per day
+it will take huge 2^31 days before the first enforced "vacuum to prevent
+wraparound"). Buffers cutting and routine vacuum are not enforced, and DBA
+can plan them independently at the time with the least system load and least
+critical for database performance. Also, it can be done less frequently
+(several times a year vs every several days) on systems with transaction rates
+similar to those mentioned above.
+
+On-disk tuple and page format
+-----------------------------
+
+On-disk tuple format remains unchanged. 32-bit t_xmin and t_xmax store the
+lower parts of 64-bit XMIN and XMAX values. Each heap page has additional
+64-bit pd_xid_base and pd_multi_base which are common for all tuples on a page.
+They are placed into a pd_special area - 16 bytes in the end of a heap page.
+Actual XMIN/XMAX for a tuple are calculated upon reading a tuple from a page
+as follows:
+
+XMIN = t_xmin + pd_xid_base. 					(1)
+XMAX = t_xmax + pd_xid_base/pd_multi_base.		(2)
+
+"Double XMAX" page format
+---------------------------------
+
+At first read of a heap page after pg_upgrade from 32-bit XID PostgreSQL
+version pd_special area with a size of 16 bytes should be added to a page.
+Though a page may not have space for this. Then it can be converted to a
+temporary format called "double XMAX".
+
+All tuples after pg-upgrade would necessarily have xmin = FrozenTransactionId.
+So we don't need tuple header t_xmin field and we reuse t_xmin to store higher
+32 bits of its XMAX.
+
+Double XMAX format is only for full pages that don't have 16 bytes for
+pd_special. So it neither has a�place for a single tuple. Insert and HOT update
+for double XMAX pages is impossible and not supported. We can only read or
+delete tuples from it.
+
+When we are able to prune page double XMAX it will be converted from it to
+general 64-bit XID page format with all operations on its tuples supported.
+
+In-memory tuple format
+----------------------
+
+In-memory tuple representation consists of two parts:
+- HeapTupleHeader from disk page (contains all heap tuple contents, not only
+header)
+- HeapTuple with additional in-memory fields
+
+HeapTuple for each tuple in memory stores t_xid_base/t_multi_base - a copies of
+page's pd_xid_base/pd_multi_base. With tuple's 32-bit t_xmin and t_xmax from
+HeapTupleHeader they are used to calculate actual 64-bit XMIN and XMAX:
+
+XMIN = t_xmin + t_xid_base. 					(3)
+XMAX = t_xmax + t_xid_base/t_multi_base.		(4)
+
+The downside of this is that we can not use tuple's XMIN and XMAX right away.
+We often need to re-read t_xmin and t_xmax - which could actually be pointers
+into a page in shared buffers and therefore they could be updated by any other
+backend.
+
+Update/delete with 64-bit XIDs and 32-bit t_xmin/t_xmax
+--------------------------------------------------------------
+
+When we try to delete/update a tuple, we check that XMAX for a page fits (2).
+I.e. that t_xmax will not be over MaxShortTransactionId relative to
+pd_xid_base/pd_multi_base of a its page.
+
+If the current XID doesn't fit a range
+(pd_xid_base, pd_xid_base + MaxShortTransactionId) (5):
+
+- heap_page_prepare_for_xid() will try to increase pd_xid_base/pd_multi_base on
+a page and update all t_xmin/t_xmax of the other tuples on the page to
+correspond new pd_xid_base/pd_multi_base.
+
+- If it was impossible, it will try to prune and freeze tuples on a page.
+
+- If this is unsuccessful it will throw an error. Normally this is very
+unlikely but if there is a very old living transaction with an age of around
+2^32 this can arise. Basically, this is a behavior similar to one during the
+vacuum to prevent wraparound when XID was 32-bit. Dba should take care and
+avoid very-long-living transactions with an age close to 2^32. So long-living
+transactions often they are most likely defunct.
+
+Insert with 64-bit XIDs and 32-bit t_xmin/t_xmax
+------------------------------------------------
+
+On insert we check if current XID fits a range (5). Otherwise:
+
+- heap_page_prepare_for_xid() will try to increase pd_xid_base for t_xmin will
+not be over MaxShortTransactionId.
+
+- If it is impossible, then it will try to prune and freeze tuples on a page.
+
+Known issue: if pd_xid_base could not be shifted to accommodate a tuple being
+inserted due to a very long-running transaction, we just throw an error. We
+neither try to insert a�tuple into another page nor mark the current page as
+full. So, in this (unlikely) case we will get regular insert errors on the next
+tries to insert to the page 'locked' by this very long-running transaction.
+
+Upgrade from 32-bit XID versions
+--------------------------------
+
+pg_upgrade doesn't change pages format itself. It is done lazily after.
+
+1. At first heap page read, tuples on a page are repacked to free 16 bytes
+at the end of a page, possibly freeing space from dead tuples.
+
+2A. 16 bytes of pd_special is added if there is a place for it
+
+2B. Page is converted to "Double XMAX" format if there is no place for
+pd_special
+
+3. If a page is in double XMAX format after its first read, and vacuum (or
+micro-vacuum at select query) could prune some tuples and free space for
+pd_special, prune_page will add pd_special and convert page from double XMAX
+to general 64-bit XID page format.
-- 
2.24.3 (Apple Git-128)


--cpok4wp6gsarlzvp--





^ permalink  raw  reply  [nested|flat] 265+ messages in thread

* [PATCH v6] README for 64bit xid
@ 2022-01-10 19:20 Pavel Borisov <[email protected]>
  0 siblings, 0 replies; 265+ messages in thread

From: Pavel Borisov @ 2022-01-10 19:20 UTC (permalink / raw)

Authors:
- Pavel Borisov <[email protected]>
- Maxim Orlov <[email protected]>
- Yura Sokolov <[email protected]> <[email protected]>
---
 src/backend/access/heap/README.XID64 | 128 +++++++++++++++++++++++++++
 1 file changed, 128 insertions(+)
 create mode 100644 src/backend/access/heap/README.XID64

