Re: Data loss, vacuum, transaction wrap-around

public inbox for [email protected]  
help / color / mirror / Atom feed

Re: Data loss, vacuum, transaction wrap-around
56+ messages / 3 participants
[nested] [flat]

* Re: Data loss, vacuum, transaction wrap-around
@ 2005-02-19 17:30 Jürgen Cappel <[email protected]>
  2005-02-20 04:19 ` Re: Data loss, vacuum, transaction wrap-around [email protected]
  0 siblings, 1 reply; 56+ messages in thread

From: Jürgen Cappel @ 2005-02-19 17:30 UTC (permalink / raw)
  To: [email protected]; pgsql-hackers

[ Shrugs ] and looks at other database systems ...

CA has put Ingres into Open Source last year.

Very reliable system with a replicator worth looking at.

Just a thought.

-------- Ursprüngliche Nachricht --------
Betreff: Re: [HACKERS] Data loss, vacuum, transaction wrap-around
Datum: Sat, 19 Feb 2005 09:15:14 -0500 (EST)
Von: [email protected]
An: "Tom Lane" <[email protected]>
CC: [email protected], "Russell Smith" <[email protected]>,	[email protected]
Referenzen: <[email protected]>    <[email protected]> 
<[email protected]>    <[email protected]> 
<[email protected]>

 > [email protected] writes:
 >> I think there should be a 100% no data loss fail safe.

OK, maybe I was overly broad in my statement, but I assumed a context that
I guess you missed. Don't you think that in normal operations, i.e. with
no hardware of OS failure, we should see any data loss as unacceptable?

If a bug causes data loss, it is a big deal right?
 >
 > Possibly we need to recalibrate our expectations here.  The current
 > situation is that PostgreSQL will not lose data if:
 >
 > 	1. Your disk drive doesn't screw up (eg, lie about write complete,
 > 	   or just plain die on you).
 > 	2. Your kernel and filesystem don't screw up.
 > 	3. You follow the instructions about routine vacuuming.
 > 	4. You don't hit any bugs that we don't know about.
 >

See, here is where I strongly disagree.Items (1) and (2) are completely
out of our control and no one would blame PostgreSQL.

Item (4) is an issue with all software, every now and then people hit bugs
and the bug is reported and assumed to get fixed.

Item (3) is just nasty, RTFM or else sucka! I think it is a very user
hostile stance.

 > I agree that it's a nice idea to be able to eliminate assumption #3 from
 > our list of gotchas, but the big picture is that it's hard to believe
 > that doing this will make for a quantum jump in the overall level of
 > reliability.  I think I listed the risks in roughly the right order of
 > severity ...

Sometimes the edge conditions of a problem are not so obscure. I think (3)
is a huge issue, iamgine I'm in this meeting:

DBA: We can't use PostgreSQL, if we forget to do normal maintenence we'll
lose all our data.

ME: Well, there is an amount of truth in that, but we just won't forget.

DBA: Sorry, I don't trust it.

CTO: Mark, I think joe has some serious issues that need to be resolved
before we can move on this.

Boom!! Lost.

 >
 > I'm willing to fix this for 8.1 (and am already in process of drafting a
 > patch), especially since it ties into some other known problems such as
 > the pg_pwd/pg_group files not being properly reconstructed after PITR
 > recovery.  But I think that a "Chinese fire drill" is not called for,
 > and backpatching a significant but poorly tested change falls into that
 > category IMHO.
 >
 > 			regards, tom lane
 >
 > ---------------------------(end of broadcast)---------------------------
 > TIP 7: don't forget to increase your free space map settings
 >

---------------------------(end of broadcast)---------------------------
TIP 1: subscribe and unsubscribe commands go to [email protected]

^ permalink  raw  reply  [nested|flat] 56+ messages in thread

* Re: Data loss, vacuum, transaction wrap-around
  2005-02-19 17:30 Re: Data loss, vacuum, transaction wrap-around Jürgen Cappel <[email protected]>
@ 2005-02-20 04:19 ` [email protected]
  0 siblings, 0 replies; 56+ messages in thread

From: [email protected] @ 2005-02-20 04:19 UTC (permalink / raw)
  To: Jürgen Cappel <[email protected]>; +Cc: pgsql-hackers

> [ Shrugs ] and looks at other database systems ...
>
> CA has put Ingres into Open Source last year.
>
> Very reliable system with a replicator worth looking at.
>
> Just a thought.

The discussion on hackers is how to make PostgreSQL better. There are many
different perspectives, differences are argued and concensus reached, and
a better PostgreSQL emerges.

Going to another database would be counter productive to the process.





^ permalink  raw  reply  [nested|flat] 56+ messages in thread

* [PATCH 7/7] Make Append node async-aware.
@ 2016-06-28 09:52 Kyotaro Horiguchi <[email protected]>
  0 siblings, 0 replies; 56+ messages in thread

From: Kyotaro Horiguchi @ 2016-06-28 09:52 UTC (permalink / raw)

Change append node to be capable to handle asynchronous children
properly. As soon as it receives !async_ready from a child, it moves
to the next child and if no child is ready, it sleeps until at least
one of them become ready.
---
 src/backend/executor/nodeAppend.c | 67 +++++++++++++++++++++++++++++++++++++--
 src/include/nodes/execnodes.h     |  2 ++
 2 files changed, 67 insertions(+), 2 deletions(-)

diff --git a/src/backend/executor/nodeAppend.c b/src/backend/executor/nodeAppend.c
index e0ce8c6..9a4063a 100644
--- a/src/backend/executor/nodeAppend.c
+++ b/src/backend/executor/nodeAppend.c
@@ -58,6 +58,7 @@
 #include "postgres.h"
 
 #include "executor/execdebug.h"
+#include "executor/execAsync.h"
 #include "executor/nodeAppend.h"
 
 static bool exec_append_initialize_next(AppendState *appendstate);
@@ -121,6 +122,7 @@ ExecInitAppend(Append *node, EState *estate, int eflags)
 {
 	AppendState *appendstate = makeNode(AppendState);
 	PlanState **appendplanstates;
+	bool	   *finished;
 	int			nplans;
 	int			i;
 	ListCell   *lc;
@@ -134,14 +136,17 @@ ExecInitAppend(Append *node, EState *estate, int eflags)
 	nplans = list_length(node->appendplans);
 
 	appendplanstates = (PlanState **) palloc0(nplans * sizeof(PlanState *));
-
+	finished = (bool *) palloc0(nplans * sizeof(bool));
+	
 	/*
 	 * create new AppendState for our append node
 	 */
 	appendstate->ps.plan = (Plan *) node;
 	appendstate->ps.state = estate;
 	appendstate->appendplans = appendplanstates;
+	appendstate->as_finished = finished;
 	appendstate->as_nplans = nplans;
+	appendstate->as_async = ((eflags & EXEC_FLAG_BACKWARD) == 0);
 
 	/*
 	 * Miscellaneous initialization
@@ -194,6 +199,8 @@ ExecInitAppend(Append *node, EState *estate, int eflags)
 void
 ExecAppend(AppendState *node)
 {
+	int stopplan = node->as_whichplan;
+
 	for (;;)
 	{
 		PlanState  *subnode;
@@ -207,7 +214,36 @@ ExecAppend(AppendState *node)
 		/*
 		 * get a tuple from the subplan
 		 */
-		result = ExecProcNode(subnode);
+		Assert(!node->as_finished[node->as_whichplan]);
+
+		if (!subnode->result_ready)
+			ExecExecuteNode(subnode);
+
+		if (!subnode->result_ready)
+		{
+			if (node->as_async)
+			{
+				/* Move to the next living node */
+				do
+				{
+					node->as_whichplan = 
+						(node->as_whichplan + 1) %  node->as_nplans;
+				} while (node->as_whichplan != stopplan &&
+						 node->as_finished[node->as_whichplan]);
+
+				/* No node is ready yet, return as not-ready */
+				if (node->as_whichplan == stopplan)
+					return;
+
+				/* Try the next node */
+				continue;
+			}
+
+			/* If not async, immediately wait for this subnode */
+			ExecAsyncWaitForNode(subnode);
+		}				
+
+		result = ExecConsumeResult((PlanState *) subnode);
 
 		if (!TupIsNull(result))
 		{
@@ -220,6 +256,31 @@ ExecAppend(AppendState *node)
 			return;
 		}
 
+		if (node->as_async)
+		{
+			node->as_finished[node->as_whichplan] = true;
+			stopplan = node->as_whichplan;
+
+			/* Find the next living subnode */
+			do
+			{
+				node->as_whichplan =
+					(node->as_whichplan + 1) % node->as_nplans;
+			} while (node->as_whichplan != stopplan &&
+					 node->as_finished[node->as_whichplan]);
+
+			if (node->as_whichplan != stopplan)
+			{
+				stopplan = node->as_whichplan;
+				continue;
+			}
+
+			/* All subnodes are exhausted. Finish this node. */
+			ExecReturnTuple(&node->ps,
+							ExecClearTuple(node->ps.ps_ResultTupleSlot));
+			return;
+		}
+
 		/*
 		 * Go on to the "next" subplan in the appropriate direction. If no
 		 * more subplans, return the empty slot set up for us by
@@ -277,6 +338,8 @@ ExecReScanAppend(AppendState *node)
 	{
 		PlanState  *subnode = node->appendplans[i];
 
+		node->as_finished[i] = false;
+
 		/*
 		 * ExecReScan doesn't know about my subplans, so I have to do
 		 * changed-parameter signaling myself.
diff --git a/src/include/nodes/execnodes.h b/src/include/nodes/execnodes.h
index b72decc..b0a86c5 100644
--- a/src/include/nodes/execnodes.h
+++ b/src/include/nodes/execnodes.h
@@ -1177,6 +1177,8 @@ typedef struct AppendState
 {
 	PlanState	ps;				/* its first field is NodeTag */
 	PlanState **appendplans;	/* array of PlanStates for my inputs */
+	bool		as_async;		/* true to allow async execution */
+	bool	   *as_finished;	/* array of the running state of subplans */
 	int			as_nplans;
 	int			as_whichplan;
 } AppendState;
-- 
2.9.2


----Next_Part(Mon_Aug_29_17_08_36_2016_213)--
Content-Type: text/plain
Content-Disposition: inline
Content-Transfer-Encoding: 8bit
MIME-Version: 1.0


-- 
Sent via pgsql-hackers mailing list ([email protected])
To make changes to your subscription:
http://www.postgresql.org/mailpref/pgsql-hackers

----Next_Part(Mon_Aug_29_17_08_36_2016_213)----






^ permalink  raw  reply  [nested|flat] 56+ messages in thread

* [PATCH 7/7] Make Append node async-aware.
@ 2016-06-28 09:52 Kyotaro Horiguchi <[email protected]>
  0 siblings, 0 replies; 56+ messages in thread

From: Kyotaro Horiguchi @ 2016-06-28 09:52 UTC (permalink / raw)

Change append node to be capable to handle asynchronous children
properly. As soon as it receives !async_ready from a child, it moves
to the next child and if no child is ready, it sleeps until at least
one of them become ready.
---
 src/backend/executor/nodeAppend.c | 94 ++++++++++++++++++++++++++++++++-------
 src/include/nodes/execnodes.h     |  2 +
 2 files changed, 80 insertions(+), 16 deletions(-)

diff --git a/src/backend/executor/nodeAppend.c b/src/backend/executor/nodeAppend.c
index e0ce8c6..1c0d26e 100644
--- a/src/backend/executor/nodeAppend.c
+++ b/src/backend/executor/nodeAppend.c
@@ -58,6 +58,7 @@
 #include "postgres.h"
 
 #include "executor/execdebug.h"
+#include "executor/execAsync.h"
 #include "executor/nodeAppend.h"
 
 static bool exec_append_initialize_next(AppendState *appendstate);
@@ -121,6 +122,7 @@ ExecInitAppend(Append *node, EState *estate, int eflags)
 {
 	AppendState *appendstate = makeNode(AppendState);
 	PlanState **appendplanstates;
+	bool	   *finished;
 	int			nplans;
 	int			i;
 	ListCell   *lc;
@@ -134,14 +136,17 @@ ExecInitAppend(Append *node, EState *estate, int eflags)
 	nplans = list_length(node->appendplans);
 
 	appendplanstates = (PlanState **) palloc0(nplans * sizeof(PlanState *));
-
+	finished = (bool *) palloc0(nplans * sizeof(bool));
+	
 	/*
 	 * create new AppendState for our append node
 	 */
 	appendstate->ps.plan = (Plan *) node;
 	appendstate->ps.state = estate;
 	appendstate->appendplans = appendplanstates;
+	appendstate->as_finished = finished;
 	appendstate->as_nplans = nplans;
+	appendstate->as_async = ((eflags & EXEC_FLAG_BACKWARD) == 0);
 
 	/*
 	 * Miscellaneous initialization
@@ -194,49 +199,104 @@ ExecInitAppend(Append *node, EState *estate, int eflags)
 void
 ExecAppend(AppendState *node)
 {
-	for (;;)
+	int n_notready = 0;
+	PlanState  *subnode;
+	int i, n;
+
+	n = node->as_whichplan;
+
+	for (i = 0 ; i < node->as_nplans ; i++)
 	{
-		PlanState  *subnode;
 		TupleTableSlot *result;
 
+		if (node->as_async)
+		{
+			if (n >= node->as_nplans)
+				n = 0;
+
+			if (node->as_finished[n])
+			{
+				n++;
+				continue;
+			}
+		}
+
 		/*
 		 * figure out which subplan we are currently processing
 		 */
-		subnode = node->appendplans[node->as_whichplan];
+		subnode = node->appendplans[n];
 
 		/*
-		 * get a tuple from the subplan
+		 * execute the subplan to get a result if it is not ready yet
 		 */
-		result = ExecProcNode(subnode);
+		if (!subnode->result_ready)
+			ExecExecuteNode(subnode);
+
+		/*
+		 * This subnode is not ready yet when asynchrony is not allowed,
+		 * immediately wait for this subnode.
+		 */
+		if (!subnode->result_ready)
+		{
+			if (node->as_async)
+			{
+				n_notready++;
+				n++;
+				continue;
+			}
+			ExecAsyncWaitForNode(subnode);
+		}
+
+		Assert(subnode->result_ready);
+
+		result = ExecConsumeResult((PlanState *)subnode);
 
 		if (!TupIsNull(result))
 		{
 			/*
-			 * If the subplan gave us something then return it as-is. We do
-			 * NOT make use of the result slot that was set up in
-			 * ExecInitAppend; there's no need for it.
+			 * If the subplan gave us something then return it
+			 * as-is. We do NOT make use of the result slot that was
+			 * set up in ExecInitAppend; there's no need for it.
 			 */
+			node->as_whichplan = n;
 			ExecReturnTuple(&node->ps, result);
 			return;
 		}
 
+		/* Tuple has been exhausted on this subnode */
+		if (node->as_async)
+		{
+			node->as_finished[n++] = true;
+			continue;
+		}
+
 		/*
 		 * Go on to the "next" subplan in the appropriate direction. If no
 		 * more subplans, return the empty slot set up for us by
 		 * ExecInitAppend.
 		 */
 		if (ScanDirectionIsForward(node->ps.state->es_direction))
-			node->as_whichplan++;
+		{
+			if (n >= node->as_nplans - 1)
+				break;
+			n++;
+		}
 		else
-			node->as_whichplan--;
-		if (!exec_append_initialize_next(node))
 		{
-			ExecReturnTuple(&node->ps,
-							ExecClearTuple(node->ps.ps_ResultTupleSlot));
-			return;
+			if (n == 0)
+				break;
+			n--;
 		}
+	}
 
-		/* Else loop back and try to get a tuple from the new subplan */
+	/*
+	 * We are finished if reached here and no subnodes are not-ready
+	 */
+	if (n_notready == 0)
+	{
+		node->as_whichplan = n;
+		ExecReturnTuple(&node->ps,
+						ExecClearTuple(node->ps.ps_ResultTupleSlot));
 	}
 }
 
@@ -277,6 +337,8 @@ ExecReScanAppend(AppendState *node)
 	{
 		PlanState  *subnode = node->appendplans[i];
 
+		node->as_finished[i] = false;
+
 		/*
 		 * ExecReScan doesn't know about my subplans, so I have to do
 		 * changed-parameter signaling myself.
diff --git a/src/include/nodes/execnodes.h b/src/include/nodes/execnodes.h
index 9121537..e4f2bb6 100644
--- a/src/include/nodes/execnodes.h
+++ b/src/include/nodes/execnodes.h
@@ -1170,6 +1170,8 @@ typedef struct AppendState
 {
 	PlanState	ps;				/* its first field is NodeTag */
 	PlanState **appendplans;	/* array of PlanStates for my inputs */
+	bool		as_async;		/* true to allow async execution */
+	bool	   *as_finished;	/* array of the running state of subplans */
 	int			as_nplans;
 	int			as_whichplan;
 } AppendState;
-- 
1.8.3.1


----Next_Part(Wed_Jul_06_16_29_14_2016_599)--
Content-Type: Text/Plain; charset=us-ascii
Content-Transfer-Encoding: 7bit
Content-Disposition: inline; filename="gentbl.sql"

CREATE EXTENSION postgres_fdw;
CREATE TABLE pl (a int, b int);
CREATE TABLE cl1 (LIKE pl) INHERITS (pl);
CREATE TABLE cl2 (LIKE pl) INHERITS (pl);
CREATE TABLE cl3 (LIKE pl) INHERITS (pl);
CREATE TABLE cl4 (LIKE pl) INHERITS (pl);
INSERT INTO cl1 (SELECT a, a FROM generate_series(0000000, 0999999) a);
INSERT INTO cl2 (SELECT a, a FROM generate_series(1000000, 1999999) a);
INSERT INTO cl3 (SELECT a, a FROM generate_series(2000000, 2999999) a);
INSERT INTO cl4 (SELECT a, a FROM generate_series(3000000, 3999999) a);

CREATE TABLE t0 (LIKE pl);
INSERT INTO t0 (SELECT a, a FROM generate_series(0000000, 9999999) a);

CREATE SERVER sv0 FOREIGN DATA WRAPPER postgres_fdw OPTIONS (host '/tmp', dbname 'postgres');
CREATE SERVER sv1 FOREIGN DATA WRAPPER postgres_fdw OPTIONS (host '/tmp', dbname 'postgres');
CREATE SERVER sv2 FOREIGN DATA WRAPPER postgres_fdw OPTIONS (host '/tmp', dbname 'postgres');
CREATE SERVER sv3 FOREIGN DATA WRAPPER postgres_fdw OPTIONS (host '/tmp', dbname 'postgres');
CREATE SERVER sv4 FOREIGN DATA WRAPPER postgres_fdw OPTIONS (host '/tmp', dbname 'postgres');
CREATE USER MAPPING FOR public SERVER sv0;
CREATE USER MAPPING FOR public SERVER sv1;
CREATE USER MAPPING FOR public SERVER sv2;
CREATE USER MAPPING FOR public SERVER sv3;
CREATE USER MAPPING FOR public SERVER sv4;

