agora inbox for [email protected]  
help / color / mirror / Atom feed
[PATCH v5 1/2] Make more clear the computation of min/max IO..
4+ messages / 2 participants
[nested] [flat]

* [PATCH v5 1/2] Make more clear the computation of min/max IO..
@ 2020-01-09 01:23 Justin Pryzby <[email protected]>
  0 siblings, 0 replies; 4+ messages in thread

From: Justin Pryzby @ 2020-01-09 01:23 UTC (permalink / raw)

..and specifically the double use and effect of correlation.

Avoid re-use of the "pages_fetched" variable
---
 src/backend/optimizer/path/costsize.c | 47 ++++++++++++++-------------
 1 file changed, 25 insertions(+), 22 deletions(-)

diff --git a/src/backend/optimizer/path/costsize.c b/src/backend/optimizer/path/costsize.c
index b5a0033721..bdc23a075f 100644
--- a/src/backend/optimizer/path/costsize.c
+++ b/src/backend/optimizer/path/costsize.c
@@ -491,12 +491,13 @@ cost_index(IndexPath *path, PlannerInfo *root, double loop_count,
 				csquared;
 	double		spc_seq_page_cost,
 				spc_random_page_cost;
-	Cost		min_IO_cost,
+	double		min_pages_fetched,	/* The min and max page count based on index correlation */
+				max_pages_fetched;
+	Cost		min_IO_cost,	/* The min and max cost based on index correlation */
 				max_IO_cost;
 	QualCost	qpqual_cost;
 	Cost		cpu_per_tuple;
 	double		tuples_fetched;
-	double		pages_fetched;
 	double		rand_heap_pages;
 	double		index_pages;
 
@@ -579,7 +580,8 @@ cost_index(IndexPath *path, PlannerInfo *root, double loop_count,
 	 * (just after a CLUSTER, for example), the number of pages fetched should
 	 * be exactly selectivity * table_size.  What's more, all but the first
 	 * will be sequential fetches, not the random fetches that occur in the
-	 * uncorrelated case.  So if the number of pages is more than 1, we
+	 * uncorrelated case (the index is expected to read fewer pages, *and* each
+	 * page read is cheaper).  So if the number of pages is more than 1, we
 	 * ought to charge
 	 *		spc_random_page_cost + (pages_fetched - 1) * spc_seq_page_cost
 	 * For partially-correlated indexes, we ought to charge somewhere between
@@ -604,17 +606,17 @@ cost_index(IndexPath *path, PlannerInfo *root, double loop_count,
 		 * pro-rate the costs for one scan.  In this case we assume all the
 		 * fetches are random accesses.
 		 */
-		pages_fetched = index_pages_fetched(tuples_fetched * loop_count,
+		max_pages_fetched = index_pages_fetched(tuples_fetched * loop_count,
 											baserel->pages,
 											(double) index->pages,
 											root);
 
 		if (indexonly)
-			pages_fetched = ceil(pages_fetched * (1.0 - baserel->allvisfrac));
+			max_pages_fetched = ceil(max_pages_fetched * (1.0 - baserel->allvisfrac));
 
-		rand_heap_pages = pages_fetched;
+		rand_heap_pages = max_pages_fetched;
 
-		max_IO_cost = (pages_fetched * spc_random_page_cost) / loop_count;
+		max_IO_cost = (max_pages_fetched * spc_random_page_cost) / loop_count;
 
 		/*
 		 * In the perfectly correlated case, the number of pages touched by
@@ -626,17 +628,17 @@ cost_index(IndexPath *path, PlannerInfo *root, double loop_count,
 		 * where such a plan is actually interesting, only one page would get
 		 * fetched per scan anyway, so it shouldn't matter much.)
 		 */
-		pages_fetched = ceil(indexSelectivity * (double) baserel->pages);
+		min_pages_fetched = ceil(indexSelectivity * (double) baserel->pages);
 
-		pages_fetched = index_pages_fetched(pages_fetched * loop_count,
+		min_pages_fetched = index_pages_fetched(min_pages_fetched * loop_count,
 											baserel->pages,
 											(double) index->pages,
 											root);
 
 		if (indexonly)
-			pages_fetched = ceil(pages_fetched * (1.0 - baserel->allvisfrac));
+			min_pages_fetched = ceil(min_pages_fetched * (1.0 - baserel->allvisfrac));
 