diff --git a/src/backend/access/heap/README.XID64 b/src/backend/access/heap/README.XID64
new file mode 100644
index 00000000000..457ba9b9ef5
--- /dev/null
+++ b/src/backend/access/heap/README.XID64
@@ -0,0 +1,128 @@
+src/backend/access/heap/README.XID64
+
+64-bit Transaction ID's (XID)
+=============================
+
+A limited number (N = 2^32) of XID's required to do vacuum freeze to prevent
+wraparound every N/2 transactions. This causes performance degradation due
+to the need to exclusively lock tables while being vacuumed. In each
+wraparound cycle, SLRU buffers are also being cut.
+
+With 64-bit XID's wraparound is effectively postponed to a very distant
+future. Even in highly loaded systems that had 2^32 transactions per day
+it will take huge 2^31 days before the first enforced "vacuum to prevent
+wraparound"). Buffers cutting and routine vacuum are not enforced, and DBA
+can plan them independently at the time with the least system load and least
+critical for database performance. Also, it can be done less frequently
+(several times a year vs every several days) on systems with transaction rates
+similar to those mentioned above.
+
+On-disk tuple and page format
+-----------------------------
+
+On-disk tuple format remains unchanged. 32-bit t_xmin and t_xmax store the
+lower parts of 64-bit XMIN and XMAX values. Each heap page has additional
+64-bit pd_xid_base and pd_multi_base which are common for all tuples on a page.
+They are placed into a pd_special area - 16 bytes in the end of a heap page.
+Actual XMIN/XMAX for a tuple are calculated upon reading a tuple from a page
+as follows:
+
+XMIN = t_xmin + pd_xid_base. 					(1)
+XMAX = t_xmax + pd_xid_base/pd_multi_base.		(2)
+
+"Double XMAX" page format
+---------------------------------
+
+At first read of a heap page after pg_upgrade from 32-bit XID PostgreSQL
+version pd_special area with a size of 16 bytes should be added to a page.
+Though a page may not have space for this. Then it can be converted to a
+temporary format called "double XMAX".
+
+All tuples after pg-upgrade would necessarily have xmin = FrozenTransactionId.
+So we don't need tuple header t_xmin field and we reuse t_xmin to store higher
+32 bits of its XMAX.
+
+Double XMAX format is only for full pages that don't have 16 bytes for
+pd_special. So it neither has a�place for a single tuple. Insert and HOT update
+for double XMAX pages is impossible and not supported. We can only read or
+delete tuples from it.
+
+When we are able to prune page double XMAX it will be converted from it to
+general 64-bit XID page format with all operations on its tuples supported.
+
+In-memory tuple format
+----------------------
+
+In-memory tuple representation consists of two parts:
+- HeapTupleHeader from disk page (contains all heap tuple contents, not only
+header)
+- HeapTuple with additional in-memory fields
+
+HeapTuple for each tuple in memory stores t_xid_base/t_multi_base - a copies of
+page's pd_xid_base/pd_multi_base. With tuple's 32-bit t_xmin and t_xmax from
+HeapTupleHeader they are used to calculate actual 64-bit XMIN and XMAX:
+
+XMIN = t_xmin + t_xid_base. 					(3)
+XMAX = t_xmax + t_xid_base/t_multi_base.		(4)
+
+The downside of this is that we can not use tuple's XMIN and XMAX right away.
+We often need to re-read t_xmin and t_xmax - which could actually be pointers
+into a page in shared buffers and therefore they could be updated by any other
+backend.
+
+Update/delete with 64-bit XIDs and 32-bit t_xmin/t_xmax
+--------------------------------------------------------------
+
+When we try to delete/update a tuple, we check that XMAX for a page fits (2).
+I.e. that t_xmax will not be over MaxShortTransactionId relative to
+pd_xid_base/pd_multi_base of a its page.
+
+If the current XID doesn't fit a range
+(pd_xid_base, pd_xid_base + MaxShortTransactionId) (5):
+
+- heap_page_prepare_for_xid() will try to increase pd_xid_base/pd_multi_base on
+a page and update all t_xmin/t_xmax of the other tuples on the page to
+correspond new pd_xid_base/pd_multi_base.
+
+- If it was impossible, it will try to prune and freeze tuples on a page.
+
+- If this is unsuccessful it will throw an error. Normally this is very
+unlikely but if there is a very old living transaction with an age of around
+2^32 this can arise. Basically, this is a behavior similar to one during the
+vacuum to prevent wraparound when XID was 32-bit. Dba should take care and
+avoid very-long-living transactions with an age close to 2^32. So long-living
+transactions often they are most likely defunct.
+
+Insert with 64-bit XIDs and 32-bit t_xmin/t_xmax
+------------------------------------------------
+
+On insert we check if current XID fits a range (5). Otherwise:
+
+- heap_page_prepare_for_xid() will try to increase pd_xid_base for t_xmin will
+not be over MaxShortTransactionId.
+
+- If it is impossible, then it will try to prune and freeze tuples on a page.
+
+Known issue: if pd_xid_base could not be shifted to accommodate a tuple being
+inserted due to a very long-running transaction, we just throw an error. We
+neither try to insert a�tuple into another page nor mark the current page as
+full. So, in this (unlikely) case we will get regular insert errors on the next
+tries to insert to the page 'locked' by this very long-running transaction.
+
+Upgrade from 32-bit XID versions
+--------------------------------
+
+pg_upgrade doesn't change pages format itself. It is done lazily after.
+
+1. At first heap page read, tuples on a page are repacked to free 16 bytes
+at the end of a page, possibly freeing space from dead tuples.
+
+2A. 16 bytes of pd_special is added if there is a place for it
+
+2B. Page is converted to "Double XMAX" format if there is no place for
+pd_special
+
+3. If a page is in double XMAX format after its first read, and vacuum (or
+micro-vacuum at select query) could prune some tuples and free space for
+pd_special, prune_page will add pd_special and convert page from double XMAX
+to general 64-bit XID page format.
-- 
2.24.3 (Apple Git-128)


--cpok4wp6gsarlzvp--





^ permalink  raw  reply  [nested|flat] 265+ messages in thread

* [PATCH v6] README for 64bit xid
@ 2022-01-10 19:20 Pavel Borisov <[email protected]>
  0 siblings, 0 replies; 265+ messages in thread

From: Pavel Borisov @ 2022-01-10 19:20 UTC (permalink / raw)

Authors:
- Pavel Borisov <[email protected]>
- Maxim Orlov <[email protected]>
- Yura Sokolov <[email protected]> <[email protected]>
---
 src/backend/access/heap/README.XID64 | 128 +++++++++++++++++++++++++++
 1 file changed, 128 insertions(+)
 create mode 100644 src/backend/access/heap/README.XID64