CREATE FOREIGN TABLE ft10 (a int, b int) SERVER sv0 OPTIONS (table_name 'cl1');
CREATE FOREIGN TABLE ft20 (a int, b int) SERVER sv0 OPTIONS (table_name 'cl2');
CREATE FOREIGN TABLE ft30 (a int, b int) SERVER sv0 OPTIONS (table_name 'cl3');
CREATE FOREIGN TABLE ft40 (a int, b int) SERVER sv0 OPTIONS (table_name 'cl4');
CREATE FOREIGN TABLE ft11 (a int, b int) SERVER sv1 OPTIONS (table_name 'cl1');
CREATE FOREIGN TABLE ft22 (a int, b int) SERVER sv2 OPTIONS (table_name 'cl2');
CREATE FOREIGN TABLE ft33 (a int, b int) SERVER sv3 OPTIONS (table_name 'cl3');
CREATE FOREIGN TABLE ft44 (a int, b int) SERVER sv4 OPTIONS (table_name 'cl4');

CREATE TABLE pf0 (LIKE pl);
ALTER FOREIGN TABLE ft10 INHERIT pf0;
ALTER FOREIGN TABLE ft20 INHERIT pf0;
ALTER FOREIGN TABLE ft30 INHERIT pf0;
ALTER FOREIGN TABLE ft40 INHERIT pf0;

CREATE table pf1 (LIKE pl);
ALTER FOREIGN TABLE ft11 INHERIT pf1;
ALTER FOREIGN TABLE ft22 INHERIT pf1;
ALTER FOREIGN TABLE ft33 INHERIT pf1;
ALTER FOREIGN TABLE ft44 INHERIT pf1;

ANALYZE;

----Next_Part(Wed_Jul_06_16_29_14_2016_599)--
Content-Type: Text/Plain; charset=us-ascii
Content-Transfer-Encoding: 7bit
Content-Disposition: inline; filename="testrun.sh"

#! /bin/bash

function do_test() {
	echo $1
	for i in $(seq 1 10); do psql postgres -c "set log_min_duration_statement = 0; set client_min_messages=log; select sum(a) from $1"; done | grep LOG
}

for i in "t0" "pl" "pf0" "pf1";
do
  do_test $i
done


----Next_Part(Wed_Jul_06_16_29_14_2016_599)--
Content-Type: text/plain
Content-Disposition: inline
Content-Transfer-Encoding: 8bit
MIME-Version: 1.0


-- 
Sent via pgsql-hackers mailing list ([email protected])
To make changes to your subscription:
http://www.postgresql.org/mailpref/pgsql-hackers

----Next_Part(Wed_Jul_06_16_29_14_2016_599)----






^ permalink  raw  reply  [nested|flat] 56+ messages in thread

* [PATCH 7/7] Make Append node async-aware.
@ 2016-06-28 09:52 Kyotaro Horiguchi <[email protected]>
  0 siblings, 0 replies; 56+ messages in thread

From: Kyotaro Horiguchi @ 2016-06-28 09:52 UTC (permalink / raw)

Change append node to be capable to handle asynchronous children
properly. As soon as it receives !async_ready from a child, it moves
to the next child and if no child is ready, it sleeps until at least
one of them become ready.
---
 src/backend/executor/nodeAppend.c | 67 +++++++++++++++++++++++++++++++++++++--
 src/include/nodes/execnodes.h     |  2 ++
 2 files changed, 67 insertions(+), 2 deletions(-)

diff --git a/src/backend/executor/nodeAppend.c b/src/backend/executor/nodeAppend.c
index e0ce8c6..9a4063a 100644
--- a/src/backend/executor/nodeAppend.c
+++ b/src/backend/executor/nodeAppend.c
@@ -58,6 +58,7 @@
 #include "postgres.h"
 
 #include "executor/execdebug.h"
+#include "executor/execAsync.h"
 #include "executor/nodeAppend.h"
 
 static bool exec_append_initialize_next(AppendState *appendstate);
@@ -121,6 +122,7 @@ ExecInitAppend(Append *node, EState *estate, int eflags)
 {
 	AppendState *appendstate = makeNode(AppendState);
 	PlanState **appendplanstates;
+	bool	   *finished;
 	int			nplans;
 	int			i;
 	ListCell   *lc;
@@ -134,14 +136,17 @@ ExecInitAppend(Append *node, EState *estate, int eflags)
 	nplans = list_length(node->appendplans);
 
 	appendplanstates = (PlanState **) palloc0(nplans * sizeof(PlanState *));
-
+	finished = (bool *) palloc0(nplans * sizeof(bool));
+	
 	/*
 	 * create new AppendState for our append node
 	 */
 	appendstate->ps.plan = (Plan *) node;
 	appendstate->ps.state = estate;
 	appendstate->appendplans = appendplanstates;
+	appendstate->as_finished = finished;
 	appendstate->as_nplans = nplans;
+	appendstate->as_async = ((eflags & EXEC_FLAG_BACKWARD) == 0);
 
 	/*
 	 * Miscellaneous initialization
@@ -194,6 +199,8 @@ ExecInitAppend(Append *node, EState *estate, int eflags)
 void
 ExecAppend(AppendState *node)
 {
+	int stopplan = node->as_whichplan;
+
 	for (;;)
 	{
 		PlanState  *subnode;
@@ -207,7 +214,36 @@ ExecAppend(AppendState *node)
 		/*
 		 * get a tuple from the subplan
 		 */
-		result = ExecProcNode(subnode);
+		Assert(!node->as_finished[node->as_whichplan]);
+
+		if (!subnode->result_ready)
+			ExecExecuteNode(subnode);
+
+		if (!subnode->result_ready)
+		{
+			if (node->as_async)
+			{
+				/* Move to the next living node */
+				do
+				{
+					node->as_whichplan = 
+						(node->as_whichplan + 1) %  node->as_nplans;
+				} while (node->as_whichplan != stopplan &&
+						 node->as_finished[node->as_whichplan]);
+
+				/* No node is ready yet, return as not-ready */
+				if (node->as_whichplan == stopplan)
+					return;
+
+				/* Try the next node */
+				continue;
+			}
+
+			/* If not async, immediately wait for this subnode */
+			ExecAsyncWaitForNode(subnode);
+		}				
+
+		result = ExecConsumeResult((PlanState *) subnode);
 
 		if (!TupIsNull(result))
 		{
@@ -220,6 +256,31 @@ ExecAppend(AppendState *node)
 			return;
 		}
 
+		if (node->as_async)
+		{
+			node->as_finished[node->as_whichplan] = true;
+			stopplan = node->as_whichplan;
+
+			/* Find the next living subnode */
+			do
+			{
+				node->as_whichplan =
+					(node->as_whichplan + 1) % node->as_nplans;
+			} while (node->as_whichplan != stopplan &&
+					 node->as_finished[node->as_whichplan]);
+
+			if (node->as_whichplan != stopplan)
+			{
+				stopplan = node->as_whichplan;
+				continue;
+			}
+
+			/* All subnodes are exhausted. Finish this node. */
+			ExecReturnTuple(&node->ps,
+							ExecClearTuple(node->ps.ps_ResultTupleSlot));
+			return;
+		}
+
 		/*
 		 * Go on to the "next" subplan in the appropriate direction. If no
 		 * more subplans, return the empty slot set up for us by
@@ -277,6 +338,8 @@ ExecReScanAppend(AppendState *node)
 	{
 		PlanState  *subnode = node->appendplans[i];
 
+		node->as_finished[i] = false;
+
 		/*
 		 * ExecReScan doesn't know about my subplans, so I have to do
 		 * changed-parameter signaling myself.
diff --git a/src/include/nodes/execnodes.h b/src/include/nodes/execnodes.h
index b72decc..b0a86c5 100644
--- a/src/include/nodes/execnodes.h
+++ b/src/include/nodes/execnodes.h
@@ -1177,6 +1177,8 @@ typedef struct AppendState
 {
 	PlanState	ps;				/* its first field is NodeTag */
 	PlanState **appendplans;	/* array of PlanStates for my inputs */
+	bool		as_async;		/* true to allow async execution */
+	bool	   *as_finished;	/* array of the running state of subplans */
 	int			as_nplans;
 	int			as_whichplan;
 } AppendState;
-- 
1.8.3.1


----Next_Part(Thu_Jul_21_18_50_07_2016_597)--
Content-Type: text/plain
Content-Disposition: inline
Content-Transfer-Encoding: 8bit
MIME-Version: 1.0


-- 
Sent via pgsql-hackers mailing list ([email protected])
To make changes to your subscription:
http://www.postgresql.org/mailpref/pgsql-hackers

----Next_Part(Thu_Jul_21_18_50_07_2016_597)----






^ permalink  raw  reply  [nested|flat] 56+ messages in thread

* [PATCH 7/7] Make Append node async-aware.
@ 2016-06-28 09:52 Kyotaro Horiguchi <[email protected]>
  0 siblings, 0 replies; 56+ messages in thread

From: Kyotaro Horiguchi @ 2016-06-28 09:52 UTC (permalink / raw)

Change append node to be capable to handle asynchronous children
properly. As soon as it receives !async_ready from a child, it moves
to the next child and if no child is ready, it sleeps until at least
one of them become ready.
---
 src/backend/executor/nodeAppend.c | 67 +++++++++++++++++++++++++++++++++++++--
 src/include/nodes/execnodes.h     |  2 ++
 2 files changed, 67 insertions(+), 2 deletions(-)

diff --git a/src/backend/executor/nodeAppend.c b/src/backend/executor/nodeAppend.c
index e0ce8c6..9a4063a 100644
--- a/src/backend/executor/nodeAppend.c
+++ b/src/backend/executor/nodeAppend.c
@@ -58,6 +58,7 @@
 #include "postgres.h"
 
 #include "executor/execdebug.h"
+#include "executor/execAsync.h"
 #include "executor/nodeAppend.h"
 
 static bool exec_append_initialize_next(AppendState *appendstate);
@@ -121,6 +122,7 @@ ExecInitAppend(Append *node, EState *estate, int eflags)
 {
 	AppendState *appendstate = makeNode(AppendState);
 	PlanState **appendplanstates;
+	bool	   *finished;
 	int			nplans;
 	int			i;
 	ListCell   *lc;
@@ -134,14 +136,17 @@ ExecInitAppend(Append *node, EState *estate, int eflags)
 	nplans = list_length(node->appendplans);
 
 	appendplanstates = (PlanState **) palloc0(nplans * sizeof(PlanState *));
-
+	finished = (bool *) palloc0(nplans * sizeof(bool));
+	
 	/*
 	 * create new AppendState for our append node
 	 */
 	appendstate->ps.plan = (Plan *) node;
 	appendstate->ps.state = estate;
 	appendstate->appendplans = appendplanstates;
+	appendstate->as_finished = finished;
 	appendstate->as_nplans = nplans;
+	appendstate->as_async = ((eflags & EXEC_FLAG_BACKWARD) == 0);
 
 	/*
 	 * Miscellaneous initialization
@@ -194,6 +199,8 @@ ExecInitAppend(Append *node, EState *estate, int eflags)
 void
 ExecAppend(AppendState *node)
 {
+	int stopplan = node->as_whichplan;
+
 	for (;;)
 	{
 		PlanState  *subnode;
@@ -207,7 +214,36 @@ ExecAppend(AppendState *node)
 		/*
 		 * get a tuple from the subplan
 		 */
-		result = ExecProcNode(subnode);
+		Assert(!node->as_finished[node->as_whichplan]);
+
+		if (!subnode->result_ready)
+			ExecExecuteNode(subnode);
+
+		if (!subnode->result_ready)
+		{
+			if (node->as_async)
+			{
+				/* Move to the next living node */
+				do
+				{
+					node->as_whichplan = 
+						(node->as_whichplan + 1) %  node->as_nplans;
+				} while (node->as_whichplan != stopplan &&
+						 node->as_finished[node->as_whichplan]);
+
+				/* No node is ready yet, return as not-ready */
+				if (node->as_whichplan == stopplan)
+					return;
+
+				/* Try the next node */
+				continue;
+			}
+
+			/* If not async, immediately wait for this subnode */
+			ExecAsyncWaitForNode(subnode);
+		}				
+
+		result = ExecConsumeResult((PlanState *) subnode);
 
 		if (!TupIsNull(result))
 		{
@@ -220,6 +256,31 @@ ExecAppend(AppendState *node)
 			return;
 		}
 
+		if (node->as_async)
+		{
+			node->as_finished[node->as_whichplan] = true;
+			stopplan = node->as_whichplan;
+
+			/* Find the next living subnode */
+			do
+			{
+				node->as_whichplan =
+					(node->as_whichplan + 1) % node->as_nplans;
+			} while (node->as_whichplan != stopplan &&
+					 node->as_finished[node->as_whichplan]);
+
+			if (node->as_whichplan != stopplan)
+			{
+				stopplan = node->as_whichplan;
+				continue;
+			}
+
+			/* All subnodes are exhausted. Finish this node. */
+			ExecReturnTuple(&node->ps,
+							ExecClearTuple(node->ps.ps_ResultTupleSlot));
+			return;
+		}
+
 		/*
 		 * Go on to the "next" subplan in the appropriate direction. If no
 		 * more subplans, return the empty slot set up for us by
@@ -277,6 +338,8 @@ ExecReScanAppend(AppendState *node)
 	{
 		PlanState  *subnode = node->appendplans[i];
 
+		node->as_finished[i] = false;
+
 		/*
 		 * ExecReScan doesn't know about my subplans, so I have to do
 		 * changed-parameter signaling myself.
diff --git a/src/include/nodes/execnodes.h b/src/include/nodes/execnodes.h
index b72decc..b0a86c5 100644
--- a/src/include/nodes/execnodes.h
+++ b/src/include/nodes/execnodes.h
@@ -1177,6 +1177,8 @@ typedef struct AppendState
 {
 	PlanState	ps;				/* its first field is NodeTag */
 	PlanState **appendplans;	/* array of PlanStates for my inputs */
+	bool		as_async;		/* true to allow async execution */
+	bool	   *as_finished;	/* array of the running state of subplans */
 	int			as_nplans;
 	int			as_whichplan;
 } AppendState;
-- 
1.8.3.1


----Next_Part(Thu_Jul_21_18_50_07_2016_597)--
Content-Type: text/plain
Content-Disposition: inline
Content-Transfer-Encoding: 8bit
MIME-Version: 1.0


-- 
Sent via pgsql-hackers mailing list ([email protected])
To make changes to your subscription:
http://www.postgresql.org/mailpref/pgsql-hackers

----Next_Part(Thu_Jul_21_18_50_07_2016_597)----






^ permalink  raw  reply  [nested|flat] 56+ messages in thread

* [PATCH 7/7] Make Append node async-aware.
@ 2016-06-28 09:52 Kyotaro Horiguchi <[email protected]>
  0 siblings, 0 replies; 56+ messages in thread

From: Kyotaro Horiguchi @ 2016-06-28 09:52 UTC (permalink / raw)

Change append node to be capable to handle asynchronous children
properly. As soon as it receives !async_ready from a child, it moves
to the next child and if no child is ready, it sleeps until at least
one of them become ready.
---
 src/backend/executor/nodeAppend.c | 67 +++++++++++++++++++++++++++++++++++++--
 src/include/nodes/execnodes.h     |  2 ++
 2 files changed, 67 insertions(+), 2 deletions(-)

diff --git a/src/backend/executor/nodeAppend.c b/src/backend/executor/nodeAppend.c
index e0ce8c6..9a4063a 100644
--- a/src/backend/executor/nodeAppend.c
+++ b/src/backend/executor/nodeAppend.c
@@ -58,6 +58,7 @@
 #include "postgres.h"
 
 #include "executor/execdebug.h"
+#include "executor/execAsync.h"
 #include "executor/nodeAppend.h"
 
 static bool exec_append_initialize_next(AppendState *appendstate);
@@ -121,6 +122,7 @@ ExecInitAppend(Append *node, EState *estate, int eflags)
 {
 	AppendState *appendstate = makeNode(AppendState);
 	PlanState **appendplanstates;
+	bool	   *finished;
 	int			nplans;
 	int			i;
 	ListCell   *lc;
@@ -134,14 +136,17 @@ ExecInitAppend(Append *node, EState *estate, int eflags)
 	nplans = list_length(node->appendplans);
 
 	appendplanstates = (PlanState **) palloc0(nplans * sizeof(PlanState *));
-
+	finished = (bool *) palloc0(nplans * sizeof(bool));
+	
 	/*
 	 * create new AppendState for our append node
 	 */
 	appendstate->ps.plan = (Plan *) node;
 	appendstate->ps.state = estate;
 	appendstate->appendplans = appendplanstates;
+	appendstate->as_finished = finished;
 	appendstate->as_nplans = nplans;
+	appendstate->as_async = ((eflags & EXEC_FLAG_BACKWARD) == 0);
 
 	/*
 	 * Miscellaneous initialization
@@ -194,6 +199,8 @@ ExecInitAppend(Append *node, EState *estate, int eflags)
 void
 ExecAppend(AppendState *node)
 {
+	int stopplan = node->as_whichplan;
+
 	for (;;)
 	{
 		PlanState  *subnode;
@@ -207,7 +214,36 @@ ExecAppend(AppendState *node)
 		/*
 		 * get a tuple from the subplan
 		 */
-		result = ExecProcNode(subnode);
+		Assert(!node->as_finished[node->as_whichplan]);
+
+		if (!subnode->result_ready)
+			ExecExecuteNode(subnode);
+
+		if (!subnode->result_ready)
+		{
+			if (node->as_async)
+			{
+				/* Move to the next living node */
+				do
+				{
+					node->as_whichplan = 
+						(node->as_whichplan + 1) %  node->as_nplans;
+				} while (node->as_whichplan != stopplan &&
+						 node->as_finished[node->as_whichplan]);
+
+				/* No node is ready yet, return as not-ready */
+				if (node->as_whichplan == stopplan)
+					return;
+
+				/* Try the next node */
+				continue;
+			}
+
+			/* If not async, immediately wait for this subnode */
+			ExecAsyncWaitForNode(subnode);
+		}				
+
+		result = ExecConsumeResult((PlanState *) subnode);
 
 		if (!TupIsNull(result))
 		{
@@ -220,6 +256,31 @@ ExecAppend(AppendState *node)
 			return;
 		}
 
+		if (node->as_async)
+		{
+			node->as_finished[node->as_whichplan] = true;
+			stopplan = node->as_whichplan;
+
+			/* Find the next living subnode */
+			do
+			{
+				node->as_whichplan =
+					(node->as_whichplan + 1) % node->as_nplans;
+			} while (node->as_whichplan != stopplan &&
+					 node->as_finished[node->as_whichplan]);
+
+			if (node->as_whichplan != stopplan)
+			{
+				stopplan = node->as_whichplan;
+				continue;
+			}
+
+			/* All subnodes are exhausted. Finish this node. */
+			ExecReturnTuple(&node->ps,
+							ExecClearTuple(node->ps.ps_ResultTupleSlot));
+			return;
+		}
+
 		/*
 		 * Go on to the "next" subplan in the appropriate direction. If no
 		 * more subplans, return the empty slot set up for us by
@@ -277,6 +338,8 @@ ExecReScanAppend(AppendState *node)
 	{
 		PlanState  *subnode = node->appendplans[i];
 