-		min_IO_cost = (pages_fetched * spc_random_page_cost) / loop_count;
+		min_IO_cost = (min_pages_fetched * spc_random_page_cost) / loop_count;
 	}
 	else
 	{
@@ -644,30 +646,31 @@ cost_index(IndexPath *path, PlannerInfo *root, double loop_count,
 		 * Normal case: apply the Mackert and Lohman formula, and then
 		 * interpolate between that and the correlation-derived result.
 		 */
-		pages_fetched = index_pages_fetched(tuples_fetched,
+
+		/* For the perfectly uncorrelated case (csquared=0) */
+		max_pages_fetched = index_pages_fetched(tuples_fetched,
 											baserel->pages,
 											(double) index->pages,
 											root);
 
 		if (indexonly)
-			pages_fetched = ceil(pages_fetched * (1.0 - baserel->allvisfrac));
+			max_pages_fetched = ceil(max_pages_fetched * (1.0 - baserel->allvisfrac));
 
-		rand_heap_pages = pages_fetched;
+		rand_heap_pages = max_pages_fetched;
 
-		/* max_IO_cost is for the perfectly uncorrelated case (csquared=0) */
-		max_IO_cost = pages_fetched * spc_random_page_cost;
+		max_IO_cost = max_pages_fetched * spc_random_page_cost;
 
-		/* min_IO_cost is for the perfectly correlated case (csquared=1) */
-		pages_fetched = ceil(indexSelectivity * (double) baserel->pages);
+		/* For the perfectly correlated case (csquared=1) */
+		min_pages_fetched = ceil(indexSelectivity * (double) baserel->pages);
 
 		if (indexonly)
-			pages_fetched = ceil(pages_fetched * (1.0 - baserel->allvisfrac));
+			min_pages_fetched = ceil(min_pages_fetched * (1.0 - baserel->allvisfrac));
 
-		if (pages_fetched > 0)
+		if (min_pages_fetched > 0)
 		{
 			min_IO_cost = spc_random_page_cost;
-			if (pages_fetched > 1)
-				min_IO_cost += (pages_fetched - 1) * spc_seq_page_cost;
+			if (min_pages_fetched > 1)
+				min_IO_cost += (min_pages_fetched - 1) * spc_seq_page_cost;
 		}
 		else
 			min_IO_cost = 0;
-- 
2.17.0


--16qp2B0xu0fRvRD7
Content-Type: text/x-diff; charset=us-ascii
Content-Disposition: attachment;
 filename="v5-0002-Use-correlation-statistic-in-costing-bitmap-scans.patch"



^ permalink  raw  reply  [nested|flat] 4+ messages in thread

* [PATCH v6 1/2] Make more clear the computation of min/max IO..
@ 2020-01-09 01:23 Justin Pryzby <[email protected]>
  0 siblings, 0 replies; 4+ messages in thread

From: Justin Pryzby @ 2020-01-09 01:23 UTC (permalink / raw)

..and specifically the double use and effect of correlation.

Avoid re-use of the "pages_fetched" variable
---
 src/backend/optimizer/path/costsize.c | 47 ++++++++++++++-------------
 1 file changed, 25 insertions(+), 22 deletions(-)

diff --git a/src/backend/optimizer/path/costsize.c b/src/backend/optimizer/path/costsize.c
index f1dfdc1a4a..083448def7 100644
--- a/src/backend/optimizer/path/costsize.c
+++ b/src/backend/optimizer/path/costsize.c
@@ -503,12 +503,13 @@ cost_index(IndexPath *path, PlannerInfo *root, double loop_count,
 				csquared;
 	double		spc_seq_page_cost,
 				spc_random_page_cost;
-	Cost		min_IO_cost,
+	double		min_pages_fetched,	/* The min and max page count based on index correlation */
+				max_pages_fetched;
+	Cost		min_IO_cost,	/* The min and max cost based on index correlation */
 				max_IO_cost;
 	QualCost	qpqual_cost;
 	Cost		cpu_per_tuple;
 	double		tuples_fetched;
-	double		pages_fetched;
 	double		rand_heap_pages;
 	double		index_pages;
 