diff --git a/src/backend/access/heap/README.XID64 b/src/backend/access/heap/README.XID64
new file mode 100644
index 00000000000..457ba9b9ef5
--- /dev/null
+++ b/src/backend/access/heap/README.XID64
@@ -0,0 +1,128 @@
+src/backend/access/heap/README.XID64
+
+64-bit Transaction ID's (XID)
+=============================
+
+A limited number (N = 2^32) of XID's required to do vacuum freeze to prevent
+wraparound every N/2 transactions. This causes performance degradation due
+to the need to exclusively lock tables while being vacuumed. In each
+wraparound cycle, SLRU buffers are also being cut.
+
+With 64-bit XID's wraparound is effectively postponed to a very distant
+future. Even in highly loaded systems that had 2^32 transactions per day
+it will take huge 2^31 days before the first enforced "vacuum to prevent
+wraparound"). Buffers cutting and routine vacuum are not enforced, and DBA
+can plan them independently at the time with the least system load and least
+critical for database performance. Also, it can be done less frequently
+(several times a year vs every several days) on systems with transaction rates
+similar to those mentioned above.
+
+On-disk tuple and page format
+-----------------------------
+
+On-disk tuple format remains unchanged. 32-bit t_xmin and t_xmax store the
+lower parts of 64-bit XMIN and XMAX values. Each heap page has additional
+64-bit pd_xid_base and pd_multi_base which are common for all tuples on a page.
+They are placed into a pd_special area - 16 bytes in the end of a heap page.
+Actual XMIN/XMAX for a tuple are calculated upon reading a tuple from a page
+as follows:
+
+XMIN = t_xmin + pd_xid_base. 					(1)
+XMAX = t_xmax + pd_xid_base/pd_multi_base.		(2)
+
+"Double XMAX" page format
+---------------------------------
+
+At first read of a heap page after pg_upgrade from 32-bit XID PostgreSQL
+version pd_special area with a size of 16 bytes should be added to a page.
+Though a page may not have space for this. Then it can be converted to a
+temporary format called "double XMAX".
+
+All tuples after pg-upgrade would necessarily have xmin = FrozenTransactionId.
+So we don't need tuple header t_xmin field and we reuse t_xmin to store higher
+32 bits of its XMAX.
+
+Double XMAX format is only for full pages that don't have 16 bytes for
+pd_special. So it neither has a�place for a single tuple. Insert and HOT update
+for double XMAX pages is impossible and not supported. We can only read or
+delete tuples from it.
+
+When we are able to prune page double XMAX it will be converted from it to
+general 64-bit XID page format with all operations on its tuples supported.
+
+In-memory tuple format
+----------------------
+
+In-memory tuple representation consists of two parts:
+- HeapTupleHeader from disk page (contains all heap tuple contents, not only
+header)
+- HeapTuple with additional in-memory fields
+
+HeapTuple for each tuple in memory stores t_xid_base/t_multi_base - a copies of
+page's pd_xid_base/pd_multi_base. With tuple's 32-bit t_xmin and t_xmax from
+HeapTupleHeader they are used to calculate actual 64-bit XMIN and XMAX:
+
+XMIN = t_xmin + t_xid_base. 					(3)
+XMAX = t_xmax + t_xid_base/t_multi_base.		(4)
+
+The downside of this is that we can not use tuple's XMIN and XMAX right away.
+We often need to re-read t_xmin and t_xmax - which could actually be pointers
+into a page in shared buffers and therefore they could be updated by any other
+backend.
+
+Update/delete with 64-bit XIDs and 32-bit t_xmin/t_xmax
+--------------------------------------------------------------
+
+When we try to delete/update a tuple, we check that XMAX for a page fits (2).
+I.e. that t_xmax will not be over MaxShortTransactionId relative to
+pd_xid_base/pd_multi_base of a its page.
+
+If the current XID doesn't fit a range
+(pd_xid_base, pd_xid_base + MaxShortTransactionId) (5):
+
+- heap_page_prepare_for_xid() will try to increase pd_xid_base/pd_multi_base on
+a page and update all t_xmin/t_xmax of the other tuples on the page to
+correspond new pd_xid_base/pd_multi_base.
+
+- If it was impossible, it will try to prune and freeze tuples on a page.
+
+- If this is unsuccessful it will throw an error. Normally this is very
+unlikely but if there is a very old living transaction with an age of around
+2^32 this can arise. Basically, this is a behavior similar to one during the
+vacuum to prevent wraparound when XID was 32-bit. Dba should take care and
+avoid very-long-living transactions with an age close to 2^32. So long-living
+transactions often they are most likely defunct.
+
+Insert with 64-bit XIDs and 32-bit t_xmin/t_xmax
+------------------------------------------------
+
+On insert we check if current XID fits a range (5). Otherwise:
+
+- heap_page_prepare_for_xid() will try to increase pd_xid_base for t_xmin will
+not be over MaxShortTransactionId.
+
+- If it is impossible, then it will try to prune and freeze tuples on a page.
+
+Known issue: if pd_xid_base could not be shifted to accommodate a tuple being
+inserted due to a very long-running transaction, we just throw an error. We
+neither try to insert a�tuple into another page nor mark the current page as
+full. So, in this (unlikely) case we will get regular insert errors on the next
+tries to insert to the page 'locked' by this very long-running transaction.
+
+Upgrade from 32-bit XID versions
+--------------------------------
+
+pg_upgrade doesn't change pages format itself. It is done lazily after.
+
+1. At first heap page read, tuples on a page are repacked to free 16 bytes
+at the end of a page, possibly freeing space from dead tuples.
+
+2A. 16 bytes of pd_special is added if there is a place for it
+
+2B. Page is converted to "Double XMAX" format if there is no place for
+pd_special
+
+3. If a page is in double XMAX format after its first read, and vacuum (or
+micro-vacuum at select query) could prune some tuples and free space for
+pd_special, prune_page will add pd_special and convert page from double XMAX
+to general 64-bit XID page format.
-- 
2.24.3 (Apple Git-128)


--cpok4wp6gsarlzvp--





^ permalink  raw  reply  [nested|flat] 265+ messages in thread

* [PATCH v6] README for 64bit xid
@ 2022-01-10 19:20 Pavel Borisov <[email protected]>
  0 siblings, 0 replies; 265+ messages in thread

From: Pavel Borisov @ 2022-01-10 19:20 UTC (permalink / raw)

Authors:
- Pavel Borisov <[email protected]>
- Maxim Orlov <[email protected]>
- Yura Sokolov <[email protected]> <[email protected]>
---
 src/backend/access/heap/README.XID64 | 128 +++++++++++++++++++++++++++
 1 file changed, 128 insertions(+)
 create mode 100644 src/backend/access/heap/README.XID64