+		node->as_finished[i] = false;
+
 		/*
 		 * ExecReScan doesn't know about my subplans, so I have to do
 		 * changed-parameter signaling myself.
diff --git a/src/include/nodes/execnodes.h b/src/include/nodes/execnodes.h
index b72decc..b0a86c5 100644
--- a/src/include/nodes/execnodes.h
+++ b/src/include/nodes/execnodes.h
@@ -1177,6 +1177,8 @@ typedef struct AppendState
 {
 	PlanState	ps;				/* its first field is NodeTag */
 	PlanState **appendplans;	/* array of PlanStates for my inputs */
+	bool		as_async;		/* true to allow async execution */
+	bool	   *as_finished;	/* array of the running state of subplans */
 	int			as_nplans;
 	int			as_whichplan;
 } AppendState;
-- 
1.8.3.1


----Next_Part(Thu_Jul_21_18_50_07_2016_597)--
Content-Type: text/plain
Content-Disposition: inline
Content-Transfer-Encoding: 8bit
MIME-Version: 1.0


-- 
Sent via pgsql-hackers mailing list ([email protected])
To make changes to your subscription:
http://www.postgresql.org/mailpref/pgsql-hackers

----Next_Part(Thu_Jul_21_18_50_07_2016_597)----






^ permalink  raw  reply  [nested|flat] 56+ messages in thread

* [PATCH 7/7] Make Append node async-aware.
@ 2016-06-28 09:52 Kyotaro Horiguchi <[email protected]>
  0 siblings, 0 replies; 56+ messages in thread

From: Kyotaro Horiguchi @ 2016-06-28 09:52 UTC (permalink / raw)

Change append node to be capable to handle asynchronous children
properly. As soon as it receives !async_ready from a child, it moves
to the next child and if no child is ready, it sleeps until at least
one of them become ready.
---
 src/backend/executor/nodeAppend.c | 67 +++++++++++++++++++++++++++++++++++++--
 src/include/nodes/execnodes.h     |  2 ++
 2 files changed, 67 insertions(+), 2 deletions(-)

diff --git a/src/backend/executor/nodeAppend.c b/src/backend/executor/nodeAppend.c
index e0ce8c6..9a4063a 100644
--- a/src/backend/executor/nodeAppend.c
+++ b/src/backend/executor/nodeAppend.c
@@ -58,6 +58,7 @@
 #include "postgres.h"
 
 #include "executor/execdebug.h"
+#include "executor/execAsync.h"
 #include "executor/nodeAppend.h"
 
 static bool exec_append_initialize_next(AppendState *appendstate);
@@ -121,6 +122,7 @@ ExecInitAppend(Append *node, EState *estate, int eflags)
 {
 	AppendState *appendstate = makeNode(AppendState);
 	PlanState **appendplanstates;
+	bool	   *finished;
 	int			nplans;
 	int			i;
 	ListCell   *lc;
@@ -134,14 +136,17 @@ ExecInitAppend(Append *node, EState *estate, int eflags)
 	nplans = list_length(node->appendplans);
 
 	appendplanstates = (PlanState **) palloc0(nplans * sizeof(PlanState *));
-
+	finished = (bool *) palloc0(nplans * sizeof(bool));
+	
 	/*
 	 * create new AppendState for our append node
 	 */
 	appendstate->ps.plan = (Plan *) node;
 	appendstate->ps.state = estate;
 	appendstate->appendplans = appendplanstates;
+	appendstate->as_finished = finished;
 	appendstate->as_nplans = nplans;
+	appendstate->as_async = ((eflags & EXEC_FLAG_BACKWARD) == 0);
 
 	/*
 	 * Miscellaneous initialization
@@ -194,6 +199,8 @@ ExecInitAppend(Append *node, EState *estate, int eflags)
 void
 ExecAppend(AppendState *node)
 {
+	int stopplan = node->as_whichplan;
+
 	for (;;)
 	{
 		PlanState  *subnode;
@@ -207,7 +214,36 @@ ExecAppend(AppendState *node)
 		/*
 		 * get a tuple from the subplan
 		 */
-		result = ExecProcNode(subnode);
+		Assert(!node->as_finished[node->as_whichplan]);
+
+		if (!subnode->result_ready)
+			ExecExecuteNode(subnode);
+
+		if (!subnode->result_ready)
+		{
+			if (node->as_async)
+			{
+				/* Move to the next living node */
+				do
+				{
+					node->as_whichplan = 
+						(node->as_whichplan + 1) %  node->as_nplans;
+				} while (node->as_whichplan != stopplan &&
+						 node->as_finished[node->as_whichplan]);
+
+				/* No node is ready yet, return as not-ready */
+				if (node->as_whichplan == stopplan)
+					return;
+
+				/* Try the next node */
+				continue;
+			}
+
+			/* If not async, immediately wait for this subnode */
+			ExecAsyncWaitForNode(subnode);
+		}				
+
+		result = ExecConsumeResult((PlanState *) subnode);
 
 		if (!TupIsNull(result))
 		{
@@ -220,6 +256,31 @@ ExecAppend(AppendState *node)
 			return;
 		}
 
+		if (node->as_async)
+		{
+			node->as_finished[node->as_whichplan] = true;
+			stopplan = node->as_whichplan;
+
+			/* Find the next living subnode */
+			do
+			{
+				node->as_whichplan =
+					(node->as_whichplan + 1) % node->as_nplans;
+			} while (node->as_whichplan != stopplan &&
+					 node->as_finished[node->as_whichplan]);
+
+			if (node->as_whichplan != stopplan)
+			{
+				stopplan = node->as_whichplan;
+				continue;
+			}
+
+			/* All subnodes are exhausted. Finish this node. */
+			ExecReturnTuple(&node->ps,
+							ExecClearTuple(node->ps.ps_ResultTupleSlot));
+			return;
+		}
+
 		/*
 		 * Go on to the "next" subplan in the appropriate direction. If no
 		 * more subplans, return the empty slot set up for us by
@@ -277,6 +338,8 @@ ExecReScanAppend(AppendState *node)
 	{
 		PlanState  *subnode = node->appendplans[i];
 
+		node->as_finished[i] = false;
+
 		/*
 		 * ExecReScan doesn't know about my subplans, so I have to do
 		 * changed-parameter signaling myself.
diff --git a/src/include/nodes/execnodes.h b/src/include/nodes/execnodes.h
index b72decc..b0a86c5 100644
--- a/src/include/nodes/execnodes.h
+++ b/src/include/nodes/execnodes.h
@@ -1177,6 +1177,8 @@ typedef struct AppendState
 {
 	PlanState	ps;				/* its first field is NodeTag */
 	PlanState **appendplans;	/* array of PlanStates for my inputs */
+	bool		as_async;		/* true to allow async execution */
+	bool	   *as_finished;	/* array of the running state of subplans */
 	int			as_nplans;
 	int			as_whichplan;
 } AppendState;
-- 
2.9.2


----Next_Part(Mon_Aug_29_17_08_36_2016_213)--
Content-Type: text/plain
Content-Disposition: inline
Content-Transfer-Encoding: 8bit
MIME-Version: 1.0


-- 
Sent via pgsql-hackers mailing list ([email protected])
To make changes to your subscription:
http://www.postgresql.org/mailpref/pgsql-hackers

----Next_Part(Mon_Aug_29_17_08_36_2016_213)----






^ permalink  raw  reply  [nested|flat] 56+ messages in thread

* [PATCH 7/7] Make Append node async-aware.
@ 2016-06-28 09:52 Kyotaro Horiguchi <[email protected]>
  0 siblings, 0 replies; 56+ messages in thread

From: Kyotaro Horiguchi @ 2016-06-28 09:52 UTC (permalink / raw)

Change append node to be capable to handle asynchronous children
properly. As soon as it receives !async_ready from a child, it moves
to the next child and if no child is ready, it sleeps until at least
one of them become ready.
---
 src/backend/executor/nodeAppend.c | 67 +++++++++++++++++++++++++++++++++++++--
 src/include/nodes/execnodes.h     |  2 ++
 2 files changed, 67 insertions(+), 2 deletions(-)

diff --git a/src/backend/executor/nodeAppend.c b/src/backend/executor/nodeAppend.c
index e0ce8c6..9a4063a 100644
--- a/src/backend/executor/nodeAppend.c
+++ b/src/backend/executor/nodeAppend.c
@@ -58,6 +58,7 @@
 #include "postgres.h"
 
 #include "executor/execdebug.h"
+#include "executor/execAsync.h"
 #include "executor/nodeAppend.h"
 
 static bool exec_append_initialize_next(AppendState *appendstate);
@@ -121,6 +122,7 @@ ExecInitAppend(Append *node, EState *estate, int eflags)
 {
 	AppendState *appendstate = makeNode(AppendState);
 	PlanState **appendplanstates;
+	bool	   *finished;
 	int			nplans;
 	int			i;
 	ListCell   *lc;
@@ -134,14 +136,17 @@ ExecInitAppend(Append *node, EState *estate, int eflags)
 	nplans = list_length(node->appendplans);
 
 	appendplanstates = (PlanState **) palloc0(nplans * sizeof(PlanState *));
-
+	finished = (bool *) palloc0(nplans * sizeof(bool));
+	
 	/*
 	 * create new AppendState for our append node
 	 */
 	appendstate->ps.plan = (Plan *) node;
 	appendstate->ps.state = estate;
 	appendstate->appendplans = appendplanstates;
+	appendstate->as_finished = finished;
 	appendstate->as_nplans = nplans;
+	appendstate->as_async = ((eflags & EXEC_FLAG_BACKWARD) == 0);
 
 	/*
 	 * Miscellaneous initialization
@@ -194,6 +199,8 @@ ExecInitAppend(Append *node, EState *estate, int eflags)
 void
 ExecAppend(AppendState *node)
 {
+	int stopplan = node->as_whichplan;
+
 	for (;;)
 	{
 		PlanState  *subnode;
@@ -207,7 +214,36 @@ ExecAppend(AppendState *node)
 		/*
 		 * get a tuple from the subplan
 		 */
-		result = ExecProcNode(subnode);
+		Assert(!node->as_finished[node->as_whichplan]);
+
+		if (!subnode->result_ready)
+			ExecExecuteNode(subnode);
+
+		if (!subnode->result_ready)
+		{
+			if (node->as_async)
+			{
+				/* Move to the next living node */
+				do
+				{
+					node->as_whichplan = 
+						(node->as_whichplan + 1) %  node->as_nplans;
+				} while (node->as_whichplan != stopplan &&
+						 node->as_finished[node->as_whichplan]);
+
+				/* No node is ready yet, return as not-ready */
+				if (node->as_whichplan == stopplan)
+					return;
+
+				/* Try the next node */
+				continue;
+			}
+
+			/* If not async, immediately wait for this subnode */
+			ExecAsyncWaitForNode(subnode);
+		}				
+
+		result = ExecConsumeResult((PlanState *) subnode);
 
 		if (!TupIsNull(result))
 		{
@@ -220,6 +256,31 @@ ExecAppend(AppendState *node)
 			return;
 		}
 
+		if (node->as_async)
+		{
+			node->as_finished[node->as_whichplan] = true;
+			stopplan = node->as_whichplan;
+
+			/* Find the next living subnode */
+			do
+			{
+				node->as_whichplan =
+					(node->as_whichplan + 1) % node->as_nplans;
+			} while (node->as_whichplan != stopplan &&
+					 node->as_finished[node->as_whichplan]);
+
+			if (node->as_whichplan != stopplan)
+			{
+				stopplan = node->as_whichplan;
+				continue;
+			}
+
+			/* All subnodes are exhausted. Finish this node. */
+			ExecReturnTuple(&node->ps,
+							ExecClearTuple(node->ps.ps_ResultTupleSlot));
+			return;
+		}
+
 		/*
 		 * Go on to the "next" subplan in the appropriate direction. If no
 		 * more subplans, return the empty slot set up for us by
@@ -277,6 +338,8 @@ ExecReScanAppend(AppendState *node)
 	{
 		PlanState  *subnode = node->appendplans[i];
 
+		node->as_finished[i] = false;
+
 		/*
 		 * ExecReScan doesn't know about my subplans, so I have to do
 		 * changed-parameter signaling myself.
diff --git a/src/include/nodes/execnodes.h b/src/include/nodes/execnodes.h
index b72decc..b0a86c5 100644
--- a/src/include/nodes/execnodes.h
+++ b/src/include/nodes/execnodes.h
@@ -1177,6 +1177,8 @@ typedef struct AppendState
 {
 	PlanState	ps;				/* its first field is NodeTag */
 	PlanState **appendplans;	/* array of PlanStates for my inputs */
+	bool		as_async;		/* true to allow async execution */
+	bool	   *as_finished;	/* array of the running state of subplans */
 	int			as_nplans;
 	int			as_whichplan;
 } AppendState;
-- 
1.8.3.1


----Next_Part(Thu_Jul_21_18_50_07_2016_597)--
Content-Type: text/plain
Content-Disposition: inline
Content-Transfer-Encoding: 8bit
MIME-Version: 1.0


-- 
Sent via pgsql-hackers mailing list ([email protected])
To make changes to your subscription:
http://www.postgresql.org/mailpref/pgsql-hackers

----Next_Part(Thu_Jul_21_18_50_07_2016_597)----






^ permalink  raw  reply  [nested|flat] 56+ messages in thread

* [PATCH 7/7] Make Append node async-aware.
@ 2016-06-28 09:52 Kyotaro Horiguchi <[email protected]>
  0 siblings, 0 replies; 56+ messages in thread

From: Kyotaro Horiguchi @ 2016-06-28 09:52 UTC (permalink / raw)

Change append node to be capable to handle asynchronous children
properly. As soon as it receives !async_ready from a child, it moves
to the next child and if no child is ready, it sleeps until at least
one of them become ready.
---
 src/backend/executor/nodeAppend.c | 94 ++++++++++++++++++++++++++++++++-------
 src/include/nodes/execnodes.h     |  2 +
 2 files changed, 80 insertions(+), 16 deletions(-)

diff --git a/src/backend/executor/nodeAppend.c b/src/backend/executor/nodeAppend.c
index e0ce8c6..1c0d26e 100644
--- a/src/backend/executor/nodeAppend.c
+++ b/src/backend/executor/nodeAppend.c
@@ -58,6 +58,7 @@
 #include "postgres.h"
 
 #include "executor/execdebug.h"
+#include "executor/execAsync.h"
 #include "executor/nodeAppend.h"
 
 static bool exec_append_initialize_next(AppendState *appendstate);
@@ -121,6 +122,7 @@ ExecInitAppend(Append *node, EState *estate, int eflags)
 {
 	AppendState *appendstate = makeNode(AppendState);
 	PlanState **appendplanstates;
+	bool	   *finished;
 	int			nplans;
 	int			i;
 	ListCell   *lc;
@@ -134,14 +136,17 @@ ExecInitAppend(Append *node, EState *estate, int eflags)
 	nplans = list_length(node->appendplans);
 
 	appendplanstates = (PlanState **) palloc0(nplans * sizeof(PlanState *));
-
+	finished = (bool *) palloc0(nplans * sizeof(bool));
+	
 	/*
 	 * create new AppendState for our append node
 	 */
 	appendstate->ps.plan = (Plan *) node;
 	appendstate->ps.state = estate;
 	appendstate->appendplans = appendplanstates;
+	appendstate->as_finished = finished;
 	appendstate->as_nplans = nplans;
+	appendstate->as_async = ((eflags & EXEC_FLAG_BACKWARD) == 0);
 
 	/*
 	 * Miscellaneous initialization
@@ -194,49 +199,104 @@ ExecInitAppend(Append *node, EState *estate, int eflags)
 void
 ExecAppend(AppendState *node)
 {
-	for (;;)
+	int n_notready = 0;
+	PlanState  *subnode;
+	int i, n;
+
+	n = node->as_whichplan;
+
+	for (i = 0 ; i < node->as_nplans ; i++)
 	{
-		PlanState  *subnode;
 		TupleTableSlot *result;
 
+		if (node->as_async)
+		{
+			if (n >= node->as_nplans)
+				n = 0;
+
+			if (node->as_finished[n])
+			{
+				n++;
+				continue;
+			}
+		}
+
 		/*
 		 * figure out which subplan we are currently processing
 		 */
-		subnode = node->appendplans[node->as_whichplan];
+		subnode = node->appendplans[n];
 
 		/*
-		 * get a tuple from the subplan
+		 * execute the subplan to get a result if it is not ready yet
 		 */
-		result = ExecProcNode(subnode);
+		if (!subnode->result_ready)
+			ExecExecuteNode(subnode);
+
+		/*
+		 * This subnode is not ready yet when asynchrony is not allowed,
+		 * immediately wait for this subnode.
+		 */
+		if (!subnode->result_ready)
+		{
+			if (node->as_async)
+			{
+				n_notready++;
+				n++;
+				continue;
+			}
+			ExecAsyncWaitForNode(subnode);
+		}
+
+		Assert(subnode->result_ready);
+
+		result = ExecConsumeResult((PlanState *)subnode);
 
 		if (!TupIsNull(result))
 		{
 			/*
-			 * If the subplan gave us something then return it as-is. We do
-			 * NOT make use of the result slot that was set up in
-			 * ExecInitAppend; there's no need for it.
+			 * If the subplan gave us something then return it
+			 * as-is. We do NOT make use of the result slot that was
+			 * set up in ExecInitAppend; there's no need for it.
 			 */
+			node->as_whichplan = n;
 			ExecReturnTuple(&node->ps, result);
 			return;
 		}
 
+		/* Tuple has been exhausted on this subnode */
+		if (node->as_async)
+		{
+			node->as_finished[n++] = true;
+			continue;
+		}
+
 		/*
 		 * Go on to the "next" subplan in the appropriate direction. If no
 		 * more subplans, return the empty slot set up for us by
 		 * ExecInitAppend.
 		 */
 		if (ScanDirectionIsForward(node->ps.state->es_direction))
-			node->as_whichplan++;
+		{
+			if (n >= node->as_nplans - 1)
+				break;
+			n++;
+		}
 		else
-			node->as_whichplan--;
-		if (!exec_append_initialize_next(node))
 		{
-			ExecReturnTuple(&node->ps,
-							ExecClearTuple(node->ps.ps_ResultTupleSlot));
-			return;
+			if (n == 0)
+				break;
+			n--;
 		}
+	}
 
-		/* Else loop back and try to get a tuple from the new subplan */
+	/*
+	 * We are finished if reached here and no subnodes are not-ready
+	 */
+	if (n_notready == 0)
+	{
+		node->as_whichplan = n;
+		ExecReturnTuple(&node->ps,
+						ExecClearTuple(node->ps.ps_ResultTupleSlot));
 	}
 }
 
@@ -277,6 +337,8 @@ ExecReScanAppend(AppendState *node)
 	{
 		PlanState  *subnode = node->appendplans[i];
 