@@ -591,7 +592,8 @@ cost_index(IndexPath *path, PlannerInfo *root, double loop_count,
 	 * (just after a CLUSTER, for example), the number of pages fetched should
 	 * be exactly selectivity * table_size.  What's more, all but the first
 	 * will be sequential fetches, not the random fetches that occur in the
-	 * uncorrelated case.  So if the number of pages is more than 1, we
+	 * uncorrelated case (the index is expected to read fewer pages, *and* each
+	 * page read is cheaper).  So if the number of pages is more than 1, we
 	 * ought to charge
 	 *		spc_random_page_cost + (pages_fetched - 1) * spc_seq_page_cost
 	 * For partially-correlated indexes, we ought to charge somewhere between
@@ -616,17 +618,17 @@ cost_index(IndexPath *path, PlannerInfo *root, double loop_count,
 		 * pro-rate the costs for one scan.  In this case we assume all the
 		 * fetches are random accesses.
 		 */
-		pages_fetched = index_pages_fetched(tuples_fetched * loop_count,
+		max_pages_fetched = index_pages_fetched(tuples_fetched * loop_count,
 											baserel->pages,
 											(double) index->pages,
 											root);
 
 		if (indexonly)
-			pages_fetched = ceil(pages_fetched * (1.0 - baserel->allvisfrac));
+			max_pages_fetched = ceil(max_pages_fetched * (1.0 - baserel->allvisfrac));
 
-		rand_heap_pages = pages_fetched;
+		rand_heap_pages = max_pages_fetched;
 
-		max_IO_cost = (pages_fetched * spc_random_page_cost) / loop_count;
+		max_IO_cost = (max_pages_fetched * spc_random_page_cost) / loop_count;
 
 		/*
 		 * In the perfectly correlated case, the number of pages touched by
@@ -638,17 +640,17 @@ cost_index(IndexPath *path, PlannerInfo *root, double loop_count,
 		 * where such a plan is actually interesting, only one page would get
 		 * fetched per scan anyway, so it shouldn't matter much.)
 		 */
-		pages_fetched = ceil(indexSelectivity * (double) baserel->pages);
+		min_pages_fetched = ceil(indexSelectivity * (double) baserel->pages);
 
-		pages_fetched = index_pages_fetched(pages_fetched * loop_count,
+		min_pages_fetched = index_pages_fetched(min_pages_fetched * loop_count,
 											baserel->pages,
 											(double) index->pages,
 											root);
 
 		if (indexonly)
-			pages_fetched = ceil(pages_fetched * (1.0 - baserel->allvisfrac));
+			min_pages_fetched = ceil(min_pages_fetched * (1.0 - baserel->allvisfrac));
 
-		min_IO_cost = (pages_fetched * spc_random_page_cost) / loop_count;
+		min_IO_cost = (min_pages_fetched * spc_random_page_cost) / loop_count;
 	}
 	else
 	{
@@ -656,30 +658,31 @@ cost_index(IndexPath *path, PlannerInfo *root, double loop_count,
 		 * Normal case: apply the Mackert and Lohman formula, and then
 		 * interpolate between that and the correlation-derived result.
 		 */
-		pages_fetched = index_pages_fetched(tuples_fetched,
+
+		/* For the perfectly uncorrelated case (csquared=0) */
+		max_pages_fetched = index_pages_fetched(tuples_fetched,
 											baserel->pages,
 											(double) index->pages,
 											root);
 
 		if (indexonly)
-			pages_fetched = ceil(pages_fetched * (1.0 - baserel->allvisfrac));
+			max_pages_fetched = ceil(max_pages_fetched * (1.0 - baserel->allvisfrac));
 
-		rand_heap_pages = pages_fetched;
+		rand_heap_pages = max_pages_fetched;
 
-		/* max_IO_cost is for the perfectly uncorrelated case (csquared=0) */
-		max_IO_cost = pages_fetched * spc_random_page_cost;
+		max_IO_cost = max_pages_fetched * spc_random_page_cost;
 
-		/* min_IO_cost is for the perfectly correlated case (csquared=1) */
-		pages_fetched = ceil(indexSelectivity * (double) baserel->pages);
+		/* For the perfectly correlated case (csquared=1) */
+		min_pages_fetched = ceil(indexSelectivity * (double) baserel->pages);
 
 		if (indexonly)
-			pages_fetched = ceil(pages_fetched * (1.0 - baserel->allvisfrac));
+			min_pages_fetched = ceil(min_pages_fetched * (1.0 - baserel->allvisfrac));
 