diff --git a/src/backend/access/heap/README.XID64 b/src/backend/access/heap/README.XID64
new file mode 100644
index 00000000000..457ba9b9ef5
--- /dev/null
+++ b/src/backend/access/heap/README.XID64
@@ -0,0 +1,128 @@
+src/backend/access/heap/README.XID64
+
+64-bit Transaction ID's (XID)
+=============================
+
+A limited number (N = 2^32) of XID's required to do vacuum freeze to prevent
+wraparound every N/2 transactions. This causes performance degradation due
+to the need to exclusively lock tables while being vacuumed. In each
+wraparound cycle, SLRU buffers are also being cut.
+
+With 64-bit XID's wraparound is effectively postponed to a very distant
+future. Even in highly loaded systems that had 2^32 transactions per day
+it will take huge 2^31 days before the first enforced "vacuum to prevent
+wraparound"). Buffers cutting and routine vacuum are not enforced, and DBA
+can plan them independently at the time with the least system load and least
+critical for database performance. Also, it can be done less frequently
+(several times a year vs every several days) on systems with transaction rates
+similar to those mentioned above.
+
+On-disk tuple and page format
+-----------------------------
+
+On-disk tuple format remains unchanged. 32-bit t_xmin and t_xmax store the
+lower parts of 64-bit XMIN and XMAX values. Each heap page has additional
+64-bit pd_xid_base and pd_multi_base which are common for all tuples on a page.
+They are placed into a pd_special area - 16 bytes in the end of a heap page.
+Actual XMIN/XMAX for a tuple are calculated upon reading a tuple from a page
+as follows:
+
+XMIN = t_xmin + pd_xid_base. 					(1)
+XMAX = t_xmax + pd_xid_base/pd_multi_base.		(2)
+
+"Double XMAX" page format
+---------------------------------
+
+At first read of a heap page after pg_upgrade from 32-bit XID PostgreSQL
+version pd_special area with a size of 16 bytes should be added to a page.
+Though a page may not have space for this. Then it can be converted to a
+temporary format called "double XMAX".
+
+All tuples after pg-upgrade would necessarily have xmin = FrozenTransactionId.
+So we don't need tuple header t_xmin field and we reuse t_xmin to store higher
+32 bits of its XMAX.
+
+Double XMAX format is only for full pages that don't have 16 bytes for
+pd_special. So it neither has a�place for a single tuple. Insert and HOT update
+for double XMAX pages is impossible and not supported. We can only read or
+delete tuples from it.
+
+When we are able to prune page double XMAX it will be converted from it to
+general 64-bit XID page format with all operations on its tuples supported.
+
+In-memory tuple format
+----------------------
+
+In-memory tuple representation consists of two parts:
+- HeapTupleHeader from disk page (contains all heap tuple contents, not only
+header)
+- HeapTuple with additional in-memory fields
+
+HeapTuple for each tuple in memory stores t_xid_base/t_multi_base - a copies of
+page's pd_xid_base/pd_multi_base. With tuple's 32-bit t_xmin and t_xmax from
+HeapTupleHeader they are used to calculate actual 64-bit XMIN and XMAX:
+
+XMIN = t_xmin + t_xid_base. 					(3)
+XMAX = t_xmax + t_xid_base/t_multi_base.		(4)
+
+The downside of this is that we can not use tuple's XMIN and XMAX right away.
+We often need to re-read t_xmin and t_xmax - which could actually be pointers
+into a page in shared buffers and therefore they could be updated by any other
+backend.
+
+Update/delete with 64-bit XIDs and 32-bit t_xmin/t_xmax
+--------------------------------------------------------------
+
+When we try to delete/update a tuple, we check that XMAX for a page fits (2).
+I.e. that t_xmax will not be over MaxShortTransactionId relative to
+pd_xid_base/pd_multi_base of a its page.
+
+If the current XID doesn't fit a range
+(pd_xid_base, pd_xid_base + MaxShortTransactionId) (5):
+
+- heap_page_prepare_for_xid() will try to increase pd_xid_base/pd_multi_base on
+a page and update all t_xmin/t_xmax of the other tuples on the page to
+correspond new pd_xid_base/pd_multi_base.
+
+- If it was impossible, it will try to prune and freeze tuples on a page.
+
+- If this is unsuccessful it will throw an error. Normally this is very
+unlikely but if there is a very old living transaction with an age of around
+2^32 this can arise. Basically, this is a behavior similar to one during the
+vacuum to prevent wraparound when XID was 32-bit. Dba should take care and
+avoid very-long-living transactions with an age close to 2^32. So long-living
+transactions often they are most likely defunct.
+
+Insert with 64-bit XIDs and 32-bit t_xmin/t_xmax
+------------------------------------------------
+
+On insert we check if current XID fits a range (5). Otherwise:
+
+- heap_page_prepare_for_xid() will try to increase pd_xid_base for t_xmin will
+not be over MaxShortTransactionId.
+
+- If it is impossible, then it will try to prune and freeze tuples on a page.
+
+Known issue: if pd_xid_base could not be shifted to accommodate a tuple being
+inserted due to a very long-running transaction, we just throw an error. We
+neither try to insert a�tuple into another page nor mark the current page as
+full. So, in this (unlikely) case we will get regular insert errors on the next
+tries to insert to the page 'locked' by this very long-running transaction.
+
+Upgrade from 32-bit XID versions
+--------------------------------
+
+pg_upgrade doesn't change pages format itself. It is done lazily after.
+
+1. At first heap page read, tuples on a page are repacked to free 16 bytes
+at the end of a page, possibly freeing space from dead tuples.
+
+2A. 16 bytes of pd_special is added if there is a place for it
+
+2B. Page is converted to "Double XMAX" format if there is no place for
+pd_special
+
+3. If a page is in double XMAX format after its first read, and vacuum (or
+micro-vacuum at select query) could prune some tuples and free space for
+pd_special, prune_page will add pd_special and convert page from double XMAX
+to general 64-bit XID page format.
-- 
2.24.3 (Apple Git-128)


--cpok4wp6gsarlzvp--





^ permalink  raw  reply  [nested|flat] 265+ messages in thread

* [PATCH v6] README for 64bit xid
@ 2022-01-10 19:20 Pavel Borisov <[email protected]>
  0 siblings, 0 replies; 265+ messages in thread

From: Pavel Borisov @ 2022-01-10 19:20 UTC (permalink / raw)

Authors:
- Pavel Borisov <[email protected]>
- Maxim Orlov <[email protected]>
- Yura Sokolov <[email protected]> <[email protected]>
---
 src/backend/access/heap/README.XID64 | 128 +++++++++++++++++++++++++++
 1 file changed, 128 insertions(+)
 create mode 100644 src/backend/access/heap/README.XID64

diff --git a/src/backend/access/heap/README.XID64 b/src/backend/access/heap/README.XID64
new file mode 100644
index 00000000000..457ba9b9ef5
--- /dev/null
+++ b/src/backend/access/heap/README.XID64
@@ -0,0 +1,128 @@
+src/backend/access/heap/README.XID64
+
+64-bit Transaction ID's (XID)
+=============================
+
+A limited number (N = 2^32) of XID's required to do vacuum freeze to prevent
+wraparound every N/2 transactions. This causes performance degradation due
+to the need to exclusively lock tables while being vacuumed. In each
+wraparound cycle, SLRU buffers are also being cut.
+
+With 64-bit XID's wraparound is effectively postponed to a very distant
+future. Even in highly loaded systems that had 2^32 transactions per day
+it will take huge 2^31 days before the first enforced "vacuum to prevent
+wraparound"). Buffers cutting and routine vacuum are not enforced, and DBA
+can plan them independently at the time with the least system load and least
+critical for database performance. Also, it can be done less frequently
+(several times a year vs every several days) on systems with transaction rates
+similar to those mentioned above.
+
+On-disk tuple and page format
+-----------------------------
+
+On-disk tuple format remains unchanged. 32-bit t_xmin and t_xmax store the
+lower parts of 64-bit XMIN and XMAX values. Each heap page has additional
+64-bit pd_xid_base and pd_multi_base which are common for all tuples on a page.
+They are placed into a pd_special area - 16 bytes in the end of a heap page.
+Actual XMIN/XMAX for a tuple are calculated upon reading a tuple from a page
+as follows:
+
+XMIN = t_xmin + pd_xid_base. 					(1)
+XMAX = t_xmax + pd_xid_base/pd_multi_base.		(2)
+
+"Double XMAX" page format
+---------------------------------
+
+At first read of a heap page after pg_upgrade from 32-bit XID PostgreSQL
+version pd_special area with a size of 16 bytes should be added to a page.
+Though a page may not have space for this. Then it can be converted to a
+temporary format called "double XMAX".
+
+All tuples after pg-upgrade would necessarily have xmin = FrozenTransactionId.
+So we don't need tuple header t_xmin field and we reuse t_xmin to store higher
+32 bits of its XMAX.
+
+Double XMAX format is only for full pages that don't have 16 bytes for
+pd_special. So it neither has a�place for a single tuple. Insert and HOT update
+for double XMAX pages is impossible and not supported. We can only read or
+delete tuples from it.
+
+When we are able to prune page double XMAX it will be converted from it to
+general 64-bit XID page format with all operations on its tuples supported.
+
+In-memory tuple format
+----------------------
+
+In-memory tuple representation consists of two parts:
+- HeapTupleHeader from disk page (contains all heap tuple contents, not only
+header)
+- HeapTuple with additional in-memory fields
+
+HeapTuple for each tuple in memory stores t_xid_base/t_multi_base - a copies of
+page's pd_xid_base/pd_multi_base. With tuple's 32-bit t_xmin and t_xmax from
+HeapTupleHeader they are used to calculate actual 64-bit XMIN and XMAX:
+
+XMIN = t_xmin + t_xid_base. 					(3)
+XMAX = t_xmax + t_xid_base/t_multi_base.		(4)
+
+The downside of this is that we can not use tuple's XMIN and XMAX right away.
+We often need to re-read t_xmin and t_xmax - which could actually be pointers
+into a page in shared buffers and therefore they could be updated by any other
+backend.
+
+Update/delete with 64-bit XIDs and 32-bit t_xmin/t_xmax
+--------------------------------------------------------------
+
+When we try to delete/update a tuple, we check that XMAX for a page fits (2).
+I.e. that t_xmax will not be over MaxShortTransactionId relative to
+pd_xid_base/pd_multi_base of a its page.
+
+If the current XID doesn't fit a range
+(pd_xid_base, pd_xid_base + MaxShortTransactionId) (5):
+
+- heap_page_prepare_for_xid() will try to increase pd_xid_base/pd_multi_base on
+a page and update all t_xmin/t_xmax of the other tuples on the page to
+correspond new pd_xid_base/pd_multi_base.
+
+- If it was impossible, it will try to prune and freeze tuples on a page.
+
+- If this is unsuccessful it will throw an error. Normally this is very
+unlikely but if there is a very old living transaction with an age of around
+2^32 this can arise. Basically, this is a behavior similar to one during the
+vacuum to prevent wraparound when XID was 32-bit. Dba should take care and
+avoid very-long-living transactions with an age close to 2^32. So long-living
+transactions often they are most likely defunct.
+
+Insert with 64-bit XIDs and 32-bit t_xmin/t_xmax
+------------------------------------------------
+
+On insert we check if current XID fits a range (5). Otherwise:
+
+- heap_page_prepare_for_xid() will try to increase pd_xid_base for t_xmin will
+not be over MaxShortTransactionId.
+
+- If it is impossible, then it will try to prune and freeze tuples on a page.
+
+Known issue: if pd_xid_base could not be shifted to accommodate a tuple being
+inserted due to a very long-running transaction, we just throw an error. We
+neither try to insert a�tuple into another page nor mark the current page as
+full. So, in this (unlikely) case we will get regular insert errors on the next
+tries to insert to the page 'locked' by this very long-running transaction.
+
+Upgrade from 32-bit XID versions
+--------------------------------
+
+pg_upgrade doesn't change pages format itself. It is done lazily after.
+
+1. At first heap page read, tuples on a page are repacked to free 16 bytes
+at the end of a page, possibly freeing space from dead tuples.
+
+2A. 16 bytes of pd_special is added if there is a place for it
+
+2B. Page is converted to "Double XMAX" format if there is no place for
+pd_special
+
+3. If a page is in double XMAX format after its first read, and vacuum (or
+micro-vacuum at select query) could prune some tuples and free space for
+pd_special, prune_page will add pd_special and convert page from double XMAX
+to general 64-bit XID page format.
-- 
2.24.3 (Apple Git-128)