+		node->as_finished[i] = false;
+
 		/*
 		 * ExecReScan doesn't know about my subplans, so I have to do
 		 * changed-parameter signaling myself.
diff --git a/src/include/nodes/execnodes.h b/src/include/nodes/execnodes.h
index 9121537..e4f2bb6 100644
--- a/src/include/nodes/execnodes.h
+++ b/src/include/nodes/execnodes.h
@@ -1170,6 +1170,8 @@ typedef struct AppendState
 {
 	PlanState	ps;				/* its first field is NodeTag */
 	PlanState **appendplans;	/* array of PlanStates for my inputs */
+	bool		as_async;		/* true to allow async execution */
+	bool	   *as_finished;	/* array of the running state of subplans */
 	int			as_nplans;
 	int			as_whichplan;
 } AppendState;
-- 
1.8.3.1


----Next_Part(Wed_Jul_06_16_29_14_2016_599)--
Content-Type: Text/Plain; charset=us-ascii
Content-Transfer-Encoding: 7bit
Content-Disposition: inline; filename="gentbl.sql"

CREATE EXTENSION postgres_fdw;
CREATE TABLE pl (a int, b int);
CREATE TABLE cl1 (LIKE pl) INHERITS (pl);
CREATE TABLE cl2 (LIKE pl) INHERITS (pl);
CREATE TABLE cl3 (LIKE pl) INHERITS (pl);
CREATE TABLE cl4 (LIKE pl) INHERITS (pl);
INSERT INTO cl1 (SELECT a, a FROM generate_series(0000000, 0999999) a);
INSERT INTO cl2 (SELECT a, a FROM generate_series(1000000, 1999999) a);
INSERT INTO cl3 (SELECT a, a FROM generate_series(2000000, 2999999) a);
INSERT INTO cl4 (SELECT a, a FROM generate_series(3000000, 3999999) a);

CREATE TABLE t0 (LIKE pl);
INSERT INTO t0 (SELECT a, a FROM generate_series(0000000, 9999999) a);

CREATE SERVER sv0 FOREIGN DATA WRAPPER postgres_fdw OPTIONS (host '/tmp', dbname 'postgres');
CREATE SERVER sv1 FOREIGN DATA WRAPPER postgres_fdw OPTIONS (host '/tmp', dbname 'postgres');
CREATE SERVER sv2 FOREIGN DATA WRAPPER postgres_fdw OPTIONS (host '/tmp', dbname 'postgres');
CREATE SERVER sv3 FOREIGN DATA WRAPPER postgres_fdw OPTIONS (host '/tmp', dbname 'postgres');
CREATE SERVER sv4 FOREIGN DATA WRAPPER postgres_fdw OPTIONS (host '/tmp', dbname 'postgres');
CREATE USER MAPPING FOR public SERVER sv0;
CREATE USER MAPPING FOR public SERVER sv1;
CREATE USER MAPPING FOR public SERVER sv2;
CREATE USER MAPPING FOR public SERVER sv3;
CREATE USER MAPPING FOR public SERVER sv4;

CREATE FOREIGN TABLE ft10 (a int, b int) SERVER sv0 OPTIONS (table_name 'cl1');
CREATE FOREIGN TABLE ft20 (a int, b int) SERVER sv0 OPTIONS (table_name 'cl2');
CREATE FOREIGN TABLE ft30 (a int, b int) SERVER sv0 OPTIONS (table_name 'cl3');
CREATE FOREIGN TABLE ft40 (a int, b int) SERVER sv0 OPTIONS (table_name 'cl4');
CREATE FOREIGN TABLE ft11 (a int, b int) SERVER sv1 OPTIONS (table_name 'cl1');
CREATE FOREIGN TABLE ft22 (a int, b int) SERVER sv2 OPTIONS (table_name 'cl2');
CREATE FOREIGN TABLE ft33 (a int, b int) SERVER sv3 OPTIONS (table_name 'cl3');
CREATE FOREIGN TABLE ft44 (a int, b int) SERVER sv4 OPTIONS (table_name 'cl4');

CREATE TABLE pf0 (LIKE pl);
ALTER FOREIGN TABLE ft10 INHERIT pf0;
ALTER FOREIGN TABLE ft20 INHERIT pf0;
ALTER FOREIGN TABLE ft30 INHERIT pf0;
ALTER FOREIGN TABLE ft40 INHERIT pf0;

CREATE table pf1 (LIKE pl);
ALTER FOREIGN TABLE ft11 INHERIT pf1;
ALTER FOREIGN TABLE ft22 INHERIT pf1;
ALTER FOREIGN TABLE ft33 INHERIT pf1;
ALTER FOREIGN TABLE ft44 INHERIT pf1;

ANALYZE;

----Next_Part(Wed_Jul_06_16_29_14_2016_599)--
Content-Type: Text/Plain; charset=us-ascii
Content-Transfer-Encoding: 7bit
Content-Disposition: inline; filename="testrun.sh"

#! /bin/bash

function do_test() {
	echo $1
	for i in $(seq 1 10); do psql postgres -c "set log_min_duration_statement = 0; set client_min_messages=log; select sum(a) from $1"; done | grep LOG
}

for i in "t0" "pl" "pf0" "pf1";
do
  do_test $i
done


----Next_Part(Wed_Jul_06_16_29_14_2016_599)--
Content-Type: text/plain
Content-Disposition: inline
Content-Transfer-Encoding: 8bit
MIME-Version: 1.0


-- 
Sent via pgsql-hackers mailing list ([email protected])
To make changes to your subscription:
http://www.postgresql.org/mailpref/pgsql-hackers

----Next_Part(Wed_Jul_06_16_29_14_2016_599)----






^ permalink  raw  reply  [nested|flat] 56+ messages in thread

* [PATCH 7/7] Make Append node async-aware.
@ 2016-06-28 09:52 Kyotaro Horiguchi <[email protected]>
  0 siblings, 0 replies; 56+ messages in thread

From: Kyotaro Horiguchi @ 2016-06-28 09:52 UTC (permalink / raw)

Change append node to be capable to handle asynchronous children
properly. As soon as it receives !async_ready from a child, it moves
to the next child and if no child is ready, it sleeps until at least
one of them become ready.
---
 src/backend/executor/nodeAppend.c | 67 +++++++++++++++++++++++++++++++++++++--
 src/include/nodes/execnodes.h     |  2 ++
 2 files changed, 67 insertions(+), 2 deletions(-)

diff --git a/src/backend/executor/nodeAppend.c b/src/backend/executor/nodeAppend.c
index e0ce8c6..9a4063a 100644
--- a/src/backend/executor/nodeAppend.c
+++ b/src/backend/executor/nodeAppend.c
@@ -58,6 +58,7 @@
 #include "postgres.h"
 
 #include "executor/execdebug.h"
+#include "executor/execAsync.h"
 #include "executor/nodeAppend.h"
 
 static bool exec_append_initialize_next(AppendState *appendstate);
@@ -121,6 +122,7 @@ ExecInitAppend(Append *node, EState *estate, int eflags)
 {
 	AppendState *appendstate = makeNode(AppendState);
 	PlanState **appendplanstates;
+	bool	   *finished;
 	int			nplans;
 	int			i;
 	ListCell   *lc;
@@ -134,14 +136,17 @@ ExecInitAppend(Append *node, EState *estate, int eflags)
 	nplans = list_length(node->appendplans);
 
 	appendplanstates = (PlanState **) palloc0(nplans * sizeof(PlanState *));
-
+	finished = (bool *) palloc0(nplans * sizeof(bool));
+	
 	/*
 	 * create new AppendState for our append node
 	 */
 	appendstate->ps.plan = (Plan *) node;
 	appendstate->ps.state = estate;
 	appendstate->appendplans = appendplanstates;
+	appendstate->as_finished = finished;
 	appendstate->as_nplans = nplans;
+	appendstate->as_async = ((eflags & EXEC_FLAG_BACKWARD) == 0);
 
 	/*
 	 * Miscellaneous initialization
@@ -194,6 +199,8 @@ ExecInitAppend(Append *node, EState *estate, int eflags)
 void
 ExecAppend(AppendState *node)
 {
+	int stopplan = node->as_whichplan;
+
 	for (;;)
 	{
 		PlanState  *subnode;
@@ -207,7 +214,36 @@ ExecAppend(AppendState *node)
 		/*
 		 * get a tuple from the subplan
 		 */
-		result = ExecProcNode(subnode);
+		Assert(!node->as_finished[node->as_whichplan]);
+
+		if (!subnode->result_ready)
+			ExecExecuteNode(subnode);
+
+		if (!subnode->result_ready)
+		{
+			if (node->as_async)
+			{
+				/* Move to the next living node */
+				do
+				{
+					node->as_whichplan = 
+						(node->as_whichplan + 1) %  node->as_nplans;
+				} while (node->as_whichplan != stopplan &&
+						 node->as_finished[node->as_whichplan]);
+
+				/* No node is ready yet, return as not-ready */
+				if (node->as_whichplan == stopplan)
+					return;
+
+				/* Try the next node */
+				continue;
+			}
+
+			/* If not async, immediately wait for this subnode */
+			ExecAsyncWaitForNode(subnode);
+		}				
+
+		result = ExecConsumeResult((PlanState *) subnode);
 
 		if (!TupIsNull(result))
 		{
@@ -220,6 +256,31 @@ ExecAppend(AppendState *node)
 			return;
 		}
 
+		if (node->as_async)
+		{
+			node->as_finished[node->as_whichplan] = true;
+			stopplan = node->as_whichplan;
+
+			/* Find the next living subnode */
+			do
+			{
+				node->as_whichplan =
+					(node->as_whichplan + 1) % node->as_nplans;
+			} while (node->as_whichplan != stopplan &&
+					 node->as_finished[node->as_whichplan]);
+
+			if (node->as_whichplan != stopplan)
+			{
+				stopplan = node->as_whichplan;
+				continue;
+			}
+
+			/* All subnodes are exhausted. Finish this node. */
+			ExecReturnTuple(&node->ps,
+							ExecClearTuple(node->ps.ps_ResultTupleSlot));
+			return;
+		}
+
 		/*
 		 * Go on to the "next" subplan in the appropriate direction. If no
 		 * more subplans, return the empty slot set up for us by
@@ -277,6 +338,8 @@ ExecReScanAppend(AppendState *node)
 	{
 		PlanState  *subnode = node->appendplans[i];
 
+		node->as_finished[i] = false;
+
 		/*
 		 * ExecReScan doesn't know about my subplans, so I have to do
 		 * changed-parameter signaling myself.
diff --git a/src/include/nodes/execnodes.h b/src/include/nodes/execnodes.h
index b72decc..b0a86c5 100644
--- a/src/include/nodes/execnodes.h
+++ b/src/include/nodes/execnodes.h
@@ -1177,6 +1177,8 @@ typedef struct AppendState
 {
 	PlanState	ps;				/* its first field is NodeTag */
 	PlanState **appendplans;	/* array of PlanStates for my inputs */
+	bool		as_async;		/* true to allow async execution */
+	bool	   *as_finished;	/* array of the running state of subplans */
 	int			as_nplans;
 	int			as_whichplan;
 } AppendState;
-- 
1.8.3.1


----Next_Part(Thu_Jul_21_18_50_07_2016_597)--
Content-Type: text/plain
Content-Disposition: inline
Content-Transfer-Encoding: 8bit
MIME-Version: 1.0


-- 
Sent via pgsql-hackers mailing list ([email protected])
To make changes to your subscription:
http://www.postgresql.org/mailpref/pgsql-hackers

----Next_Part(Thu_Jul_21_18_50_07_2016_597)----






^ permalink  raw  reply  [nested|flat] 56+ messages in thread

* [PATCH 7/7] Make Append node async-aware.
@ 2016-06-28 09:52 Kyotaro Horiguchi <[email protected]>
  0 siblings, 0 replies; 56+ messages in thread

From: Kyotaro Horiguchi @ 2016-06-28 09:52 UTC (permalink / raw)

Change append node to be capable to handle asynchronous children
properly. As soon as it receives !async_ready from a child, it moves
to the next child and if no child is ready, it sleeps until at least
one of them become ready.
---
 src/backend/executor/nodeAppend.c | 94 ++++++++++++++++++++++++++++++++-------
 src/include/nodes/execnodes.h     |  2 +
 2 files changed, 80 insertions(+), 16 deletions(-)

diff --git a/src/backend/executor/nodeAppend.c b/src/backend/executor/nodeAppend.c
index e0ce8c6..1c0d26e 100644
--- a/src/backend/executor/nodeAppend.c
+++ b/src/backend/executor/nodeAppend.c
@@ -58,6 +58,7 @@
 #include "postgres.h"
 
 #include "executor/execdebug.h"
+#include "executor/execAsync.h"
 #include "executor/nodeAppend.h"
 
 static bool exec_append_initialize_next(AppendState *appendstate);
@@ -121,6 +122,7 @@ ExecInitAppend(Append *node, EState *estate, int eflags)
 {
 	AppendState *appendstate = makeNode(AppendState);
 	PlanState **appendplanstates;
+	bool	   *finished;
 	int			nplans;
 	int			i;
 	ListCell   *lc;
@@ -134,14 +136,17 @@ ExecInitAppend(Append *node, EState *estate, int eflags)
 	nplans = list_length(node->appendplans);
 
 	appendplanstates = (PlanState **) palloc0(nplans * sizeof(PlanState *));
-
+	finished = (bool *) palloc0(nplans * sizeof(bool));
+	
 	/*
 	 * create new AppendState for our append node
 	 */
 	appendstate->ps.plan = (Plan *) node;
 	appendstate->ps.state = estate;
 	appendstate->appendplans = appendplanstates;
+	appendstate->as_finished = finished;
 	appendstate->as_nplans = nplans;
+	appendstate->as_async = ((eflags & EXEC_FLAG_BACKWARD) == 0);
 
 	/*
 	 * Miscellaneous initialization
@@ -194,49 +199,104 @@ ExecInitAppend(Append *node, EState *estate, int eflags)
 void
 ExecAppend(AppendState *node)
 {
-	for (;;)
+	int n_notready = 0;
+	PlanState  *subnode;
+	int i, n;
+
+	n = node->as_whichplan;
+
+	for (i = 0 ; i < node->as_nplans ; i++)
 	{
-		PlanState  *subnode;
 		TupleTableSlot *result;
 
+		if (node->as_async)
+		{
+			if (n >= node->as_nplans)
+				n = 0;
+
+			if (node->as_finished[n])
+			{
+				n++;
+				continue;
+			}
+		}
+
 		/*
 		 * figure out which subplan we are currently processing
 		 */
-		subnode = node->appendplans[node->as_whichplan];
+		subnode = node->appendplans[n];
 
 		/*
-		 * get a tuple from the subplan
+		 * execute the subplan to get a result if it is not ready yet
 		 */
-		result = ExecProcNode(subnode);
+		if (!subnode->result_ready)
+			ExecExecuteNode(subnode);
+
+		/*
+		 * This subnode is not ready yet when asynchrony is not allowed,
+		 * immediately wait for this subnode.
+		 */
+		if (!subnode->result_ready)
+		{
+			if (node->as_async)
+			{
+				n_notready++;
+				n++;
+				continue;
+			}
+			ExecAsyncWaitForNode(subnode);
+		}
+
+		Assert(subnode->result_ready);
+
+		result = ExecConsumeResult((PlanState *)subnode);
 
 		if (!TupIsNull(result))
 		{
 			/*
-			 * If the subplan gave us something then return it as-is. We do
-			 * NOT make use of the result slot that was set up in
-			 * ExecInitAppend; there's no need for it.
+			 * If the subplan gave us something then return it
+			 * as-is. We do NOT make use of the result slot that was
+			 * set up in ExecInitAppend; there's no need for it.
 			 */
+			node->as_whichplan = n;
 			ExecReturnTuple(&node->ps, result);
 			return;
 		}
 
+		/* Tuple has been exhausted on this subnode */
+		if (node->as_async)
+		{
+			node->as_finished[n++] = true;
+			continue;
+		}
+
 		/*
 		 * Go on to the "next" subplan in the appropriate direction. If no
 		 * more subplans, return the empty slot set up for us by
 		 * ExecInitAppend.
 		 */
 		if (ScanDirectionIsForward(node->ps.state->es_direction))
-			node->as_whichplan++;
+		{
+			if (n >= node->as_nplans - 1)
+				break;
+			n++;
+		}
 		else
-			node->as_whichplan--;
-		if (!exec_append_initialize_next(node))
 		{
-			ExecReturnTuple(&node->ps,
-							ExecClearTuple(node->ps.ps_ResultTupleSlot));
-			return;
+			if (n == 0)
+				break;
+			n--;
 		}
+	}
 
-		/* Else loop back and try to get a tuple from the new subplan */
+	/*
+	 * We are finished if reached here and no subnodes are not-ready
+	 */
+	if (n_notready == 0)
+	{
+		node->as_whichplan = n;
+		ExecReturnTuple(&node->ps,
+						ExecClearTuple(node->ps.ps_ResultTupleSlot));
 	}
 }
 
@@ -277,6 +337,8 @@ ExecReScanAppend(AppendState *node)
 	{
 		PlanState  *subnode = node->appendplans[i];
 
+		node->as_finished[i] = false;
+
 		/*
 		 * ExecReScan doesn't know about my subplans, so I have to do
 		 * changed-parameter signaling myself.
diff --git a/src/include/nodes/execnodes.h b/src/include/nodes/execnodes.h
index 9121537..e4f2bb6 100644
--- a/src/include/nodes/execnodes.h
+++ b/src/include/nodes/execnodes.h
@@ -1170,6 +1170,8 @@ typedef struct AppendState
 {
 	PlanState	ps;				/* its first field is NodeTag */
 	PlanState **appendplans;	/* array of PlanStates for my inputs */
+	bool		as_async;		/* true to allow async execution */
+	bool	   *as_finished;	/* array of the running state of subplans */
 	int			as_nplans;
 	int			as_whichplan;
 } AppendState;
-- 
1.8.3.1


----Next_Part(Wed_Jul_06_16_29_14_2016_599)--
Content-Type: Text/Plain; charset=us-ascii
Content-Transfer-Encoding: 7bit
Content-Disposition: inline; filename="gentbl.sql"

CREATE EXTENSION postgres_fdw;
CREATE TABLE pl (a int, b int);
CREATE TABLE cl1 (LIKE pl) INHERITS (pl);
CREATE TABLE cl2 (LIKE pl) INHERITS (pl);
CREATE TABLE cl3 (LIKE pl) INHERITS (pl);
CREATE TABLE cl4 (LIKE pl) INHERITS (pl);
INSERT INTO cl1 (SELECT a, a FROM generate_series(0000000, 0999999) a);
INSERT INTO cl2 (SELECT a, a FROM generate_series(1000000, 1999999) a);
INSERT INTO cl3 (SELECT a, a FROM generate_series(2000000, 2999999) a);
INSERT INTO cl4 (SELECT a, a FROM generate_series(3000000, 3999999) a);

CREATE TABLE t0 (LIKE pl);
INSERT INTO t0 (SELECT a, a FROM generate_series(0000000, 9999999) a);