-		if (pages_fetched > 0)
+		if (min_pages_fetched > 0)
 		{
 			min_IO_cost = spc_random_page_cost;
-			if (pages_fetched > 1)
-				min_IO_cost += (pages_fetched - 1) * spc_seq_page_cost;
+			if (min_pages_fetched > 1)
+				min_IO_cost += (min_pages_fetched - 1) * spc_seq_page_cost;
 		}
 		else
 			min_IO_cost = 0;
-- 
2.17.0


--VV4b6MQE+OnNyhkM
Content-Type: text/x-diff; charset=us-ascii
Content-Disposition: attachment;
 filename="v6-0002-Use-correlation-statistic-in-costing-bitmap-scans.patch"



^ permalink  raw  reply  [nested|flat] 4+ messages in thread

* [PATCH v4 1/2] Make more clear the computation of min/max IO..
@ 2020-01-09 01:23 Justin Pryzby <[email protected]>
  0 siblings, 0 replies; 4+ messages in thread

From: Justin Pryzby @ 2020-01-09 01:23 UTC (permalink / raw)

..and specifically the double use and effect of correlation.

Avoid re-use of the "pages_fetched" variable
---
 src/backend/optimizer/path/costsize.c | 47 +++++++++++++++++++----------------
 1 file changed, 25 insertions(+), 22 deletions(-)

diff --git a/src/backend/optimizer/path/costsize.c b/src/backend/optimizer/path/costsize.c
index b5a0033..bdc23a0 100644
--- a/src/backend/optimizer/path/costsize.c
+++ b/src/backend/optimizer/path/costsize.c
@@ -491,12 +491,13 @@ cost_index(IndexPath *path, PlannerInfo *root, double loop_count,
 				csquared;
 	double		spc_seq_page_cost,
 				spc_random_page_cost;
-	Cost		min_IO_cost,
+	double		min_pages_fetched,	/* The min and max page count based on index correlation */
+				max_pages_fetched;
+	Cost		min_IO_cost,	/* The min and max cost based on index correlation */
 				max_IO_cost;
 	QualCost	qpqual_cost;
 	Cost		cpu_per_tuple;
 	double		tuples_fetched;
-	double		pages_fetched;
 	double		rand_heap_pages;
 	double		index_pages;
 
@@ -579,7 +580,8 @@ cost_index(IndexPath *path, PlannerInfo *root, double loop_count,
 	 * (just after a CLUSTER, for example), the number of pages fetched should
 	 * be exactly selectivity * table_size.  What's more, all but the first
 	 * will be sequential fetches, not the random fetches that occur in the
-	 * uncorrelated case.  So if the number of pages is more than 1, we
+	 * uncorrelated case (the index is expected to read fewer pages, *and* each
+	 * page read is cheaper).  So if the number of pages is more than 1, we
 	 * ought to charge
 	 *		spc_random_page_cost + (pages_fetched - 1) * spc_seq_page_cost
 	 * For partially-correlated indexes, we ought to charge somewhere between
@@ -604,17 +606,17 @@ cost_index(IndexPath *path, PlannerInfo *root, double loop_count,
 		 * pro-rate the costs for one scan.  In this case we assume all the
 		 * fetches are random accesses.
 		 */
-		pages_fetched = index_pages_fetched(tuples_fetched * loop_count,
+		max_pages_fetched = index_pages_fetched(tuples_fetched * loop_count,
 											baserel->pages,
 											(double) index->pages,
 											root);
 
 		if (indexonly)
-			pages_fetched = ceil(pages_fetched * (1.0 - baserel->allvisfrac));
+			max_pages_fetched = ceil(max_pages_fetched * (1.0 - baserel->allvisfrac));
 
-		rand_heap_pages = pages_fetched;
+		rand_heap_pages = max_pages_fetched;
 
-		max_IO_cost = (pages_fetched * spc_random_page_cost) / loop_count;
+		max_IO_cost = (max_pages_fetched * spc_random_page_cost) / loop_count;
 
 		/*
 		 * In the perfectly correlated case, the number of pages touched by
@@ -626,17 +628,17 @@ cost_index(IndexPath *path, PlannerInfo *root, double loop_count,
 		 * where such a plan is actually interesting, only one page would get
 		 * fetched per scan anyway, so it shouldn't matter much.)
 		 */
-		pages_fetched = ceil(indexSelectivity * (double) baserel->pages);
+		min_pages_fetched = ceil(indexSelectivity * (double) baserel->pages);
 