--cpok4wp6gsarlzvp--





^ permalink  raw  reply  [nested|flat] 265+ messages in thread

* [PATCH v6] README for 64bit xid
@ 2022-01-10 19:20 Pavel Borisov <[email protected]>
  0 siblings, 0 replies; 265+ messages in thread

From: Pavel Borisov @ 2022-01-10 19:20 UTC (permalink / raw)

Authors:
- Pavel Borisov <[email protected]>
- Maxim Orlov <[email protected]>
- Yura Sokolov <[email protected]> <[email protected]>
---
 src/backend/access/heap/README.XID64 | 128 +++++++++++++++++++++++++++
 1 file changed, 128 insertions(+)
 create mode 100644 src/backend/access/heap/README.XID64

diff --git a/src/backend/access/heap/README.XID64 b/src/backend/access/heap/README.XID64
new file mode 100644
index 00000000000..457ba9b9ef5
--- /dev/null
+++ b/src/backend/access/heap/README.XID64
@@ -0,0 +1,128 @@
+src/backend/access/heap/README.XID64
+
+64-bit Transaction ID's (XID)
+=============================
+
+A limited number (N = 2^32) of XID's required to do vacuum freeze to prevent
+wraparound every N/2 transactions. This causes performance degradation due
+to the need to exclusively lock tables while being vacuumed. In each
+wraparound cycle, SLRU buffers are also being cut.
+
+With 64-bit XID's wraparound is effectively postponed to a very distant
+future. Even in highly loaded systems that had 2^32 transactions per day
+it will take huge 2^31 days before the first enforced "vacuum to prevent
+wraparound"). Buffers cutting and routine vacuum are not enforced, and DBA
+can plan them independently at the time with the least system load and least
+critical for database performance. Also, it can be done less frequently
+(several times a year vs every several days) on systems with transaction rates
+similar to those mentioned above.
+
+On-disk tuple and page format
+-----------------------------
+
+On-disk tuple format remains unchanged. 32-bit t_xmin and t_xmax store the
+lower parts of 64-bit XMIN and XMAX values. Each heap page has additional
+64-bit pd_xid_base and pd_multi_base which are common for all tuples on a page.
+They are placed into a pd_special area - 16 bytes in the end of a heap page.
+Actual XMIN/XMAX for a tuple are calculated upon reading a tuple from a page
+as follows:
+
+XMIN = t_xmin + pd_xid_base. 					(1)
+XMAX = t_xmax + pd_xid_base/pd_multi_base.		(2)
+
+"Double XMAX" page format
+---------------------------------
+
+At first read of a heap page after pg_upgrade from 32-bit XID PostgreSQL
+version pd_special area with a size of 16 bytes should be added to a page.
+Though a page may not have space for this. Then it can be converted to a
+temporary format called "double XMAX".
+
+All tuples after pg-upgrade would necessarily have xmin = FrozenTransactionId.
+So we don't need tuple header t_xmin field and we reuse t_xmin to store higher
+32 bits of its XMAX.
+
+Double XMAX format is only for full pages that don't have 16 bytes for
+pd_special. So it neither has a�place for a single tuple. Insert and HOT update
+for double XMAX pages is impossible and not supported. We can only read or
+delete tuples from it.
+
+When we are able to prune page double XMAX it will be converted from it to
+general 64-bit XID page format with all operations on its tuples supported.
+
+In-memory tuple format
+----------------------
+
+In-memory tuple representation consists of two parts:
+- HeapTupleHeader from disk page (contains all heap tuple contents, not only
+header)
+- HeapTuple with additional in-memory fields
+
+HeapTuple for each tuple in memory stores t_xid_base/t_multi_base - a copies of
+page's pd_xid_base/pd_multi_base. With tuple's 32-bit t_xmin and t_xmax from
+HeapTupleHeader they are used to calculate actual 64-bit XMIN and XMAX:
+
+XMIN = t_xmin + t_xid_base. 					(3)
+XMAX = t_xmax + t_xid_base/t_multi_base.		(4)
+
+The downside of this is that we can not use tuple's XMIN and XMAX right away.
+We often need to re-read t_xmin and t_xmax - which could actually be pointers
+into a page in shared buffers and therefore they could be updated by any other
+backend.
+
+Update/delete with 64-bit XIDs and 32-bit t_xmin/t_xmax
+--------------------------------------------------------------
+
+When we try to delete/update a tuple, we check that XMAX for a page fits (2).
+I.e. that t_xmax will not be over MaxShortTransactionId relative to
+pd_xid_base/pd_multi_base of a its page.
+
+If the current XID doesn't fit a range
+(pd_xid_base, pd_xid_base + MaxShortTransactionId) (5):
+
+- heap_page_prepare_for_xid() will try to increase pd_xid_base/pd_multi_base on
+a page and update all t_xmin/t_xmax of the other tuples on the page to
+correspond new pd_xid_base/pd_multi_base.
+
+- If it was impossible, it will try to prune and freeze tuples on a page.
+
+- If this is unsuccessful it will throw an error. Normally this is very
+unlikely but if there is a very old living transaction with an age of around
+2^32 this can arise. Basically, this is a behavior similar to one during the
+vacuum to prevent wraparound when XID was 32-bit. Dba should take care and
+avoid very-long-living transactions with an age close to 2^32. So long-living
+transactions often they are most likely defunct.
+
+Insert with 64-bit XIDs and 32-bit t_xmin/t_xmax
+------------------------------------------------
+
+On insert we check if current XID fits a range (5). Otherwise:
+
+- heap_page_prepare_for_xid() will try to increase pd_xid_base for t_xmin will
+not be over MaxShortTransactionId.
+
+- If it is impossible, then it will try to prune and freeze tuples on a page.
+
+Known issue: if pd_xid_base could not be shifted to accommodate a tuple being
+inserted due to a very long-running transaction, we just throw an error. We
+neither try to insert a�tuple into another page nor mark the current page as
+full. So, in this (unlikely) case we will get regular insert errors on the next
+tries to insert to the page 'locked' by this very long-running transaction.
+
+Upgrade from 32-bit XID versions
+--------------------------------
+
+pg_upgrade doesn't change pages format itself. It is done lazily after.
+
+1. At first heap page read, tuples on a page are repacked to free 16 bytes
+at the end of a page, possibly freeing space from dead tuples.
+
+2A. 16 bytes of pd_special is added if there is a place for it
+
+2B. Page is converted to "Double XMAX" format if there is no place for
+pd_special
+
+3. If a page is in double XMAX format after its first read, and vacuum (or
+micro-vacuum at select query) could prune some tuples and free space for
+pd_special, prune_page will add pd_special and convert page from double XMAX
+to general 64-bit XID page format.
-- 
2.24.3 (Apple Git-128)