CREATE SERVER sv0 FOREIGN DATA WRAPPER postgres_fdw OPTIONS (host '/tmp', dbname 'postgres');
CREATE SERVER sv1 FOREIGN DATA WRAPPER postgres_fdw OPTIONS (host '/tmp', dbname 'postgres');
CREATE SERVER sv2 FOREIGN DATA WRAPPER postgres_fdw OPTIONS (host '/tmp', dbname 'postgres');
CREATE SERVER sv3 FOREIGN DATA WRAPPER postgres_fdw OPTIONS (host '/tmp', dbname 'postgres');
CREATE SERVER sv4 FOREIGN DATA WRAPPER postgres_fdw OPTIONS (host '/tmp', dbname 'postgres');
CREATE USER MAPPING FOR public SERVER sv0;
CREATE USER MAPPING FOR public SERVER sv1;
CREATE USER MAPPING FOR public SERVER sv2;
CREATE USER MAPPING FOR public SERVER sv3;
CREATE USER MAPPING FOR public SERVER sv4;

CREATE FOREIGN TABLE ft10 (a int, b int) SERVER sv0 OPTIONS (table_name 'cl1');
CREATE FOREIGN TABLE ft20 (a int, b int) SERVER sv0 OPTIONS (table_name 'cl2');
CREATE FOREIGN TABLE ft30 (a int, b int) SERVER sv0 OPTIONS (table_name 'cl3');
CREATE FOREIGN TABLE ft40 (a int, b int) SERVER sv0 OPTIONS (table_name 'cl4');
CREATE FOREIGN TABLE ft11 (a int, b int) SERVER sv1 OPTIONS (table_name 'cl1');
CREATE FOREIGN TABLE ft22 (a int, b int) SERVER sv2 OPTIONS (table_name 'cl2');
CREATE FOREIGN TABLE ft33 (a int, b int) SERVER sv3 OPTIONS (table_name 'cl3');
CREATE FOREIGN TABLE ft44 (a int, b int) SERVER sv4 OPTIONS (table_name 'cl4');

CREATE TABLE pf0 (LIKE pl);
ALTER FOREIGN TABLE ft10 INHERIT pf0;
ALTER FOREIGN TABLE ft20 INHERIT pf0;
ALTER FOREIGN TABLE ft30 INHERIT pf0;
ALTER FOREIGN TABLE ft40 INHERIT pf0;

CREATE table pf1 (LIKE pl);
ALTER FOREIGN TABLE ft11 INHERIT pf1;
ALTER FOREIGN TABLE ft22 INHERIT pf1;
ALTER FOREIGN TABLE ft33 INHERIT pf1;
ALTER FOREIGN TABLE ft44 INHERIT pf1;

ANALYZE;

----Next_Part(Wed_Jul_06_16_29_14_2016_599)--
Content-Type: Text/Plain; charset=us-ascii
Content-Transfer-Encoding: 7bit
Content-Disposition: inline; filename="testrun.sh"

#! /bin/bash

function do_test() {
	echo $1
	for i in $(seq 1 10); do psql postgres -c "set log_min_duration_statement = 0; set client_min_messages=log; select sum(a) from $1"; done | grep LOG
}

for i in "t0" "pl" "pf0" "pf1";
do
  do_test $i
done


----Next_Part(Wed_Jul_06_16_29_14_2016_599)--
Content-Type: text/plain
Content-Disposition: inline
Content-Transfer-Encoding: 8bit
MIME-Version: 1.0


-- 
Sent via pgsql-hackers mailing list ([email protected])
To make changes to your subscription:
http://www.postgresql.org/mailpref/pgsql-hackers

----Next_Part(Wed_Jul_06_16_29_14_2016_599)----






^ permalink  raw  reply  [nested|flat] 56+ messages in thread

* [PATCH 7/7] Make Append node async-aware.
@ 2016-06-28 09:52 Kyotaro Horiguchi <[email protected]>
  0 siblings, 0 replies; 56+ messages in thread

From: Kyotaro Horiguchi @ 2016-06-28 09:52 UTC (permalink / raw)

Change append node to be capable to handle asynchronous children
properly. As soon as it receives !async_ready from a child, it moves
to the next child and if no child is ready, it sleeps until at least
one of them become ready.
---
 src/backend/executor/nodeAppend.c | 94 ++++++++++++++++++++++++++++++++-------
 src/include/nodes/execnodes.h     |  2 +
 2 files changed, 80 insertions(+), 16 deletions(-)

diff --git a/src/backend/executor/nodeAppend.c b/src/backend/executor/nodeAppend.c
index e0ce8c6..1c0d26e 100644
--- a/src/backend/executor/nodeAppend.c
+++ b/src/backend/executor/nodeAppend.c
@@ -58,6 +58,7 @@
 #include "postgres.h"
 
 #include "executor/execdebug.h"
+#include "executor/execAsync.h"
 #include "executor/nodeAppend.h"
 
 static bool exec_append_initialize_next(AppendState *appendstate);
@@ -121,6 +122,7 @@ ExecInitAppend(Append *node, EState *estate, int eflags)
 {
 	AppendState *appendstate = makeNode(AppendState);
 	PlanState **appendplanstates;
+	bool	   *finished;
 	int			nplans;
 	int			i;
 	ListCell   *lc;
@@ -134,14 +136,17 @@ ExecInitAppend(Append *node, EState *estate, int eflags)
 	nplans = list_length(node->appendplans);
 
 	appendplanstates = (PlanState **) palloc0(nplans * sizeof(PlanState *));
-
+	finished = (bool *) palloc0(nplans * sizeof(bool));
+	
 	/*
 	 * create new AppendState for our append node
 	 */
 	appendstate->ps.plan = (Plan *) node;
 	appendstate->ps.state = estate;
 	appendstate->appendplans = appendplanstates;
+	appendstate->as_finished = finished;
 	appendstate->as_nplans = nplans;
+	appendstate->as_async = ((eflags & EXEC_FLAG_BACKWARD) == 0);
 
 	/*
 	 * Miscellaneous initialization
@@ -194,49 +199,104 @@ ExecInitAppend(Append *node, EState *estate, int eflags)
 void
 ExecAppend(AppendState *node)
 {
-	for (;;)
+	int n_notready = 0;
+	PlanState  *subnode;
+	int i, n;
+
+	n = node->as_whichplan;
+
+	for (i = 0 ; i < node->as_nplans ; i++)
 	{
-		PlanState  *subnode;
 		TupleTableSlot *result;
 
+		if (node->as_async)
+		{
+			if (n >= node->as_nplans)
+				n = 0;
+
+			if (node->as_finished[n])
+			{
+				n++;
+				continue;
+			}
+		}
+
 		/*
 		 * figure out which subplan we are currently processing
 		 */
-		subnode = node->appendplans[node->as_whichplan];
+		subnode = node->appendplans[n];
 
 		/*
-		 * get a tuple from the subplan
+		 * execute the subplan to get a result if it is not ready yet
 		 */
-		result = ExecProcNode(subnode);
+		if (!subnode->result_ready)
+			ExecExecuteNode(subnode);
+
+		/*
+		 * This subnode is not ready yet when asynchrony is not allowed,
+		 * immediately wait for this subnode.
+		 */
+		if (!subnode->result_ready)
+		{
+			if (node->as_async)
+			{
+				n_notready++;
+				n++;
+				continue;
+			}
+			ExecAsyncWaitForNode(subnode);
+		}
+
+		Assert(subnode->result_ready);
+
+		result = ExecConsumeResult((PlanState *)subnode);
 
 		if (!TupIsNull(result))
 		{
 			/*
-			 * If the subplan gave us something then return it as-is. We do
-			 * NOT make use of the result slot that was set up in
-			 * ExecInitAppend; there's no need for it.
+			 * If the subplan gave us something then return it
+			 * as-is. We do NOT make use of the result slot that was
+			 * set up in ExecInitAppend; there's no need for it.
 			 */
+			node->as_whichplan = n;
 			ExecReturnTuple(&node->ps, result);
 			return;
 		}
 
+		/* Tuple has been exhausted on this subnode */
+		if (node->as_async)
+		{
+			node->as_finished[n++] = true;
+			continue;
+		}
+
 		/*
 		 * Go on to the "next" subplan in the appropriate direction. If no
 		 * more subplans, return the empty slot set up for us by
 		 * ExecInitAppend.
 		 */
 		if (ScanDirectionIsForward(node->ps.state->es_direction))
-			node->as_whichplan++;
+		{
+			if (n >= node->as_nplans - 1)
+				break;
+			n++;
+		}
 		else
-			node->as_whichplan--;
-		if (!exec_append_initialize_next(node))
 		{
-			ExecReturnTuple(&node->ps,
-							ExecClearTuple(node->ps.ps_ResultTupleSlot));
-			return;
+			if (n == 0)
+				break;
+			n--;
 		}
+	}
 
-		/* Else loop back and try to get a tuple from the new subplan */
+	/*
+	 * We are finished if reached here and no subnodes are not-ready
+	 */
+	if (n_notready == 0)
+	{
+		node->as_whichplan = n;
+		ExecReturnTuple(&node->ps,
+						ExecClearTuple(node->ps.ps_ResultTupleSlot));
 	}
 }
 
@@ -277,6 +337,8 @@ ExecReScanAppend(AppendState *node)
 	{
 		PlanState  *subnode = node->appendplans[i];
 
+		node->as_finished[i] = false;
+
 		/*
 		 * ExecReScan doesn't know about my subplans, so I have to do
 		 * changed-parameter signaling myself.
diff --git a/src/include/nodes/execnodes.h b/src/include/nodes/execnodes.h
index 9121537..e4f2bb6 100644
--- a/src/include/nodes/execnodes.h
+++ b/src/include/nodes/execnodes.h
@@ -1170,6 +1170,8 @@ typedef struct AppendState
 {
 	PlanState	ps;				/* its first field is NodeTag */
 	PlanState **appendplans;	/* array of PlanStates for my inputs */
+	bool		as_async;		/* true to allow async execution */
+	bool	   *as_finished;	/* array of the running state of subplans */
 	int			as_nplans;
 	int			as_whichplan;
 } AppendState;
-- 
1.8.3.1


----Next_Part(Wed_Jul_06_16_29_14_2016_599)--
Content-Type: Text/Plain; charset=us-ascii
Content-Transfer-Encoding: 7bit
Content-Disposition: inline; filename="gentbl.sql"

CREATE EXTENSION postgres_fdw;
CREATE TABLE pl (a int, b int);
CREATE TABLE cl1 (LIKE pl) INHERITS (pl);
CREATE TABLE cl2 (LIKE pl) INHERITS (pl);
CREATE TABLE cl3 (LIKE pl) INHERITS (pl);
CREATE TABLE cl4 (LIKE pl) INHERITS (pl);
INSERT INTO cl1 (SELECT a, a FROM generate_series(0000000, 0999999) a);
INSERT INTO cl2 (SELECT a, a FROM generate_series(1000000, 1999999) a);
INSERT INTO cl3 (SELECT a, a FROM generate_series(2000000, 2999999) a);
INSERT INTO cl4 (SELECT a, a FROM generate_series(3000000, 3999999) a);

CREATE TABLE t0 (LIKE pl);
INSERT INTO t0 (SELECT a, a FROM generate_series(0000000, 9999999) a);

CREATE SERVER sv0 FOREIGN DATA WRAPPER postgres_fdw OPTIONS (host '/tmp', dbname 'postgres');
CREATE SERVER sv1 FOREIGN DATA WRAPPER postgres_fdw OPTIONS (host '/tmp', dbname 'postgres');
CREATE SERVER sv2 FOREIGN DATA WRAPPER postgres_fdw OPTIONS (host '/tmp', dbname 'postgres');
CREATE SERVER sv3 FOREIGN DATA WRAPPER postgres_fdw OPTIONS (host '/tmp', dbname 'postgres');
CREATE SERVER sv4 FOREIGN DATA WRAPPER postgres_fdw OPTIONS (host '/tmp', dbname 'postgres');
CREATE USER MAPPING FOR public SERVER sv0;
CREATE USER MAPPING FOR public SERVER sv1;
CREATE USER MAPPING FOR public SERVER sv2;
CREATE USER MAPPING FOR public SERVER sv3;
CREATE USER MAPPING FOR public SERVER sv4;

CREATE FOREIGN TABLE ft10 (a int, b int) SERVER sv0 OPTIONS (table_name 'cl1');
CREATE FOREIGN TABLE ft20 (a int, b int) SERVER sv0 OPTIONS (table_name 'cl2');
CREATE FOREIGN TABLE ft30 (a int, b int) SERVER sv0 OPTIONS (table_name 'cl3');
CREATE FOREIGN TABLE ft40 (a int, b int) SERVER sv0 OPTIONS (table_name 'cl4');
CREATE FOREIGN TABLE ft11 (a int, b int) SERVER sv1 OPTIONS (table_name 'cl1');
CREATE FOREIGN TABLE ft22 (a int, b int) SERVER sv2 OPTIONS (table_name 'cl2');
CREATE FOREIGN TABLE ft33 (a int, b int) SERVER sv3 OPTIONS (table_name 'cl3');
CREATE FOREIGN TABLE ft44 (a int, b int) SERVER sv4 OPTIONS (table_name 'cl4');

CREATE TABLE pf0 (LIKE pl);
ALTER FOREIGN TABLE ft10 INHERIT pf0;
ALTER FOREIGN TABLE ft20 INHERIT pf0;
ALTER FOREIGN TABLE ft30 INHERIT pf0;
ALTER FOREIGN TABLE ft40 INHERIT pf0;

CREATE table pf1 (LIKE pl);
ALTER FOREIGN TABLE ft11 INHERIT pf1;
ALTER FOREIGN TABLE ft22 INHERIT pf1;
ALTER FOREIGN TABLE ft33 INHERIT pf1;
ALTER FOREIGN TABLE ft44 INHERIT pf1;

ANALYZE;

----Next_Part(Wed_Jul_06_16_29_14_2016_599)--
Content-Type: Text/Plain; charset=us-ascii
Content-Transfer-Encoding: 7bit
Content-Disposition: inline; filename="testrun.sh"

#! /bin/bash

function do_test() {
	echo $1
	for i in $(seq 1 10); do psql postgres -c "set log_min_duration_statement = 0; set client_min_messages=log; select sum(a) from $1"; done | grep LOG
}

for i in "t0" "pl" "pf0" "pf1";
do
  do_test $i
done


----Next_Part(Wed_Jul_06_16_29_14_2016_599)--
Content-Type: text/plain
Content-Disposition: inline
Content-Transfer-Encoding: 8bit
MIME-Version: 1.0


-- 
Sent via pgsql-hackers mailing list ([email protected])
To make changes to your subscription:
http://www.postgresql.org/mailpref/pgsql-hackers

----Next_Part(Wed_Jul_06_16_29_14_2016_599)----






^ permalink  raw  reply  [nested|flat] 56+ messages in thread

* [PATCH 7/7] Make Append node async-aware.
@ 2016-06-28 09:52 Kyotaro Horiguchi <[email protected]>
  0 siblings, 0 replies; 56+ messages in thread

From: Kyotaro Horiguchi @ 2016-06-28 09:52 UTC (permalink / raw)

Change append node to be capable to handle asynchronous children
properly. As soon as it receives !async_ready from a child, it moves
to the next child and if no child is ready, it sleeps until at least
one of them become ready.
---
 src/backend/executor/nodeAppend.c | 67 +++++++++++++++++++++++++++++++++++++--
 src/include/nodes/execnodes.h     |  2 ++
 2 files changed, 67 insertions(+), 2 deletions(-)

diff --git a/src/backend/executor/nodeAppend.c b/src/backend/executor/nodeAppend.c
index e0ce8c6..9a4063a 100644
--- a/src/backend/executor/nodeAppend.c
+++ b/src/backend/executor/nodeAppend.c
@@ -58,6 +58,7 @@
 #include "postgres.h"
 
 #include "executor/execdebug.h"
+#include "executor/execAsync.h"
 #include "executor/nodeAppend.h"
 
 static bool exec_append_initialize_next(AppendState *appendstate);
@@ -121,6 +122,7 @@ ExecInitAppend(Append *node, EState *estate, int eflags)
 {
 	AppendState *appendstate = makeNode(AppendState);
 	PlanState **appendplanstates;
+	bool	   *finished;
 	int			nplans;
 	int			i;
 	ListCell   *lc;
@@ -134,14 +136,17 @@ ExecInitAppend(Append *node, EState *estate, int eflags)
 	nplans = list_length(node->appendplans);
 
 	appendplanstates = (PlanState **) palloc0(nplans * sizeof(PlanState *));
-
+	finished = (bool *) palloc0(nplans * sizeof(bool));
+	
 	/*
 	 * create new AppendState for our append node
 	 */
 	appendstate->ps.plan = (Plan *) node;
 	appendstate->ps.state = estate;
 	appendstate->appendplans = appendplanstates;
+	appendstate->as_finished = finished;
 	appendstate->as_nplans = nplans;
+	appendstate->as_async = ((eflags & EXEC_FLAG_BACKWARD) == 0);
 
 	/*
 	 * Miscellaneous initialization
@@ -194,6 +199,8 @@ ExecInitAppend(Append *node, EState *estate, int eflags)
 void
 ExecAppend(AppendState *node)
 {
+	int stopplan = node->as_whichplan;
+
 	for (;;)
 	{
 		PlanState  *subnode;
@@ -207,7 +214,36 @@ ExecAppend(AppendState *node)
 		/*
 		 * get a tuple from the subplan
 		 */
-		result = ExecProcNode(subnode);
+		Assert(!node->as_finished[node->as_whichplan]);
+
+		if (!subnode->result_ready)
+			ExecExecuteNode(subnode);
+
+		if (!subnode->result_ready)
+		{
+			if (node->as_async)
+			{
+				/* Move to the next living node */
+				do
+				{
+					node->as_whichplan = 
+						(node->as_whichplan + 1) %  node->as_nplans;
+				} while (node->as_whichplan != stopplan &&
+						 node->as_finished[node->as_whichplan]);
+
+				/* No node is ready yet, return as not-ready */
+				if (node->as_whichplan == stopplan)
+					return;
+
+				/* Try the next node */
+				continue;
+			}
+
+			/* If not async, immediately wait for this subnode */
+			ExecAsyncWaitForNode(subnode);
+		}				
+
+		result = ExecConsumeResult((PlanState *) subnode);
 
 		if (!TupIsNull(result))
 		{
@@ -220,6 +256,31 @@ ExecAppend(AppendState *node)
 			return;
 		}
 
+		if (node->as_async)
+		{
+			node->as_finished[node->as_whichplan] = true;
+			stopplan = node->as_whichplan;
+
+			/* Find the next living subnode */
+			do
+			{
+				node->as_whichplan =
+					(node->as_whichplan + 1) % node->as_nplans;
+			} while (node->as_whichplan != stopplan &&
+					 node->as_finished[node->as_whichplan]);
+
+			if (node->as_whichplan != stopplan)
+			{
+				stopplan = node->as_whichplan;
+				continue;
+			}
+
+			/* All subnodes are exhausted. Finish this node. */
+			ExecReturnTuple(&node->ps,
+							ExecClearTuple(node->ps.ps_ResultTupleSlot));
+			return;
+		}
+
 		/*
 		 * Go on to the "next" subplan in the appropriate direction. If no
 		 * more subplans, return the empty slot set up for us by
@@ -277,6 +338,8 @@ ExecReScanAppend(AppendState *node)
 	{
 		PlanState  *subnode = node->appendplans[i];
 