-		pages_fetched = index_pages_fetched(pages_fetched * loop_count,
+		min_pages_fetched = index_pages_fetched(min_pages_fetched * loop_count,
 											baserel->pages,
 											(double) index->pages,
 											root);
 
 		if (indexonly)
-			pages_fetched = ceil(pages_fetched * (1.0 - baserel->allvisfrac));
+			min_pages_fetched = ceil(min_pages_fetched * (1.0 - baserel->allvisfrac));
 
-		min_IO_cost = (pages_fetched * spc_random_page_cost) / loop_count;
+		min_IO_cost = (min_pages_fetched * spc_random_page_cost) / loop_count;
 	}
 	else
 	{
@@ -644,30 +646,31 @@ cost_index(IndexPath *path, PlannerInfo *root, double loop_count,
 		 * Normal case: apply the Mackert and Lohman formula, and then
 		 * interpolate between that and the correlation-derived result.
 		 */
-		pages_fetched = index_pages_fetched(tuples_fetched,
+
+		/* For the perfectly uncorrelated case (csquared=0) */
+		max_pages_fetched = index_pages_fetched(tuples_fetched,
 											baserel->pages,
 											(double) index->pages,
 											root);
 
 		if (indexonly)
-			pages_fetched = ceil(pages_fetched * (1.0 - baserel->allvisfrac));
+			max_pages_fetched = ceil(max_pages_fetched * (1.0 - baserel->allvisfrac));
 
-		rand_heap_pages = pages_fetched;
+		rand_heap_pages = max_pages_fetched;
 
-		/* max_IO_cost is for the perfectly uncorrelated case (csquared=0) */
-		max_IO_cost = pages_fetched * spc_random_page_cost;
+		max_IO_cost = max_pages_fetched * spc_random_page_cost;
 
-		/* min_IO_cost is for the perfectly correlated case (csquared=1) */
-		pages_fetched = ceil(indexSelectivity * (double) baserel->pages);
+		/* For the perfectly correlated case (csquared=1) */
+		min_pages_fetched = ceil(indexSelectivity * (double) baserel->pages);
 
 		if (indexonly)
-			pages_fetched = ceil(pages_fetched * (1.0 - baserel->allvisfrac));
+			min_pages_fetched = ceil(min_pages_fetched * (1.0 - baserel->allvisfrac));
 
-		if (pages_fetched > 0)
+		if (min_pages_fetched > 0)
 		{
 			min_IO_cost = spc_random_page_cost;
-			if (pages_fetched > 1)
-				min_IO_cost += (pages_fetched - 1) * spc_seq_page_cost;
+			if (min_pages_fetched > 1)
+				min_IO_cost += (min_pages_fetched - 1) * spc_seq_page_cost;
 		}
 		else
 			min_IO_cost = 0;
-- 
2.7.4


--gBdJBemW82xJqIAr
Content-Type: text/x-diff; charset=us-ascii
Content-Disposition: attachment;
 filename="v4-0002-Use-correlation-statistic-in-costing-bitmap-scans.patch"



^ permalink  raw  reply  [nested|flat] 4+ messages in thread

* [PATCH v3 1/2] Test exposing bug when foreign table points to a partitioned table
@ 2025-07-15 10:45 Jehan-Guillaume de Rorthais <[email protected]>
  0 siblings, 0 replies; 4+ messages in thread

From: Jehan-Guillaume de Rorthais @ 2025-07-15 10:45 UTC (permalink / raw)

When a foreign table points to a partitioned table or an inheritance
parent on the foreign server, a non-direct DML can affect multiple
rows when only one row is intended to be affected. This
happens because postgres_fdw uses only ctid to identify a row to work
on. Though ctid uniquely identifies a row in a single table, in a
partitioned table or in an inheritance hierarchy, there can be be
multiple rows, in different partitions, with the same ctid. So
DML statement sent to the foreign server by postgres_fdw ends up
affecting more than one rows, only one of which is intended to be
affected.

This commit adds testcases to show the problem. A subsequent commit
would have a fix to the problem.

Author: Ashutosh Bapat <[email protected]>
Reviewed-by: Kyotaro Horiguchi <[email protected]>
Discussion: https://www.postgresql.org/message-id/flat/CAFjFpRfcgwsHRmpvoOK-GUQi-n8MgAS%2BOxcQo%3DaBDn1COywmcg%4...