--cpok4wp6gsarlzvp--





^ permalink  raw  reply  [nested|flat] 265+ messages in thread

* [PATCH v6] README for 64bit xid
@ 2022-01-10 19:20 Pavel Borisov <[email protected]>
  0 siblings, 0 replies; 265+ messages in thread

From: Pavel Borisov @ 2022-01-10 19:20 UTC (permalink / raw)

Authors:
- Pavel Borisov <[email protected]>
- Maxim Orlov <[email protected]>
- Yura Sokolov <[email protected]> <[email protected]>
---
 src/backend/access/heap/README.XID64 | 128 +++++++++++++++++++++++++++
 1 file changed, 128 insertions(+)
 create mode 100644 src/backend/access/heap/README.XID64

diff --git a/src/backend/access/heap/README.XID64 b/src/backend/access/heap/README.XID64
new file mode 100644
index 00000000000..457ba9b9ef5
--- /dev/null
+++ b/src/backend/access/heap/README.XID64
@@ -0,0 +1,128 @@
+src/backend/access/heap/README.XID64
+
+64-bit Transaction ID's (XID)
+=============================
+
+A limited number (N = 2^32) of XID's required to do vacuum freeze to prevent
+wraparound every N/2 transactions. This causes performance degradation due
+to the need to exclusively lock tables while being vacuumed. In each
+wraparound cycle, SLRU buffers are also being cut.
+
+With 64-bit XID's wraparound is effectively postponed to a very distant
+future. Even in highly loaded systems that had 2^32 transactions per day
+it will take huge 2^31 days before the first enforced "vacuum to prevent
+wraparound"). Buffers cutting and routine vacuum are not enforced, and DBA
+can plan them independently at the time with the least system load and least
+critical for database performance. Also, it can be done less frequently
+(several times a year vs every several days) on systems with transaction rates
+similar to those mentioned above.
+
+On-disk tuple and page format
+-----------------------------
+
+On-disk tuple format remains unchanged. 32-bit t_xmin and t_xmax store the
+lower parts of 64-bit XMIN and XMAX values. Each heap page has additional
+64-bit pd_xid_base and pd_multi_base which are common for all tuples on a page.
+They are placed into a pd_special area - 16 bytes in the end of a heap page.
+Actual XMIN/XMAX for a tuple are calculated upon reading a tuple from a page
+as follows:
+
+XMIN = t_xmin + pd_xid_base. 					(1)
+XMAX = t_xmax + pd_xid_base/pd_multi_base.		(2)
+
+"Double XMAX" page format
+---------------------------------
+
+At first read of a heap page after pg_upgrade from 32-bit XID PostgreSQL
+version pd_special area with a size of 16 bytes should be added to a page.
+Though a page may not have space for this. Then it can be converted to a
+temporary format called "double XMAX".
+
+All tuples after pg-upgrade would necessarily have xmin = FrozenTransactionId.
+So we don't need tuple header t_xmin field and we reuse t_xmin to store higher
+32 bits of its XMAX.
+
+Double XMAX format is only for full pages that don't have 16 bytes for
+pd_special. So it neither has a�place for a single tuple. Insert and HOT update
+for double XMAX pages is impossible and not supported. We can only read or
+delete tuples from it.
+
+When we are able to prune page double XMAX it will be converted from it to
+general 64-bit XID page format with all operations on its tuples supported.
+
+In-memory tuple format
+----------------------
+
+In-memory tuple representation consists of two parts:
+- HeapTupleHeader from disk page (contains all heap tuple contents, not only
+header)
+- HeapTuple with additional in-memory fields
+
+HeapTuple for each tuple in memory stores t_xid_base/t_multi_base - a copies of
+page's pd_xid_base/pd_multi_base. With tuple's 32-bit t_xmin and t_xmax from
+HeapTupleHeader they are used to calculate actual 64-bit XMIN and XMAX:
+
+XMIN = t_xmin + t_xid_base. 					(3)
+XMAX = t_xmax + t_xid_base/t_multi_base.		(4)
+
+The downside of this is that we can not use tuple's XMIN and XMAX right away.
+We often need to re-read t_xmin and t_xmax - which could actually be pointers
+into a page in shared buffers and therefore they could be updated by any other
+backend.
+
+Update/delete with 64-bit XIDs and 32-bit t_xmin/t_xmax
+--------------------------------------------------------------
+
+When we try to delete/update a tuple, we check that XMAX for a page fits (2).
+I.e. that t_xmax will not be over MaxShortTransactionId relative to
+pd_xid_base/pd_multi_base of a its page.
+
+If the current XID doesn't fit a range
+(pd_xid_base, pd_xid_base + MaxShortTransactionId) (5):
+
+- heap_page_prepare_for_xid() will try to increase pd_xid_base/pd_multi_base on
+a page and update all t_xmin/t_xmax of the other tuples on the page to
+correspond new pd_xid_base/pd_multi_base.
+
+- If it was impossible, it will try to prune and freeze tuples on a page.
+
+- If this is unsuccessful it will throw an error. Normally this is very
+unlikely but if there is a very old living transaction with an age of around
+2^32 this can arise. Basically, this is a behavior similar to one during the
+vacuum to prevent wraparound when XID was 32-bit. Dba should take care and
+avoid very-long-living transactions with an age close to 2^32. So long-living
+transactions often they are most likely defunct.
+
+Insert with 64-bit XIDs and 32-bit t_xmin/t_xmax
+------------------------------------------------
+
+On insert we check if current XID fits a range (5). Otherwise:
+
+- heap_page_prepare_for_xid() will try to increase pd_xid_base for t_xmin will
+not be over MaxShortTransactionId.
+
+- If it is impossible, then it will try to prune and freeze tuples on a page.
+
+Known issue: if pd_xid_base could not be shifted to accommodate a tuple being
+inserted due to a very long-running transaction, we just throw an error. We
+neither try to insert a�tuple into another page nor mark the current page as
+full. So, in this (unlikely) case we will get regular insert errors on the next
+tries to insert to the page 'locked' by this very long-running transaction.
+
+Upgrade from 32-bit XID versions
+--------------------------------
+
+pg_upgrade doesn't change pages format itself. It is done lazily after.
+
+1. At first heap page read, tuples on a page are repacked to free 16 bytes
+at the end of a page, possibly freeing space from dead tuples.
+
+2A. 16 bytes of pd_special is added if there is a place for it
+
+2B. Page is converted to "Double XMAX" format if there is no place for
+pd_special
+
+3. If a page is in double XMAX format after its first read, and vacuum (or
+micro-vacuum at select query) could prune some tuples and free space for
+pd_special, prune_page will add pd_special and convert page from double XMAX
+to general 64-bit XID page format.
-- 
2.24.3 (Apple Git-128)