+		node->as_finished[i] = false;
+
 		/*
 		 * ExecReScan doesn't know about my subplans, so I have to do
 		 * changed-parameter signaling myself.
diff --git a/src/include/nodes/execnodes.h b/src/include/nodes/execnodes.h
index b72decc..b0a86c5 100644
--- a/src/include/nodes/execnodes.h
+++ b/src/include/nodes/execnodes.h
@@ -1177,6 +1177,8 @@ typedef struct AppendState
 {
 	PlanState	ps;				/* its first field is NodeTag */
 	PlanState **appendplans;	/* array of PlanStates for my inputs */
+	bool		as_async;		/* true to allow async execution */
+	bool	   *as_finished;	/* array of the running state of subplans */
 	int			as_nplans;
 	int			as_whichplan;
 } AppendState;
-- 
2.9.2


----Next_Part(Mon_Aug_29_17_08_36_2016_213)--
Content-Type: text/plain
Content-Disposition: inline
Content-Transfer-Encoding: 8bit
MIME-Version: 1.0


-- 
Sent via pgsql-hackers mailing list ([email protected])
To make changes to your subscription:
http://www.postgresql.org/mailpref/pgsql-hackers

----Next_Part(Mon_Aug_29_17_08_36_2016_213)----






^ permalink  raw  reply  [nested|flat] 56+ messages in thread

* [PATCH 7/7] Make Append node async-aware.
@ 2016-06-28 09:52 Kyotaro Horiguchi <[email protected]>
  0 siblings, 0 replies; 56+ messages in thread

From: Kyotaro Horiguchi @ 2016-06-28 09:52 UTC (permalink / raw)

Change append node to be capable to handle asynchronous children
properly. As soon as it receives !async_ready from a child, it moves
to the next child and if no child is ready, it sleeps until at least
one of them become ready.
---
 src/backend/executor/nodeAppend.c | 94 ++++++++++++++++++++++++++++++++-------
 src/include/nodes/execnodes.h     |  2 +
 2 files changed, 80 insertions(+), 16 deletions(-)

diff --git a/src/backend/executor/nodeAppend.c b/src/backend/executor/nodeAppend.c
index e0ce8c6..1c0d26e 100644
--- a/src/backend/executor/nodeAppend.c
+++ b/src/backend/executor/nodeAppend.c
@@ -58,6 +58,7 @@
 #include "postgres.h"
 
 #include "executor/execdebug.h"
+#include "executor/execAsync.h"
 #include "executor/nodeAppend.h"
 
 static bool exec_append_initialize_next(AppendState *appendstate);
@@ -121,6 +122,7 @@ ExecInitAppend(Append *node, EState *estate, int eflags)
 {
 	AppendState *appendstate = makeNode(AppendState);
 	PlanState **appendplanstates;
+	bool	   *finished;
 	int			nplans;
 	int			i;
 	ListCell   *lc;
@@ -134,14 +136,17 @@ ExecInitAppend(Append *node, EState *estate, int eflags)
 	nplans = list_length(node->appendplans);
 
 	appendplanstates = (PlanState **) palloc0(nplans * sizeof(PlanState *));
-
+	finished = (bool *) palloc0(nplans * sizeof(bool));
+	
 	/*
 	 * create new AppendState for our append node
 	 */
 	appendstate->ps.plan = (Plan *) node;
 	appendstate->ps.state = estate;
 	appendstate->appendplans = appendplanstates;
+	appendstate->as_finished = finished;
 	appendstate->as_nplans = nplans;
+	appendstate->as_async = ((eflags & EXEC_FLAG_BACKWARD) == 0);
 
 	/*
 	 * Miscellaneous initialization
@@ -194,49 +199,104 @@ ExecInitAppend(Append *node, EState *estate, int eflags)
 void
 ExecAppend(AppendState *node)
 {
-	for (;;)
+	int n_notready = 0;
+	PlanState  *subnode;
+	int i, n;
+
+	n = node->as_whichplan;
+
+	for (i = 0 ; i < node->as_nplans ; i++)
 	{
-		PlanState  *subnode;
 		TupleTableSlot *result;
 
+		if (node->as_async)
+		{
+			if (n >= node->as_nplans)
+				n = 0;
+
+			if (node->as_finished[n])
+			{
+				n++;
+				continue;
+			}
+		}
+
 		/*
 		 * figure out which subplan we are currently processing
 		 */
-		subnode = node->appendplans[node->as_whichplan];
+		subnode = node->appendplans[n];
 
 		/*
-		 * get a tuple from the subplan
+		 * execute the subplan to get a result if it is not ready yet
 		 */
-		result = ExecProcNode(subnode);
+		if (!subnode->result_ready)
+			ExecExecuteNode(subnode);
+
+		/*
+		 * This subnode is not ready yet when asynchrony is not allowed,
+		 * immediately wait for this subnode.
+		 */
+		if (!subnode->result_ready)
+		{
+			if (node->as_async)
+			{
+				n_notready++;
+				n++;
+				continue;
+			}
+			ExecAsyncWaitForNode(subnode);
+		}
+
+		Assert(subnode->result_ready);
+
+		result = ExecConsumeResult((PlanState *)subnode);
 
 		if (!TupIsNull(result))
 		{
 			/*
-			 * If the subplan gave us something then return it as-is. We do
-			 * NOT make use of the result slot that was set up in
-			 * ExecInitAppend; there's no need for it.
+			 * If the subplan gave us something then return it
+			 * as-is. We do NOT make use of the result slot that was
+			 * set up in ExecInitAppend; there's no need for it.
 			 */
+			node->as_whichplan = n;
 			ExecReturnTuple(&node->ps, result);
 			return;
 		}
 
+		/* Tuple has been exhausted on this subnode */
+		if (node->as_async)
+		{
+			node->as_finished[n++] = true;
+			continue;
+		}
+
 		/*
 		 * Go on to the "next" subplan in the appropriate direction. If no
 		 * more subplans, return the empty slot set up for us by
 		 * ExecInitAppend.
 		 */
 		if (ScanDirectionIsForward(node->ps.state->es_direction))
-			node->as_whichplan++;
+		{
+			if (n >= node->as_nplans - 1)
+				break;
+			n++;
+		}
 		else
-			node->as_whichplan--;
-		if (!exec_append_initialize_next(node))
 		{
-			ExecReturnTuple(&node->ps,
-							ExecClearTuple(node->ps.ps_ResultTupleSlot));
-			return;
+			if (n == 0)
+				break;
+			n--;
 		}
+	}
 
-		/* Else loop back and try to get a tuple from the new subplan */
+	/*
+	 * We are finished if reached here and no subnodes are not-ready
+	 */
+	if (n_notready == 0)
+	{
+		node->as_whichplan = n;
+		ExecReturnTuple(&node->ps,
+						ExecClearTuple(node->ps.ps_ResultTupleSlot));
 	}
 }
 
@@ -277,6 +337,8 @@ ExecReScanAppend(AppendState *node)
 	{
 		PlanState  *subnode = node->appendplans[i];
 
+		node->as_finished[i] = false;
+
 		/*
 		 * ExecReScan doesn't know about my subplans, so I have to do
 		 * changed-parameter signaling myself.
diff --git a/src/include/nodes/execnodes.h b/src/include/nodes/execnodes.h
index 9121537..e4f2bb6 100644
--- a/src/include/nodes/execnodes.h
+++ b/src/include/nodes/execnodes.h
@@ -1170,6 +1170,8 @@ typedef struct AppendState
 {
 	PlanState	ps;				/* its first field is NodeTag */
 	PlanState **appendplans;	/* array of PlanStates for my inputs */
+	bool		as_async;		/* true to allow async execution */
+	bool	   *as_finished;	/* array of the running state of subplans */
 	int			as_nplans;
 	int			as_whichplan;
 } AppendState;
-- 
1.8.3.1


----Next_Part(Wed_Jul_06_16_29_14_2016_599)--
Content-Type: Text/Plain; charset=us-ascii
Content-Transfer-Encoding: 7bit
Content-Disposition: inline; filename="gentbl.sql"

CREATE EXTENSION postgres_fdw;
CREATE TABLE pl (a int, b int);
CREATE TABLE cl1 (LIKE pl) INHERITS (pl);
CREATE TABLE cl2 (LIKE pl) INHERITS (pl);
CREATE TABLE cl3 (LIKE pl) INHERITS (pl);
CREATE TABLE cl4 (LIKE pl) INHERITS (pl);
INSERT INTO cl1 (SELECT a, a FROM generate_series(0000000, 0999999) a);
INSERT INTO cl2 (SELECT a, a FROM generate_series(1000000, 1999999) a);
INSERT INTO cl3 (SELECT a, a FROM generate_series(2000000, 2999999) a);
INSERT INTO cl4 (SELECT a, a FROM generate_series(3000000, 3999999) a);

CREATE TABLE t0 (LIKE pl);
INSERT INTO t0 (SELECT a, a FROM generate_series(0000000, 9999999) a);

CREATE SERVER sv0 FOREIGN DATA WRAPPER postgres_fdw OPTIONS (host '/tmp', dbname 'postgres');
CREATE SERVER sv1 FOREIGN DATA WRAPPER postgres_fdw OPTIONS (host '/tmp', dbname 'postgres');
CREATE SERVER sv2 FOREIGN DATA WRAPPER postgres_fdw OPTIONS (host '/tmp', dbname 'postgres');
CREATE SERVER sv3 FOREIGN DATA WRAPPER postgres_fdw OPTIONS (host '/tmp', dbname 'postgres');
CREATE SERVER sv4 FOREIGN DATA WRAPPER postgres_fdw OPTIONS (host '/tmp', dbname 'postgres');
CREATE USER MAPPING FOR public SERVER sv0;
CREATE USER MAPPING FOR public SERVER sv1;
CREATE USER MAPPING FOR public SERVER sv2;
CREATE USER MAPPING FOR public SERVER sv3;
CREATE USER MAPPING FOR public SERVER sv4;

CREATE FOREIGN TABLE ft10 (a int, b int) SERVER sv0 OPTIONS (table_name 'cl1');
CREATE FOREIGN TABLE ft20 (a int, b int) SERVER sv0 OPTIONS (table_name 'cl2');
CREATE FOREIGN TABLE ft30 (a int, b int) SERVER sv0 OPTIONS (table_name 'cl3');
CREATE FOREIGN TABLE ft40 (a int, b int) SERVER sv0 OPTIONS (table_name 'cl4');
CREATE FOREIGN TABLE ft11 (a int, b int) SERVER sv1 OPTIONS (table_name 'cl1');
CREATE FOREIGN TABLE ft22 (a int, b int) SERVER sv2 OPTIONS (table_name 'cl2');
CREATE FOREIGN TABLE ft33 (a int, b int) SERVER sv3 OPTIONS (table_name 'cl3');
CREATE FOREIGN TABLE ft44 (a int, b int) SERVER sv4 OPTIONS (table_name 'cl4');

CREATE TABLE pf0 (LIKE pl);
ALTER FOREIGN TABLE ft10 INHERIT pf0;
ALTER FOREIGN TABLE ft20 INHERIT pf0;
ALTER FOREIGN TABLE ft30 INHERIT pf0;
ALTER FOREIGN TABLE ft40 INHERIT pf0;

CREATE table pf1 (LIKE pl);
ALTER FOREIGN TABLE ft11 INHERIT pf1;
ALTER FOREIGN TABLE ft22 INHERIT pf1;
ALTER FOREIGN TABLE ft33 INHERIT pf1;
ALTER FOREIGN TABLE ft44 INHERIT pf1;

ANALYZE;

----Next_Part(Wed_Jul_06_16_29_14_2016_599)--
Content-Type: Text/Plain; charset=us-ascii
Content-Transfer-Encoding: 7bit
Content-Disposition: inline; filename="testrun.sh"

#! /bin/bash

function do_test() {
	echo $1
	for i in $(seq 1 10); do psql postgres -c "set log_min_duration_statement = 0; set client_min_messages=log; select sum(a) from $1"; done | grep LOG
}

for i in "t0" "pl" "pf0" "pf1";
do
  do_test $i
done


----Next_Part(Wed_Jul_06_16_29_14_2016_599)--
Content-Type: text/plain
Content-Disposition: inline
Content-Transfer-Encoding: 8bit
MIME-Version: 1.0


-- 
Sent via pgsql-hackers mailing list ([email protected])
To make changes to your subscription:
http://www.postgresql.org/mailpref/pgsql-hackers

----Next_Part(Wed_Jul_06_16_29_14_2016_599)----






^ permalink  raw  reply  [nested|flat] 56+ messages in thread

* [PATCH 7/7] Make Append node async-aware.
@ 2016-06-28 09:52 Kyotaro Horiguchi <[email protected]>
  0 siblings, 0 replies; 56+ messages in thread

From: Kyotaro Horiguchi @ 2016-06-28 09:52 UTC (permalink / raw)

Change append node to be capable to handle asynchronous children
properly. As soon as it receives !async_ready from a child, it moves
to the next child and if no child is ready, it sleeps until at least
one of them become ready.
---
 src/backend/executor/nodeAppend.c | 67 +++++++++++++++++++++++++++++++++++++--
 src/include/nodes/execnodes.h     |  2 ++
 2 files changed, 67 insertions(+), 2 deletions(-)

diff --git a/src/backend/executor/nodeAppend.c b/src/backend/executor/nodeAppend.c
index e0ce8c6..9a4063a 100644
--- a/src/backend/executor/nodeAppend.c
+++ b/src/backend/executor/nodeAppend.c
@@ -58,6 +58,7 @@
 #include "postgres.h"
 
 #include "executor/execdebug.h"
+#include "executor/execAsync.h"
 #include "executor/nodeAppend.h"
 
 static bool exec_append_initialize_next(AppendState *appendstate);
@@ -121,6 +122,7 @@ ExecInitAppend(Append *node, EState *estate, int eflags)
 {
 	AppendState *appendstate = makeNode(AppendState);
 	PlanState **appendplanstates;
+	bool	   *finished;
 	int			nplans;
 	int			i;
 	ListCell   *lc;
@@ -134,14 +136,17 @@ ExecInitAppend(Append *node, EState *estate, int eflags)
 	nplans = list_length(node->appendplans);
 
 	appendplanstates = (PlanState **) palloc0(nplans * sizeof(PlanState *));
-
+	finished = (bool *) palloc0(nplans * sizeof(bool));
+	
 	/*
 	 * create new AppendState for our append node
 	 */
 	appendstate->ps.plan = (Plan *) node;
 	appendstate->ps.state = estate;
 	appendstate->appendplans = appendplanstates;
+	appendstate->as_finished = finished;
 	appendstate->as_nplans = nplans;
+	appendstate->as_async = ((eflags & EXEC_FLAG_BACKWARD) == 0);
 
 	/*
 	 * Miscellaneous initialization
@@ -194,6 +199,8 @@ ExecInitAppend(Append *node, EState *estate, int eflags)
 void
 ExecAppend(AppendState *node)
 {
+	int stopplan = node->as_whichplan;
+
 	for (;;)
 	{
 		PlanState  *subnode;
@@ -207,7 +214,36 @@ ExecAppend(AppendState *node)
 		/*
 		 * get a tuple from the subplan
 		 */
-		result = ExecProcNode(subnode);
+		Assert(!node->as_finished[node->as_whichplan]);
+
+		if (!subnode->result_ready)
+			ExecExecuteNode(subnode);
+
+		if (!subnode->result_ready)
+		{
+			if (node->as_async)
+			{
+				/* Move to the next living node */
+				do
+				{
+					node->as_whichplan = 
+						(node->as_whichplan + 1) %  node->as_nplans;
+				} while (node->as_whichplan != stopplan &&
+						 node->as_finished[node->as_whichplan]);
+
+				/* No node is ready yet, return as not-ready */
+				if (node->as_whichplan == stopplan)
+					return;
+
+				/* Try the next node */
+				continue;
+			}
+
+			/* If not async, immediately wait for this subnode */
+			ExecAsyncWaitForNode(subnode);
+		}				
+
+		result = ExecConsumeResult((PlanState *) subnode);
 
 		if (!TupIsNull(result))
 		{
@@ -220,6 +256,31 @@ ExecAppend(AppendState *node)
 			return;
 		}
 
+		if (node->as_async)
+		{
+			node->as_finished[node->as_whichplan] = true;
+			stopplan = node->as_whichplan;
+
+			/* Find the next living subnode */
+			do
+			{
+				node->as_whichplan =
+					(node->as_whichplan + 1) % node->as_nplans;
+			} while (node->as_whichplan != stopplan &&
+					 node->as_finished[node->as_whichplan]);
+
+			if (node->as_whichplan != stopplan)
+			{
+				stopplan = node->as_whichplan;
+				continue;
+			}
+
+			/* All subnodes are exhausted. Finish this node. */
+			ExecReturnTuple(&node->ps,
+							ExecClearTuple(node->ps.ps_ResultTupleSlot));
+			return;
+		}
+
 		/*
 		 * Go on to the "next" subplan in the appropriate direction. If no
 		 * more subplans, return the empty slot set up for us by
@@ -277,6 +338,8 @@ ExecReScanAppend(AppendState *node)
 	{
 		PlanState  *subnode = node->appendplans[i];
 
+		node->as_finished[i] = false;
+
 		/*
 		 * ExecReScan doesn't know about my subplans, so I have to do
 		 * changed-parameter signaling myself.
diff --git a/src/include/nodes/execnodes.h b/src/include/nodes/execnodes.h
index b72decc..b0a86c5 100644
--- a/src/include/nodes/execnodes.h
+++ b/src/include/nodes/execnodes.h
@@ -1177,6 +1177,8 @@ typedef struct AppendState
 {
 	PlanState	ps;				/* its first field is NodeTag */
 	PlanState **appendplans;	/* array of PlanStates for my inputs */
+	bool		as_async;		/* true to allow async execution */
+	bool	   *as_finished;	/* array of the running state of subplans */
 	int			as_nplans;
 	int			as_whichplan;
 } AppendState;
-- 
2.9.2


----Next_Part(Mon_Aug_29_17_08_36_2016_213)--
Content-Type: text/plain
Content-Disposition: inline
Content-Transfer-Encoding: 8bit
MIME-Version: 1.0


-- 
Sent via pgsql-hackers mailing list ([email protected])
To make changes to your subscription:
http://www.postgresql.org/mailpref/pgsql-hackers

----Next_Part(Mon_Aug_29_17_08_36_2016_213)----






^ permalink  raw  reply  [nested|flat] 56+ messages in thread

* [PATCH 7/7] Make Append node async-aware.
@ 2016-06-28 09:52 Kyotaro Horiguchi <[email protected]>
  0 siblings, 0 replies; 56+ messages in thread