Rebased by Jehan-Guillaume de Rorthais <[email protected]>
---
 .../postgres_fdw/expected/postgres_fdw.out    | 125 ++++++++++++++++++
 contrib/postgres_fdw/sql/postgres_fdw.sql     |  58 ++++++++
 2 files changed, 183 insertions(+)

diff --git a/contrib/postgres_fdw/expected/postgres_fdw.out b/contrib/postgres_fdw/expected/postgres_fdw.out
index 2185b42bb4f..62019eaa881 100644
--- a/contrib/postgres_fdw/expected/postgres_fdw.out
+++ b/contrib/postgres_fdw/expected/postgres_fdw.out
@@ -8954,6 +8954,131 @@ drop foreign table remt2;
 drop table loct1;
 drop table loct2;
 drop table parent;
+-- test DML statement on a foreign table pointing to an inheritance hierarchy
+-- on the remote server
+CREATE TABLE a(aa TEXT);
+ALTER TABLE a SET (autovacuum_enabled = 'false');
+CREATE TABLE b() INHERITS(a);
+ALTER TABLE b SET (autovacuum_enabled = 'false');
+INSERT INTO a(aa) VALUES('aaa');
+INSERT INTO b(aa) VALUES('bbb');
+CREATE FOREIGN TABLE fa (aa TEXT) SERVER loopback OPTIONS (table_name 'a');
+SELECT tableoid::regclass, ctid, * FROM fa;
+ tableoid | ctid  | aa  
+----------+-------+-----
+ fa       | (0,1) | aaa
+ fa       | (0,1) | bbb
+(2 rows)
+
+-- use random() so that DML statement is not pushed down to the foreign
+-- server
+EXPLAIN (VERBOSE, COSTS OFF)
+UPDATE fa SET aa = (CASE WHEN random() <= 1 THEN 'zzzz' ELSE NULL END) WHERE aa = 'aaa';
+                                                   QUERY PLAN                                                    
+-----------------------------------------------------------------------------------------------------------------
+ Update on public.fa
+   Remote SQL: UPDATE public.a SET aa = $2 WHERE ctid = $1
+   ->  Foreign Scan on public.fa
+         Output: CASE WHEN (random() <= '1'::double precision) THEN 'zzzz'::text ELSE NULL::text END, ctid, fa.*
+         Remote SQL: SELECT aa, ctid FROM public.a WHERE ((aa = 'aaa')) FOR UPDATE
+(5 rows)
+
+UPDATE fa SET aa = (CASE WHEN random() <= 1 THEN 'zzzz' ELSE NULL END) WHERE aa = 'aaa';
+SELECT tableoid::regclass, ctid, * FROM fa;
+ tableoid | ctid  |  aa  
+----------+-------+------
+ fa       | (0,2) | zzzz
+ fa       | (0,1) | bbb
+(2 rows)
+
+-- repopulate tables so that we have rows with same ctid
+TRUNCATE a, b;
+INSERT INTO a(aa) VALUES('aaa');
+INSERT INTO b(aa) VALUES('bbb');
+EXPLAIN (VERBOSE, COSTS OFF)
+DELETE FROM fa WHERE aa = (CASE WHEN random() <= 1 THEN 'aaa' ELSE 'bbb' END);
+                                                  QUERY PLAN                                                   
+---------------------------------------------------------------------------------------------------------------
+ Delete on public.fa
+   Remote SQL: DELETE FROM public.a WHERE ctid = $1
+   ->  Foreign Scan on public.fa
+         Output: ctid
+         Filter: (fa.aa = CASE WHEN (random() <= '1'::double precision) THEN 'aaa'::text ELSE 'bbb'::text END)
+         Remote SQL: SELECT aa, ctid FROM public.a FOR UPDATE
+(6 rows)
+
+DELETE FROM fa WHERE aa = (CASE WHEN random() <= 1 THEN 'aaa' ELSE 'bbb' END);
+SELECT tableoid::regclass, ctid, * FROM fa;
+ tableoid | ctid  | aa 
+----------+-------+-----
+ fa       | (0,1) | bbb
+(1 row)
+
+-- cleanup
+DROP FOREIGN TABLE fa;
+DROP TABLE a CASCADE;