--cpok4wp6gsarlzvp--





^ permalink  raw  reply  [nested|flat] 265+ messages in thread

* [PATCH v6] README for 64bit xid
@ 2022-01-10 19:20 Pavel Borisov <[email protected]>
  0 siblings, 0 replies; 265+ messages in thread

From: Pavel Borisov @ 2022-01-10 19:20 UTC (permalink / raw)

Authors:
- Pavel Borisov <[email protected]>
- Maxim Orlov <[email protected]>
- Yura Sokolov <[email protected]> <[email protected]>
---
 src/backend/access/heap/README.XID64 | 128 +++++++++++++++++++++++++++
 1 file changed, 128 insertions(+)
 create mode 100644 src/backend/access/heap/README.XID64

diff --git a/src/backend/access/heap/README.XID64 b/src/backend/access/heap/README.XID64
new file mode 100644
index 00000000000..457ba9b9ef5
--- /dev/null
+++ b/src/backend/access/heap/README.XID64
@@ -0,0 +1,128 @@
+src/backend/access/heap/README.XID64
+
+64-bit Transaction ID's (XID)
+=============================
+
+A limited number (N = 2^32) of XID's required to do vacuum freeze to prevent
+wraparound every N/2 transactions. This causes performance degradation due
+to the need to exclusively lock tables while being vacuumed. In each
+wraparound cycle, SLRU buffers are also being cut.
+
+With 64-bit XID's wraparound is effectively postponed to a very distant
+future. Even in highly loaded systems that had 2^32 transactions per day
+it will take huge 2^31 days before the first enforced "vacuum to prevent
+wraparound"). Buffers cutting and routine vacuum are not enforced, and DBA
+can plan them independently at the time with the least system load and least
+critical for database performance. Also, it can be done less frequently
+(several times a year vs every several days) on systems with transaction rates
+similar to those mentioned above.
+
+On-disk tuple and page format
+-----------------------------
+
+On-disk tuple format remains unchanged. 32-bit t_xmin and t_xmax store the
+lower parts of 64-bit XMIN and XMAX values. Each heap page has additional
+64-bit pd_xid_base and pd_multi_base which are common for all tuples on a page.
+They are placed into a pd_special area - 16 bytes in the end of a heap page.
+Actual XMIN/XMAX for a tuple are calculated upon reading a tuple from a page
+as follows:
+
+XMIN = t_xmin + pd_xid_base. 					(1)
+XMAX = t_xmax + pd_xid_base/pd_multi_base.		(2)
+
+"Double XMAX" page format
+---------------------------------
+
+At first read of a heap page after pg_upgrade from 32-bit XID PostgreSQL
+version pd_special area with a size of 16 bytes should be added to a page.
+Though a page may not have space for this. Then it can be converted to a
+temporary format called "double XMAX".
+
+All tuples after pg-upgrade would necessarily have xmin = FrozenTransactionId.
+So we don't need tuple header t_xmin field and we reuse t_xmin to store higher
+32 bits of its XMAX.
+
+Double XMAX format is only for full pages that don't have 16 bytes for
+pd_special. So it neither has a�place for a single tuple. Insert and HOT update
+for double XMAX pages is impossible and not supported. We can only read or
+delete tuples from it.
+
+When we are able to prune page double XMAX it will be converted from it to
+general 64-bit XID page format with all operations on its tuples supported.
+
+In-memory tuple format
+----------------------
+
+In-memory tuple representation consists of two parts:
+- HeapTupleHeader from disk page (contains all heap tuple contents, not only
+header)
+- HeapTuple with additional in-memory fields
+
+HeapTuple for each tuple in memory stores t_xid_base/t_multi_base - a copies of
+page's pd_xid_base/pd_multi_base. With tuple's 32-bit t_xmin and t_xmax from
+HeapTupleHeader they are used to calculate actual 64-bit XMIN and XMAX:
+
+XMIN = t_xmin + t_xid_base. 					(3)
+XMAX = t_xmax + t_xid_base/t_multi_base.		(4)
+
+The downside of this is that we can not use tuple's XMIN and XMAX right away.
+We often need to re-read t_xmin and t_xmax - which could actually be pointers
+into a page in shared buffers and therefore they could be updated by any other
+backend.
+
+Update/delete with 64-bit XIDs and 32-bit t_xmin/t_xmax
+--------------------------------------------------------------
+
+When we try to delete/update a tuple, we check that XMAX for a page fits (2).
+I.e. that t_xmax will not be over MaxShortTransactionId relative to
+pd_xid_base/pd_multi_base of a its page.
+
+If the current XID doesn't fit a range
+(pd_xid_base, pd_xid_base + MaxShortTransactionId) (5):
+
+- heap_page_prepare_for_xid() will try to increase pd_xid_base/pd_multi_base on
+a page and update all t_xmin/t_xmax of the other tuples on the page to
+correspond new pd_xid_base/pd_multi_base.
+
+- If it was impossible, it will try to prune and freeze tuples on a page.
+
+- If this is unsuccessful it will throw an error. Normally this is very
+unlikely but if there is a very old living transaction with an age of around
+2^32 this can arise. Basically, this is a behavior similar to one during the
+vacuum to prevent wraparound when XID was 32-bit. Dba should take care and
+avoid very-long-living transactions with an age close to 2^32. So long-living
+transactions often they are most likely defunct.
+
+Insert with 64-bit XIDs and 32-bit t_xmin/t_xmax
+------------------------------------------------
+
+On insert we check if current XID fits a range (5). Otherwise:
+
+- heap_page_prepare_for_xid() will try to increase pd_xid_base for t_xmin will
+not be over MaxShortTransactionId.
+
+- If it is impossible, then it will try to prune and freeze tuples on a page.
+
+Known issue: if pd_xid_base could not be shifted to accommodate a tuple being
+inserted due to a very long-running transaction, we just throw an error. We
+neither try to insert a�tuple into another page nor mark the current page as
+full. So, in this (unlikely) case we will get regular insert errors on the next
+tries to insert to the page 'locked' by this very long-running transaction.
+
+Upgrade from 32-bit XID versions
+--------------------------------
+
+pg_upgrade doesn't change pages format itself. It is done lazily after.
+
+1. At first heap page read, tuples on a page are repacked to free 16 bytes
+at the end of a page, possibly freeing space from dead tuples.
+
+2A. 16 bytes of pd_special is added if there is a place for it
+
+2B. Page is converted to "Double XMAX" format if there is no place for
+pd_special
+
+3. If a page is in double XMAX format after its first read, and vacuum (or
+micro-vacuum at select query) could prune some tuples and free space for
+pd_special, prune_page will add pd_special and convert page from double XMAX
+to general 64-bit XID page format.
-- 
2.24.3 (Apple Git-128)