From: Kyotaro Horiguchi @ 2016-06-28 09:52 UTC (permalink / raw)

Change append node to be capable to handle asynchronous children
properly. As soon as it receives !async_ready from a child, it moves
to the next child and if no child is ready, it sleeps until at least
one of them become ready.
---
 src/backend/executor/nodeAppend.c | 67 +++++++++++++++++++++++++++++++++++++--
 src/include/nodes/execnodes.h     |  2 ++
 2 files changed, 67 insertions(+), 2 deletions(-)

diff --git a/src/backend/executor/nodeAppend.c b/src/backend/executor/nodeAppend.c
index e0ce8c6..9a4063a 100644
--- a/src/backend/executor/nodeAppend.c
+++ b/src/backend/executor/nodeAppend.c
@@ -58,6 +58,7 @@
 #include "postgres.h"
 
 #include "executor/execdebug.h"
+#include "executor/execAsync.h"
 #include "executor/nodeAppend.h"
 
 static bool exec_append_initialize_next(AppendState *appendstate);
@@ -121,6 +122,7 @@ ExecInitAppend(Append *node, EState *estate, int eflags)
 {
 	AppendState *appendstate = makeNode(AppendState);
 	PlanState **appendplanstates;
+	bool	   *finished;
 	int			nplans;
 	int			i;
 	ListCell   *lc;
@@ -134,14 +136,17 @@ ExecInitAppend(Append *node, EState *estate, int eflags)
 	nplans = list_length(node->appendplans);
 
 	appendplanstates = (PlanState **) palloc0(nplans * sizeof(PlanState *));
-
+	finished = (bool *) palloc0(nplans * sizeof(bool));
+	
 	/*
 	 * create new AppendState for our append node
 	 */
 	appendstate->ps.plan = (Plan *) node;
 	appendstate->ps.state = estate;
 	appendstate->appendplans = appendplanstates;
+	appendstate->as_finished = finished;
 	appendstate->as_nplans = nplans;
+	appendstate->as_async = ((eflags & EXEC_FLAG_BACKWARD) == 0);
 
 	/*
 	 * Miscellaneous initialization
@@ -194,6 +199,8 @@ ExecInitAppend(Append *node, EState *estate, int eflags)
 void
 ExecAppend(AppendState *node)
 {
+	int stopplan = node->as_whichplan;
+
 	for (;;)
 	{
 		PlanState  *subnode;
@@ -207,7 +214,36 @@ ExecAppend(AppendState *node)
 		/*
 		 * get a tuple from the subplan
 		 */
-		result = ExecProcNode(subnode);
+		Assert(!node->as_finished[node->as_whichplan]);
+
+		if (!subnode->result_ready)
+			ExecExecuteNode(subnode);
+
+		if (!subnode->result_ready)
+		{
+			if (node->as_async)
+			{
+				/* Move to the next living node */
+				do
+				{
+					node->as_whichplan = 
+						(node->as_whichplan + 1) %  node->as_nplans;
+				} while (node->as_whichplan != stopplan &&
+						 node->as_finished[node->as_whichplan]);
+
+				/* No node is ready yet, return as not-ready */
+				if (node->as_whichplan == stopplan)
+					return;
+
+				/* Try the next node */
+				continue;
+			}
+
+			/* If not async, immediately wait for this subnode */
+			ExecAsyncWaitForNode(subnode);
+		}				
+
+		result = ExecConsumeResult((PlanState *) subnode);
 
 		if (!TupIsNull(result))
 		{
@@ -220,6 +256,31 @@ ExecAppend(AppendState *node)
 			return;
 		}
 
+		if (node->as_async)
+		{
+			node->as_finished[node->as_whichplan] = true;
+			stopplan = node->as_whichplan;
+
+			/* Find the next living subnode */
+			do
+			{
+				node->as_whichplan =
+					(node->as_whichplan + 1) % node->as_nplans;
+			} while (node->as_whichplan != stopplan &&
+					 node->as_finished[node->as_whichplan]);
+
+			if (node->as_whichplan != stopplan)
+			{
+				stopplan = node->as_whichplan;
+				continue;
+			}
+
+			/* All subnodes are exhausted. Finish this node. */
+			ExecReturnTuple(&node->ps,
+							ExecClearTuple(node->ps.ps_ResultTupleSlot));
+			return;
+		}
+
 		/*
 		 * Go on to the "next" subplan in the appropriate direction. If no
 		 * more subplans, return the empty slot set up for us by
@@ -277,6 +338,8 @@ ExecReScanAppend(AppendState *node)
 	{
 		PlanState  *subnode = node->appendplans[i];
 
+		node->as_finished[i] = false;
+
 		/*
 		 * ExecReScan doesn't know about my subplans, so I have to do
 		 * changed-parameter signaling myself.
diff --git a/src/include/nodes/execnodes.h b/src/include/nodes/execnodes.h
index b72decc..b0a86c5 100644
--- a/src/include/nodes/execnodes.h
+++ b/src/include/nodes/execnodes.h
@@ -1177,6 +1177,8 @@ typedef struct AppendState
 {
 	PlanState	ps;				/* its first field is NodeTag */
 	PlanState **appendplans;	/* array of PlanStates for my inputs */
+	bool		as_async;		/* true to allow async execution */
+	bool	   *as_finished;	/* array of the running state of subplans */
 	int			as_nplans;
 	int			as_whichplan;
 } AppendState;
-- 
1.8.3.1


----Next_Part(Thu_Jul_21_18_50_07_2016_597)--
Content-Type: text/plain
Content-Disposition: inline
Content-Transfer-Encoding: 8bit
MIME-Version: 1.0


-- 
Sent via pgsql-hackers mailing list ([email protected])
To make changes to your subscription:
http://www.postgresql.org/mailpref/pgsql-hackers

----Next_Part(Thu_Jul_21_18_50_07_2016_597)----






^ permalink  raw  reply  [nested|flat] 56+ messages in thread

* [PATCH 7/7] Make Append node async-aware.
@ 2016-06-28 09:52 Kyotaro Horiguchi <[email protected]>
  0 siblings, 0 replies; 56+ messages in thread

From: Kyotaro Horiguchi @ 2016-06-28 09:52 UTC (permalink / raw)

Change append node to be capable to handle asynchronous children
properly. As soon as it receives !async_ready from a child, it moves
to the next child and if no child is ready, it sleeps until at least
one of them become ready.
---
 src/backend/executor/nodeAppend.c | 94 ++++++++++++++++++++++++++++++++-------
 src/include/nodes/execnodes.h     |  2 +
 2 files changed, 80 insertions(+), 16 deletions(-)

diff --git a/src/backend/executor/nodeAppend.c b/src/backend/executor/nodeAppend.c
index e0ce8c6..1c0d26e 100644
--- a/src/backend/executor/nodeAppend.c
+++ b/src/backend/executor/nodeAppend.c
@@ -58,6 +58,7 @@
 #include "postgres.h"
 
 #include "executor/execdebug.h"
+#include "executor/execAsync.h"
 #include "executor/nodeAppend.h"
 
 static bool exec_append_initialize_next(AppendState *appendstate);
@@ -121,6 +122,7 @@ ExecInitAppend(Append *node, EState *estate, int eflags)
 {
 	AppendState *appendstate = makeNode(AppendState);
 	PlanState **appendplanstates;
+	bool	   *finished;
 	int			nplans;
 	int			i;
 	ListCell   *lc;
@@ -134,14 +136,17 @@ ExecInitAppend(Append *node, EState *estate, int eflags)
 	nplans = list_length(node->appendplans);
 
 	appendplanstates = (PlanState **) palloc0(nplans * sizeof(PlanState *));
-
+	finished = (bool *) palloc0(nplans * sizeof(bool));
+	
 	/*
 	 * create new AppendState for our append node
 	 */
 	appendstate->ps.plan = (Plan *) node;
 	appendstate->ps.state = estate;
 	appendstate->appendplans = appendplanstates;
+	appendstate->as_finished = finished;
 	appendstate->as_nplans = nplans;
+	appendstate->as_async = ((eflags & EXEC_FLAG_BACKWARD) == 0);
 
 	/*
 	 * Miscellaneous initialization
@@ -194,49 +199,104 @@ ExecInitAppend(Append *node, EState *estate, int eflags)
 void
 ExecAppend(AppendState *node)
 {
-	for (;;)
+	int n_notready = 0;
+	PlanState  *subnode;
+	int i, n;
+
+	n = node->as_whichplan;
+
+	for (i = 0 ; i < node->as_nplans ; i++)
 	{
-		PlanState  *subnode;
 		TupleTableSlot *result;
 
+		if (node->as_async)
+		{
+			if (n >= node->as_nplans)
+				n = 0;
+
+			if (node->as_finished[n])
+			{
+				n++;
+				continue;
+			}
+		}
+
 		/*
 		 * figure out which subplan we are currently processing
 		 */
-		subnode = node->appendplans[node->as_whichplan];
+		subnode = node->appendplans[n];
 
 		/*
-		 * get a tuple from the subplan
+		 * execute the subplan to get a result if it is not ready yet
 		 */
-		result = ExecProcNode(subnode);
+		if (!subnode->result_ready)
+			ExecExecuteNode(subnode);
+
+		/*
+		 * This subnode is not ready yet when asynchrony is not allowed,
+		 * immediately wait for this subnode.
+		 */
+		if (!subnode->result_ready)
+		{
+			if (node->as_async)
+			{
+				n_notready++;
+				n++;
+				continue;
+			}
+			ExecAsyncWaitForNode(subnode);
+		}
+
+		Assert(subnode->result_ready);
+
+		result = ExecConsumeResult((PlanState *)subnode);
 
 		if (!TupIsNull(result))
 		{
 			/*
-			 * If the subplan gave us something then return it as-is. We do
-			 * NOT make use of the result slot that was set up in
-			 * ExecInitAppend; there's no need for it.
+			 * If the subplan gave us something then return it
+			 * as-is. We do NOT make use of the result slot that was
+			 * set up in ExecInitAppend; there's no need for it.
 			 */
+			node->as_whichplan = n;
 			ExecReturnTuple(&node->ps, result);
 			return;
 		}
 
+		/* Tuple has been exhausted on this subnode */
+		if (node->as_async)
+		{
+			node->as_finished[n++] = true;
+			continue;
+		}
+
 		/*
 		 * Go on to the "next" subplan in the appropriate direction. If no
 		 * more subplans, return the empty slot set up for us by
 		 * ExecInitAppend.
 		 */
 		if (ScanDirectionIsForward(node->ps.state->es_direction))
-			node->as_whichplan++;
+		{
+			if (n >= node->as_nplans - 1)
+				break;
+			n++;
+		}
 		else
-			node->as_whichplan--;
-		if (!exec_append_initialize_next(node))
 		{
-			ExecReturnTuple(&node->ps,
-							ExecClearTuple(node->ps.ps_ResultTupleSlot));
-			return;
+			if (n == 0)
+				break;
+			n--;
 		}
+	}
 
-		/* Else loop back and try to get a tuple from the new subplan */
+	/*
+	 * We are finished if reached here and no subnodes are not-ready
+	 */
+	if (n_notready == 0)
+	{
+		node->as_whichplan = n;
+		ExecReturnTuple(&node->ps,
+						ExecClearTuple(node->ps.ps_ResultTupleSlot));
 	}
 }
 
@@ -277,6 +337,8 @@ ExecReScanAppend(AppendState *node)
 	{
 		PlanState  *subnode = node->appendplans[i];
 
+		node->as_finished[i] = false;
+
 		/*
 		 * ExecReScan doesn't know about my subplans, so I have to do
 		 * changed-parameter signaling myself.
diff --git a/src/include/nodes/execnodes.h b/src/include/nodes/execnodes.h
index 9121537..e4f2bb6 100644
--- a/src/include/nodes/execnodes.h
+++ b/src/include/nodes/execnodes.h
@@ -1170,6 +1170,8 @@ typedef struct AppendState
 {
 	PlanState	ps;				/* its first field is NodeTag */
 	PlanState **appendplans;	/* array of PlanStates for my inputs */
+	bool		as_async;		/* true to allow async execution */
+	bool	   *as_finished;	/* array of the running state of subplans */
 	int			as_nplans;
 	int			as_whichplan;
 } AppendState;
-- 
1.8.3.1


----Next_Part(Wed_Jul_06_16_29_14_2016_599)--
Content-Type: Text/Plain; charset=us-ascii
Content-Transfer-Encoding: 7bit
Content-Disposition: inline; filename="gentbl.sql"

CREATE EXTENSION postgres_fdw;
CREATE TABLE pl (a int, b int);
CREATE TABLE cl1 (LIKE pl) INHERITS (pl);
CREATE TABLE cl2 (LIKE pl) INHERITS (pl);
CREATE TABLE cl3 (LIKE pl) INHERITS (pl);
CREATE TABLE cl4 (LIKE pl) INHERITS (pl);
INSERT INTO cl1 (SELECT a, a FROM generate_series(0000000, 0999999) a);
INSERT INTO cl2 (SELECT a, a FROM generate_series(1000000, 1999999) a);
INSERT INTO cl3 (SELECT a, a FROM generate_series(2000000, 2999999) a);
INSERT INTO cl4 (SELECT a, a FROM generate_series(3000000, 3999999) a);

CREATE TABLE t0 (LIKE pl);
INSERT INTO t0 (SELECT a, a FROM generate_series(0000000, 9999999) a);

CREATE SERVER sv0 FOREIGN DATA WRAPPER postgres_fdw OPTIONS (host '/tmp', dbname 'postgres');
CREATE SERVER sv1 FOREIGN DATA WRAPPER postgres_fdw OPTIONS (host '/tmp', dbname 'postgres');
CREATE SERVER sv2 FOREIGN DATA WRAPPER postgres_fdw OPTIONS (host '/tmp', dbname 'postgres');
CREATE SERVER sv3 FOREIGN DATA WRAPPER postgres_fdw OPTIONS (host '/tmp', dbname 'postgres');
CREATE SERVER sv4 FOREIGN DATA WRAPPER postgres_fdw OPTIONS (host '/tmp', dbname 'postgres');
CREATE USER MAPPING FOR public SERVER sv0;
CREATE USER MAPPING FOR public SERVER sv1;
CREATE USER MAPPING FOR public SERVER sv2;
CREATE USER MAPPING FOR public SERVER sv3;
CREATE USER MAPPING FOR public SERVER sv4;

CREATE FOREIGN TABLE ft10 (a int, b int) SERVER sv0 OPTIONS (table_name 'cl1');
CREATE FOREIGN TABLE ft20 (a int, b int) SERVER sv0 OPTIONS (table_name 'cl2');
CREATE FOREIGN TABLE ft30 (a int, b int) SERVER sv0 OPTIONS (table_name 'cl3');
CREATE FOREIGN TABLE ft40 (a int, b int) SERVER sv0 OPTIONS (table_name 'cl4');
CREATE FOREIGN TABLE ft11 (a int, b int) SERVER sv1 OPTIONS (table_name 'cl1');
CREATE FOREIGN TABLE ft22 (a int, b int) SERVER sv2 OPTIONS (table_name 'cl2');
CREATE FOREIGN TABLE ft33 (a int, b int) SERVER sv3 OPTIONS (table_name 'cl3');
CREATE FOREIGN TABLE ft44 (a int, b int) SERVER sv4 OPTIONS (table_name 'cl4');

CREATE TABLE pf0 (LIKE pl);
ALTER FOREIGN TABLE ft10 INHERIT pf0;
ALTER FOREIGN TABLE ft20 INHERIT pf0;
ALTER FOREIGN TABLE ft30 INHERIT pf0;
ALTER FOREIGN TABLE ft40 INHERIT pf0;

CREATE table pf1 (LIKE pl);
ALTER FOREIGN TABLE ft11 INHERIT pf1;
ALTER FOREIGN TABLE ft22 INHERIT pf1;
ALTER FOREIGN TABLE ft33 INHERIT pf1;
ALTER FOREIGN TABLE ft44 INHERIT pf1;

ANALYZE;

----Next_Part(Wed_Jul_06_16_29_14_2016_599)--
Content-Type: Text/Plain; charset=us-ascii
Content-Transfer-Encoding: 7bit
Content-Disposition: inline; filename="testrun.sh"

#! /bin/bash

function do_test() {
	echo $1
	for i in $(seq 1 10); do psql postgres -c "set log_min_duration_statement = 0; set client_min_messages=log; select sum(a) from $1"; done | grep LOG
}

for i in "t0" "pl" "pf0" "pf1";
do
  do_test $i
done


----Next_Part(Wed_Jul_06_16_29_14_2016_599)--
Content-Type: text/plain
Content-Disposition: inline
Content-Transfer-Encoding: 8bit
MIME-Version: 1.0


-- 
Sent via pgsql-hackers mailing list ([email protected])
To make changes to your subscription:
http://www.postgresql.org/mailpref/pgsql-hackers

----Next_Part(Wed_Jul_06_16_29_14_2016_599)----






^ permalink  raw  reply  [nested|flat] 56+ messages in thread

* [PATCH 7/7] Make Append node async-aware.
@ 2016-06-28 09:52 Kyotaro Horiguchi <[email protected]>
  0 siblings, 0 replies; 56+ messages in thread

From: Kyotaro Horiguchi @ 2016-06-28 09:52 UTC (permalink / raw)

Change append node to be capable to handle asynchronous children
properly. As soon as it receives !async_ready from a child, it moves
to the next child and if no child is ready, it sleeps until at least
one of them become ready.
---
 src/backend/executor/nodeAppend.c | 67 +++++++++++++++++++++++++++++++++++++--
 src/include/nodes/execnodes.h     |  2 ++
 2 files changed, 67 insertions(+), 2 deletions(-)

diff --git a/src/backend/executor/nodeAppend.c b/src/backend/executor/nodeAppend.c
index e0ce8c6..9a4063a 100644
--- a/src/backend/executor/nodeAppend.c
+++ b/src/backend/executor/nodeAppend.c
@@ -58,6 +58,7 @@
 #include "postgres.h"
 
 #include "executor/execdebug.h"
+#include "executor/execAsync.h"
 #include "executor/nodeAppend.h"
 
 static bool exec_append_initialize_next(AppendState *appendstate);
@@ -121,6 +122,7 @@ ExecInitAppend(Append *node, EState *estate, int eflags)
 {
 	AppendState *appendstate = makeNode(AppendState);
 	PlanState **appendplanstates;
+	bool	   *finished;
 	int			nplans;
 	int			i;
 	ListCell   *lc;
@@ -134,14 +136,17 @@ ExecInitAppend(Append *node, EState *estate, int eflags)
 	nplans = list_length(node->appendplans);
 
 	appendplanstates = (PlanState **) palloc0(nplans * sizeof(PlanState *));
-
+	finished = (bool *) palloc0(nplans * sizeof(bool));
+	
 	/*
 	 * create new AppendState for our append node
 	 */
 	appendstate->ps.plan = (Plan *) node;
 	appendstate->ps.state = estate;
 	appendstate->appendplans = appendplanstates;
+	appendstate->as_finished = finished;
 	appendstate->as_nplans = nplans;
+	appendstate->as_async = ((eflags & EXEC_FLAG_BACKWARD) == 0);
 