+NOTICE:  drop cascades to table b
+-- ===================================================================
+-- test foreign table pointing to a remote partitioned table
+-- ===================================================================
+-- test DML statement on foreign table pointing to a foreign partitioned table
+CREATE TABLE plt (a int, b int) PARTITION BY LIST(a);
+CREATE TABLE plt_p1 PARTITION OF plt FOR VALUES IN (1);
+CREATE TABLE plt_p2 PARTITION OF plt FOR VALUES IN (2);
+INSERT INTO plt VALUES (1, 1), (2, 2);
+CREATE FOREIGN TABLE fplt (a int, b int) SERVER loopback OPTIONS (table_name 'plt');
+SELECT tableoid::regclass, ctid, * FROM fplt;
+ tableoid | ctid  | a | b 
+----------+-------+---+---
+ fplt     | (0,1) | 1 | 1
+ fplt     | (0,1) | 2 | 2
+(2 rows)
+
+-- use random() so that DML statement is not pushed down to the foreign
+-- server
+EXPLAIN (VERBOSE, COSTS OFF)
+UPDATE fplt SET b = (CASE WHEN random() <= 1 THEN 10 ELSE 20 END) WHERE a = 1;
+                                           QUERY PLAN                                            
+-------------------------------------------------------------------------------------------------
+ Update on public.fplt
+   Remote SQL: UPDATE public.plt SET b = $2 WHERE ctid = $1
+   ->  Foreign Scan on public.fplt
+         Output: CASE WHEN (random() <= '1'::double precision) THEN 10 ELSE 20 END, ctid, fplt.*
+         Remote SQL: SELECT a, b, ctid FROM public.plt WHERE ((a = 1)) FOR UPDATE
+(5 rows)
+
+UPDATE fplt SET b = (CASE WHEN random() <= 1 THEN 10 ELSE 20 END) WHERE a = 1;
+SELECT tableoid::regclass, ctid, * FROM fplt;
+ tableoid | ctid  | a | b  
+----------+-------+---+----
+ fplt     | (0,2) | 1 | 10
+ fplt     | (0,1) | 2 | 2
+(2 rows)
+
+-- repopulate partitioned table so that we have rows with same ctid
+TRUNCATE plt;
+INSERT INTO plt VALUES (1, 1), (2, 2);
+EXPLAIN (VERBOSE, COSTS OFF)
+DELETE FROM fplt WHERE a = (CASE WHEN random() <=  1 THEN 1 ELSE 10 END);
+                                         QUERY PLAN                                          
+---------------------------------------------------------------------------------------------
+ Delete on public.fplt
+   Remote SQL: DELETE FROM public.plt WHERE ctid = $1
+   ->  Foreign Scan on public.fplt
+         Output: ctid
+         Filter: (fplt.a = CASE WHEN (random() <= '1'::double precision) THEN 1 ELSE 10 END)
+         Remote SQL: SELECT a, ctid FROM public.plt FOR UPDATE
+(6 rows)
+
+DELETE FROM fplt WHERE a = (CASE WHEN random() <=  1 THEN 1 ELSE 10 END);
+SELECT tableoid::regclass, ctid, * FROM fplt;
+ tableoid | ctid  | a | b 
+----------+-------+---+---
+ fplt     | (0,1) | 2 | 2 
+(1 row)
+
+DROP TABLE plt;
+DROP FOREIGN TABLE fplt;
 -- ===================================================================
 -- test tuple routing for foreign-table partitions
 -- ===================================================================
diff --git a/contrib/postgres_fdw/sql/postgres_fdw.sql b/contrib/postgres_fdw/sql/postgres_fdw.sql
index e534b40de3c..ae3777b7edb 100644
--- a/contrib/postgres_fdw/sql/postgres_fdw.sql
+++ b/contrib/postgres_fdw/sql/postgres_fdw.sql
@@ -2494,6 +2494,64 @@ drop table loct1;
 drop table loct2;
 drop table parent;
 