--cpok4wp6gsarlzvp--





^ permalink  raw  reply  [nested|flat] 265+ messages in thread

* [PATCH v6] README for 64bit xid
@ 2022-01-10 19:20 Pavel Borisov <[email protected]>
  0 siblings, 0 replies; 265+ messages in thread

From: Pavel Borisov @ 2022-01-10 19:20 UTC (permalink / raw)

Authors:
- Pavel Borisov <[email protected]>
- Maxim Orlov <[email protected]>
- Yura Sokolov <[email protected]> <[email protected]>
---
 src/backend/access/heap/README.XID64 | 128 +++++++++++++++++++++++++++
 1 file changed, 128 insertions(+)
 create mode 100644 src/backend/access/heap/README.XID64

diff --git a/src/backend/access/heap/README.XID64 b/src/backend/access/heap/README.XID64
new file mode 100644
index 00000000000..457ba9b9ef5
--- /dev/null
+++ b/src/backend/access/heap/README.XID64
@@ -0,0 +1,128 @@
+src/backend/access/heap/README.XID64
+
+64-bit Transaction ID's (XID)
+=============================
+
+A limited number (N = 2^32) of XID's required to do vacuum freeze to prevent
+wraparound every N/2 transactions. This causes performance degradation due
+to the need to exclusively lock tables while being vacuumed. In each
+wraparound cycle, SLRU buffers are also being cut.
+
+With 64-bit XID's wraparound is effectively postponed to a very distant
+future. Even in highly loaded systems that had 2^32 transactions per day
+it will take huge 2^31 days before the first enforced "vacuum to prevent
+wraparound"). Buffers cutting and routine vacuum are not enforced, and DBA
+can plan them independently at the time with the least system load and least
+critical for database performance. Also, it can be done less frequently
+(several times a year vs every several days) on systems with transaction rates
+similar to those mentioned above.
+
+On-disk tuple and page format
+-----------------------------
+
+On-disk tuple format remains unchanged. 32-bit t_xmin and t_xmax store the
+lower parts of 64-bit XMIN and XMAX values. Each heap page has additional
+64-bit pd_xid_base and pd_multi_base which are common for all tuples on a page.
+They are placed into a pd_special area - 16 bytes in the end of a heap page.
+Actual XMIN/XMAX for a tuple are calculated upon reading a tuple from a page
+as follows:
+
+XMIN = t_xmin + pd_xid_base. 					(1)
+XMAX = t_xmax + pd_xid_base/pd_multi_base.		(2)
+
+"Double XMAX" page format
+---------------------------------
+
+At first read of a heap page after pg_upgrade from 32-bit XID PostgreSQL
+version pd_special area with a size of 16 bytes should be added to a page.
+Though a page may not have space for this. Then it can be converted to a
+temporary format called "double XMAX".
+
+All tuples after pg-upgrade would necessarily have xmin = FrozenTransactionId.
+So we don't need tuple header t_xmin field and we reuse t_xmin to store higher
+32 bits of its XMAX.
+
+Double XMAX format is only for full pages that don't have 16 bytes for
+pd_special. So it neither has a�place for a single tuple. Insert and HOT update
+for double XMAX pages is impossible and not supported. We can only read or
+delete tuples from it.
+
+When we are able to prune page double XMAX it will be converted from it to
+general 64-bit XID page format with all operations on its tuples supported.
+
+In-memory tuple format
+----------------------
+
+In-memory tuple representation consists of two parts:
+- HeapTupleHeader from disk page (contains all heap tuple contents, not only
+header)
+- HeapTuple with additional in-memory fields
+
+HeapTuple for each tuple in memory stores t_xid_base/t_multi_base - a copies of
+page's pd_xid_base/pd_multi_base. With tuple's 32-bit t_xmin and t_xmax from
+HeapTupleHeader they are used to calculate actual 64-bit XMIN and XMAX:
+
+XMIN = t_xmin + t_xid_base. 					(3)
+XMAX = t_xmax + t_xid_base/t_multi_base.		(4)
+
+The downside of this is that we can not use tuple's XMIN and XMAX right away.
+We often need to re-read t_xmin and t_xmax - which could actually be pointers
+into a page in shared buffers and therefore they could be updated by any other
+backend.
+
+Update/delete with 64-bit XIDs and 32-bit t_xmin/t_xmax
+--------------------------------------------------------------
+
+When we try to delete/update a tuple, we check that XMAX for a page fits (2).
+I.e. that t_xmax will not be over MaxShortTransactionId relative to
+pd_xid_base/pd_multi_base of a its page.
+
+If the current XID doesn't fit a range
+(pd_xid_base, pd_xid_base + MaxShortTransactionId) (5):
+
+- heap_page_prepare_for_xid() will try to increase pd_xid_base/pd_multi_base on
+a page and update all t_xmin/t_xmax of the other tuples on the page to
+correspond new pd_xid_base/pd_multi_base.
+
+- If it was impossible, it will try to prune and freeze tuples on a page.
+
+- If this is unsuccessful it will throw an error. Normally this is very
+unlikely but if there is a very old living transaction with an age of around
+2^32 this can arise. Basically, this is a behavior similar to one during the
+vacuum to prevent wraparound when XID was 32-bit. Dba should take care and
+avoid very-long-living transactions with an age close to 2^32. So long-living
+transactions often they are most likely defunct.
+
+Insert with 64-bit XIDs and 32-bit t_xmin/t_xmax
+------------------------------------------------
+
+On insert we check if current XID fits a range (5). Otherwise:
+
+- heap_page_prepare_for_xid() will try to increase pd_xid_base for t_xmin will
+not be over MaxShortTransactionId.
+
+- If it is impossible, then it will try to prune and freeze tuples on a page.
+
+Known issue: if pd_xid_base could not be shifted to accommodate a tuple being
+inserted due to a very long-running transaction, we just throw an error. We
+neither try to insert a�tuple into another page nor mark the current page as
+full. So, in this (unlikely) case we will get regular insert errors on the next
+tries to insert to the page 'locked' by this very long-running transaction.
+
+Upgrade from 32-bit XID versions
+--------------------------------
+
+pg_upgrade doesn't change pages format itself. It is done lazily after.
+
+1. At first heap page read, tuples on a page are repacked to free 16 bytes
+at the end of a page, possibly freeing space from dead tuples.
+
+2A. 16 bytes of pd_special is added if there is a place for it
+
+2B. Page is converted to "Double XMAX" format if there is no place for
+pd_special
+
+3. If a page is in double XMAX format after its first read, and vacuum (or
+micro-vacuum at select query) could prune some tuples and free space for
+pd_special, prune_page will add pd_special and convert page from double XMAX
+to general 64-bit XID page format.
-- 
2.24.3 (Apple Git-128)


--cpok4wp6gsarlzvp--





^ permalink  raw  reply  [nested|flat] 265+ messages in thread