+-- test DML statement on a foreign table pointing to an inheritance hierarchy
+-- on the remote server
+CREATE TABLE a(aa TEXT);
+ALTER TABLE a SET (autovacuum_enabled = 'false');
+CREATE TABLE b() INHERITS(a);
+ALTER TABLE b SET (autovacuum_enabled = 'false');
+INSERT INTO a(aa) VALUES('aaa');
+INSERT INTO b(aa) VALUES('bbb');
+CREATE FOREIGN TABLE fa (aa TEXT) SERVER loopback OPTIONS (table_name 'a');
+
+SELECT tableoid::regclass, ctid, * FROM fa;
+-- use random() so that DML statement is not pushed down to the foreign
+-- server
+EXPLAIN (VERBOSE, COSTS OFF)
+UPDATE fa SET aa = (CASE WHEN random() <= 1 THEN 'zzzz' ELSE NULL END) WHERE aa = 'aaa';
+UPDATE fa SET aa = (CASE WHEN random() <= 1 THEN 'zzzz' ELSE NULL END) WHERE aa = 'aaa';
+SELECT tableoid::regclass, ctid, * FROM fa;
+-- repopulate tables so that we have rows with same ctid
+TRUNCATE a, b;
+INSERT INTO a(aa) VALUES('aaa');
+INSERT INTO b(aa) VALUES('bbb');
+EXPLAIN (VERBOSE, COSTS OFF)
+DELETE FROM fa WHERE aa = (CASE WHEN random() <= 1 THEN 'aaa' ELSE 'bbb' END);
+DELETE FROM fa WHERE aa = (CASE WHEN random() <= 1 THEN 'aaa' ELSE 'bbb' END);
+SELECT tableoid::regclass, ctid, * FROM fa;
+
+-- cleanup
+DROP FOREIGN TABLE fa;
+DROP TABLE a CASCADE;
+
+-- ===================================================================
+-- test foreign table pointing to a remote partitioned table
+-- ===================================================================
+
+-- test DML statement on foreign table pointing to a foreign partitioned table
+CREATE TABLE plt (a int, b int) PARTITION BY LIST(a);
+CREATE TABLE plt_p1 PARTITION OF plt FOR VALUES IN (1);
+CREATE TABLE plt_p2 PARTITION OF plt FOR VALUES IN (2);
+INSERT INTO plt VALUES (1, 1), (2, 2);
+CREATE FOREIGN TABLE fplt (a int, b int) SERVER loopback OPTIONS (table_name 'plt');
+SELECT tableoid::regclass, ctid, * FROM fplt;
+-- use random() so that DML statement is not pushed down to the foreign
+-- server
+EXPLAIN (VERBOSE, COSTS OFF)
+UPDATE fplt SET b = (CASE WHEN random() <= 1 THEN 10 ELSE 20 END) WHERE a = 1;
+UPDATE fplt SET b = (CASE WHEN random() <= 1 THEN 10 ELSE 20 END) WHERE a = 1;
+SELECT tableoid::regclass, ctid, * FROM fplt;
+-- repopulate partitioned table so that we have rows with same ctid
+TRUNCATE plt;
+INSERT INTO plt VALUES (1, 1), (2, 2);
+EXPLAIN (VERBOSE, COSTS OFF)
+DELETE FROM fplt WHERE a = (CASE WHEN random() <=  1 THEN 1 ELSE 10 END);
+DELETE FROM fplt WHERE a = (CASE WHEN random() <=  1 THEN 1 ELSE 10 END);
+SELECT tableoid::regclass, ctid, * FROM fplt;
+
+DROP TABLE plt;
+DROP FOREIGN TABLE fplt;
+
 -- ===================================================================
 -- test tuple routing for foreign-table partitions
 -- ===================================================================
-- 
2.50.0


--MP_/4ZRYdF7Ah.pt5w65PuZdaPz
Content-Type: text/x-patch
Content-Transfer-Encoding: 7bit
Content-Disposition: attachment;
 filename=v3-0002-Error-out-if-one-iteration-of-non-direct-DML-affe.patch



^ permalink  raw  reply  [nested|flat] 4+ messages in thread


end of thread, other threads:[~2025-07-15 10:45 UTC | newest]

Thread overview: 4+ messages (download: mbox mbox.gz follow: Atom feed)
-- links below jump to the message on this page --
2020-01-09 01:23 [PATCH v5 1/2] Make more clear the computation of min/max IO.. Justin Pryzby <[email protected]>
2020-01-09 01:23 [PATCH v6 1/2] Make more clear the computation of min/max IO.. Justin Pryzby <[email protected]>
2020-01-09 01:23 [PATCH v4 1/2] Make more clear the computation of min/max IO.. Justin Pryzby <[email protected]>
2025-07-15 10:45 [PATCH v3 1/2] Test exposing bug when foreign table points to a partitioned table Jehan-Guillaume de Rorthais <[email protected]>

This inbox is served by agora; see mirroring instructions
for how to clone and mirror all data and code used for this inbox