 	/*
 	 * Miscellaneous initialization
@@ -194,6 +199,8 @@ ExecInitAppend(Append *node, EState *estate, int eflags)
 void
 ExecAppend(AppendState *node)
 {
+	int stopplan = node->as_whichplan;
+
 	for (;;)
 	{
 		PlanState  *subnode;
@@ -207,7 +214,36 @@ ExecAppend(AppendState *node)
 		/*
 		 * get a tuple from the subplan
 		 */
-		result = ExecProcNode(subnode);
+		Assert(!node->as_finished[node->as_whichplan]);
+
+		if (!subnode->result_ready)
+			ExecExecuteNode(subnode);
+
+		if (!subnode->result_ready)
+		{
+			if (node->as_async)
+			{
+				/* Move to the next living node */
+				do
+				{
+					node->as_whichplan = 
+						(node->as_whichplan + 1) %  node->as_nplans;
+				} while (node->as_whichplan != stopplan &&
+						 node->as_finished[node->as_whichplan]);
+
+				/* No node is ready yet, return as not-ready */
+				if (node->as_whichplan == stopplan)
+					return;
+
+				/* Try the next node */
+				continue;
+			}
+
+			/* If not async, immediately wait for this subnode */
+			ExecAsyncWaitForNode(subnode);
+		}				
+
+		result = ExecConsumeResult((PlanState *) subnode);
 
 		if (!TupIsNull(result))
 		{
@@ -220,6 +256,31 @@ ExecAppend(AppendState *node)
 			return;
 		}
 
+		if (node->as_async)
+		{
+			node->as_finished[node->as_whichplan] = true;
+			stopplan = node->as_whichplan;
+
+			/* Find the next living subnode */
+			do
+			{
+				node->as_whichplan =
+					(node->as_whichplan + 1) % node->as_nplans;
+			} while (node->as_whichplan != stopplan &&
+					 node->as_finished[node->as_whichplan]);
+
+			if (node->as_whichplan != stopplan)
+			{
+				stopplan = node->as_whichplan;
+				continue;
+			}
+
+			/* All subnodes are exhausted. Finish this node. */
+			ExecReturnTuple(&node->ps,
+							ExecClearTuple(node->ps.ps_ResultTupleSlot));
+			return;
+		}
+
 		/*
 		 * Go on to the "next" subplan in the appropriate direction. If no
 		 * more subplans, return the empty slot set up for us by
@@ -277,6 +338,8 @@ ExecReScanAppend(AppendState *node)
 	{
 		PlanState  *subnode = node->appendplans[i];
 
+		node->as_finished[i] = false;
+
 		/*
 		 * ExecReScan doesn't know about my subplans, so I have to do
 		 * changed-parameter signaling myself.
diff --git a/src/include/nodes/execnodes.h b/src/include/nodes/execnodes.h
index b72decc..b0a86c5 100644
--- a/src/include/nodes/execnodes.h
+++ b/src/include/nodes/execnodes.h
@@ -1177,6 +1177,8 @@ typedef struct AppendState
 {
 	PlanState	ps;				/* its first field is NodeTag */
 	PlanState **appendplans;	/* array of PlanStates for my inputs */
+	bool		as_async;		/* true to allow async execution */
+	bool	   *as_finished;	/* array of the running state of subplans */
 	int			as_nplans;
 	int			as_whichplan;
 } AppendState;
-- 
2.9.2


----Next_Part(Mon_Aug_29_17_08_36_2016_213)--
Content-Type: text/plain
Content-Disposition: inline
Content-Transfer-Encoding: 8bit
MIME-Version: 1.0


-- 
Sent via pgsql-hackers mailing list ([email protected])
To make changes to your subscription:
http://www.postgresql.org/mailpref/pgsql-hackers

----Next_Part(Mon_Aug_29_17_08_36_2016_213)----






^ permalink  raw  reply  [nested|flat] 56+ messages in thread

* [PATCH 7/7] Make Append node async-aware.
@ 2016-06-28 09:52 Kyotaro Horiguchi <[email protected]>
  0 siblings, 0 replies; 56+ messages in thread

From: Kyotaro Horiguchi @ 2016-06-28 09:52 UTC (permalink / raw)

Change append node to be capable to handle asynchronous children
properly. As soon as it receives !async_ready from a child, it moves
to the next child and if no child is ready, it sleeps until at least
one of them become ready.
---
 src/backend/executor/nodeAppend.c | 67 +++++++++++++++++++++++++++++++++++++--
 src/include/nodes/execnodes.h     |  2 ++
 2 files changed, 67 insertions(+), 2 deletions(-)

diff --git a/src/backend/executor/nodeAppend.c b/src/backend/executor/nodeAppend.c
index e0ce8c6..9a4063a 100644
--- a/src/backend/executor/nodeAppend.c
+++ b/src/backend/executor/nodeAppend.c
@@ -58,6 +58,7 @@
 #include "postgres.h"
 
 #include "executor/execdebug.h"
+#include "executor/execAsync.h"
 #include "executor/nodeAppend.h"
 
 static bool exec_append_initialize_next(AppendState *appendstate);
@@ -121,6 +122,7 @@ ExecInitAppend(Append *node, EState *estate, int eflags)
 {
 	AppendState *appendstate = makeNode(AppendState);
 	PlanState **appendplanstates;
+	bool	   *finished;
 	int			nplans;
 	int			i;
 	ListCell   *lc;
@@ -134,14 +136,17 @@ ExecInitAppend(Append *node, EState *estate, int eflags)
 	nplans = list_length(node->appendplans);
 
 	appendplanstates = (PlanState **) palloc0(nplans * sizeof(PlanState *));
-
+	finished = (bool *) palloc0(nplans * sizeof(bool));
+	
 	/*
 	 * create new AppendState for our append node
 	 */
 	appendstate->ps.plan = (Plan *) node;
 	appendstate->ps.state = estate;
 	appendstate->appendplans = appendplanstates;
+	appendstate->as_finished = finished;
 	appendstate->as_nplans = nplans;
+	appendstate->as_async = ((eflags & EXEC_FLAG_BACKWARD) == 0);
 
 	/*
 	 * Miscellaneous initialization
@@ -194,6 +199,8 @@ ExecInitAppend(Append *node, EState *estate, int eflags)
 void
 ExecAppend(AppendState *node)
 {
+	int stopplan = node->as_whichplan;
+
 	for (;;)
 	{
 		PlanState  *subnode;
@@ -207,7 +214,36 @@ ExecAppend(AppendState *node)
 		/*
 		 * get a tuple from the subplan
 		 */
-		result = ExecProcNode(subnode);
+		Assert(!node->as_finished[node->as_whichplan]);
+
+		if (!subnode->result_ready)
+			ExecExecuteNode(subnode);
+
+		if (!subnode->result_ready)
+		{
+			if (node->as_async)
+			{
+				/* Move to the next living node */
+				do
+				{
+					node->as_whichplan = 
+						(node->as_whichplan + 1) %  node->as_nplans;
+				} while (node->as_whichplan != stopplan &&
+						 node->as_finished[node->as_whichplan]);
+
+				/* No node is ready yet, return as not-ready */
+				if (node->as_whichplan == stopplan)
+					return;
+
+				/* Try the next node */
+				continue;
+			}
+
+			/* If not async, immediately wait for this subnode */
+			ExecAsyncWaitForNode(subnode);
+		}				
+
+		result = ExecConsumeResult((PlanState *) subnode);
 
 		if (!TupIsNull(result))
 		{
@@ -220,6 +256,31 @@ ExecAppend(AppendState *node)
 			return;
 		}
 
+		if (node->as_async)
+		{
+			node->as_finished[node->as_whichplan] = true;
+			stopplan = node->as_whichplan;
+
+			/* Find the next living subnode */
+			do
+			{
+				node->as_whichplan =
+					(node->as_whichplan + 1) % node->as_nplans;
+			} while (node->as_whichplan != stopplan &&
+					 node->as_finished[node->as_whichplan]);
+
+			if (node->as_whichplan != stopplan)
+			{
+				stopplan = node->as_whichplan;
+				continue;
+			}
+
+			/* All subnodes are exhausted. Finish this node. */
+			ExecReturnTuple(&node->ps,
+							ExecClearTuple(node->ps.ps_ResultTupleSlot));
+			return;
+		}
+
 		/*
 		 * Go on to the "next" subplan in the appropriate direction. If no
 		 * more subplans, return the empty slot set up for us by
@@ -277,6 +338,8 @@ ExecReScanAppend(AppendState *node)
 	{
 		PlanState  *subnode = node->appendplans[i];
 
+		node->as_finished[i] = false;
+
 		/*
 		 * ExecReScan doesn't know about my subplans, so I have to do
 		 * changed-parameter signaling myself.
diff --git a/src/include/nodes/execnodes.h b/src/include/nodes/execnodes.h
index b72decc..b0a86c5 100644
--- a/src/include/nodes/execnodes.h
+++ b/src/include/nodes/execnodes.h
@@ -1177,6 +1177,8 @@ typedef struct AppendState
 {
 	PlanState	ps;				/* its first field is NodeTag */
 	PlanState **appendplans;	/* array of PlanStates for my inputs */
+	bool		as_async;		/* true to allow async execution */
+	bool	   *as_finished;	/* array of the running state of subplans */
 	int			as_nplans;
 	int			as_whichplan;
 } AppendState;
-- 
2.9.2


----Next_Part(Mon_Aug_29_17_08_36_2016_213)--
Content-Type: text/plain
Content-Disposition: inline
Content-Transfer-Encoding: 8bit
MIME-Version: 1.0


-- 
Sent via pgsql-hackers mailing list ([email protected])
To make changes to your subscription:
http://www.postgresql.org/mailpref/pgsql-hackers

----Next_Part(Mon_Aug_29_17_08_36_2016_213)----






^ permalink  raw  reply  [nested|flat] 56+ messages in thread

* [PATCH 7/7] Make Append node async-aware.
@ 2016-06-28 09:52 Kyotaro Horiguchi <[email protected]>
  0 siblings, 0 replies; 56+ messages in thread

From: Kyotaro Horiguchi @ 2016-06-28 09:52 UTC (permalink / raw)

Change append node to be capable to handle asynchronous children
properly. As soon as it receives !async_ready from a child, it moves
to the next child and if no child is ready, it sleeps until at least
one of them become ready.
---
 src/backend/executor/nodeAppend.c | 67 +++++++++++++++++++++++++++++++++++++--
 src/include/nodes/execnodes.h     |  2 ++
 2 files changed, 67 insertions(+), 2 deletions(-)

diff --git a/src/backend/executor/nodeAppend.c b/src/backend/executor/nodeAppend.c
index e0ce8c6..9a4063a 100644
--- a/src/backend/executor/nodeAppend.c
+++ b/src/backend/executor/nodeAppend.c
@@ -58,6 +58,7 @@
 #include "postgres.h"
 
 #include "executor/execdebug.h"
+#include "executor/execAsync.h"
 #include "executor/nodeAppend.h"
 
 static bool exec_append_initialize_next(AppendState *appendstate);
@@ -121,6 +122,7 @@ ExecInitAppend(Append *node, EState *estate, int eflags)
 {
 	AppendState *appendstate = makeNode(AppendState);
 	PlanState **appendplanstates;
+	bool	   *finished;
 	int			nplans;
 	int			i;
 	ListCell   *lc;
@@ -134,14 +136,17 @@ ExecInitAppend(Append *node, EState *estate, int eflags)
 	nplans = list_length(node->appendplans);
 
 	appendplanstates = (PlanState **) palloc0(nplans * sizeof(PlanState *));
-
+	finished = (bool *) palloc0(nplans * sizeof(bool));
+	
 	/*
 	 * create new AppendState for our append node
 	 */
 	appendstate->ps.plan = (Plan *) node;
 	appendstate->ps.state = estate;
 	appendstate->appendplans = appendplanstates;
+	appendstate->as_finished = finished;
 	appendstate->as_nplans = nplans;
+	appendstate->as_async = ((eflags & EXEC_FLAG_BACKWARD) == 0);
 
 	/*
 	 * Miscellaneous initialization
@@ -194,6 +199,8 @@ ExecInitAppend(Append *node, EState *estate, int eflags)
 void
 ExecAppend(AppendState *node)
 {
+	int stopplan = node->as_whichplan;
+
 	for (;;)
 	{
 		PlanState  *subnode;
@@ -207,7 +214,36 @@ ExecAppend(AppendState *node)
 		/*
 		 * get a tuple from the subplan
 		 */
-		result = ExecProcNode(subnode);
+		Assert(!node->as_finished[node->as_whichplan]);
+
+		if (!subnode->result_ready)
+			ExecExecuteNode(subnode);
+
+		if (!subnode->result_ready)
+		{
+			if (node->as_async)
+			{
+				/* Move to the next living node */
+				do
+				{
+					node->as_whichplan = 
+						(node->as_whichplan + 1) %  node->as_nplans;
+				} while (node->as_whichplan != stopplan &&
+						 node->as_finished[node->as_whichplan]);
+
+				/* No node is ready yet, return as not-ready */
+				if (node->as_whichplan == stopplan)
+					return;
+
+				/* Try the next node */
+				continue;
+			}
+
+			/* If not async, immediately wait for this subnode */
+			ExecAsyncWaitForNode(subnode);
+		}				
+
+		result = ExecConsumeResult((PlanState *) subnode);
 
 		if (!TupIsNull(result))
 		{
@@ -220,6 +256,31 @@ ExecAppend(AppendState *node)
 			return;
 		}
 
+		if (node->as_async)
+		{
+			node->as_finished[node->as_whichplan] = true;
+			stopplan = node->as_whichplan;
+
+			/* Find the next living subnode */
+			do
+			{
+				node->as_whichplan =
+					(node->as_whichplan + 1) % node->as_nplans;
+			} while (node->as_whichplan != stopplan &&
+					 node->as_finished[node->as_whichplan]);
+
+			if (node->as_whichplan != stopplan)
+			{
+				stopplan = node->as_whichplan;
+				continue;
+			}
+
+			/* All subnodes are exhausted. Finish this node. */
+			ExecReturnTuple(&node->ps,
+							ExecClearTuple(node->ps.ps_ResultTupleSlot));
+			return;
+		}
+
 		/*
 		 * Go on to the "next" subplan in the appropriate direction. If no
 		 * more subplans, return the empty slot set up for us by
@@ -277,6 +338,8 @@ ExecReScanAppend(AppendState *node)
 	{
 		PlanState  *subnode = node->appendplans[i];
 
+		node->as_finished[i] = false;
+
 		/*
 		 * ExecReScan doesn't know about my subplans, so I have to do
 		 * changed-parameter signaling myself.
diff --git a/src/include/nodes/execnodes.h b/src/include/nodes/execnodes.h
index b72decc..b0a86c5 100644
--- a/src/include/nodes/execnodes.h
+++ b/src/include/nodes/execnodes.h
@@ -1177,6 +1177,8 @@ typedef struct AppendState
 {
 	PlanState	ps;				/* its first field is NodeTag */
 	PlanState **appendplans;	/* array of PlanStates for my inputs */
+	bool		as_async;		/* true to allow async execution */
+	bool	   *as_finished;	/* array of the running state of subplans */
 	int			as_nplans;
 	int			as_whichplan;
 } AppendState;
-- 
2.9.2


----Next_Part(Mon_Aug_29_17_08_36_2016_213)--
Content-Type: text/plain
Content-Disposition: inline
Content-Transfer-Encoding: 8bit
MIME-Version: 1.0


-- 
Sent via pgsql-hackers mailing list ([email protected])
To make changes to your subscription:
http://www.postgresql.org/mailpref/pgsql-hackers

----Next_Part(Mon_Aug_29_17_08_36_2016_213)----






^ permalink  raw  reply  [nested|flat] 56+ messages in thread

* [PATCH 7/7] Make Append node async-aware.
@ 2016-06-28 09:52 Kyotaro Horiguchi <[email protected]>
  0 siblings, 0 replies; 56+ messages in thread

From: Kyotaro Horiguchi @ 2016-06-28 09:52 UTC (permalink / raw)

Change append node to be capable to handle asynchronous children
properly. As soon as it receives !async_ready from a child, it moves
to the next child and if no child is ready, it sleeps until at least
one of them become ready.
---
 src/backend/executor/nodeAppend.c | 67 +++++++++++++++++++++++++++++++++++++--
 src/include/nodes/execnodes.h     |  2 ++
 2 files changed, 67 insertions(+), 2 deletions(-)

diff --git a/src/backend/executor/nodeAppend.c b/src/backend/executor/nodeAppend.c
index e0ce8c6..9a4063a 100644
--- a/src/backend/executor/nodeAppend.c
+++ b/src/backend/executor/nodeAppend.c
@@ -58,6 +58,7 @@
 #include "postgres.h"
 
 #include "executor/execdebug.h"
+#include "executor/execAsync.h"
 #include "executor/nodeAppend.h"
 
 static bool exec_append_initialize_next(AppendState *appendstate);
@@ -121,6 +122,7 @@ ExecInitAppend(Append *node, EState *estate, int eflags)
 {
 	AppendState *appendstate = makeNode(AppendState);
 	PlanState **appendplanstates;
+	bool	   *finished;
 	int			nplans;
 	int			i;
 	ListCell   *lc;
@@ -134,14 +136,17 @@ ExecInitAppend(Append *node, EState *estate, int eflags)
 	nplans = list_length(node->appendplans);
 
 	appendplanstates = (PlanState **) palloc0(nplans * sizeof(PlanState *));
-
+	finished = (bool *) palloc0(nplans * sizeof(bool));
+	
 	/*
 	 * create new AppendState for our append node
 	 */
 	appendstate->ps.plan = (Plan *) node;
 	appendstate->ps.state = estate;
 	appendstate->appendplans = appendplanstates;
+	appendstate->as_finished = finished;
 	appendstate->as_nplans = nplans;
+	appendstate->as_async = ((eflags & EXEC_FLAG_BACKWARD) == 0);
 
 	/*
 	 * Miscellaneous initialization
@@ -194,6 +199,8 @@ ExecInitAppend(Append *node, EState *estate, int eflags)
 void
 ExecAppend(AppendState *node)
 {
+	int stopplan = node->as_whichplan;
+
 	for (;;)
 	{
 		PlanState  *subnode;
@@ -207,7 +214,36 @@ ExecAppend(AppendState *node)
 		/*
 		 * get a tuple from the subplan
 		 */
-		result = ExecProcNode(subnode);
+		Assert(!node->as_finished[node->as_whichplan]);
+
+		if (!subnode->result_ready)
+			ExecExecuteNode(subnode);
+
+		if (!subnode->result_ready)
+		{
+			if (node->as_async)
+			{
+				/* Move to the next living node */
+				do
+				{
+					node->as_whichplan = 
+						(node->as_whichplan + 1) %  node->as_nplans;
+				} while (node->as_whichplan != stopplan &&
+						 node->as_finished[node->as_whichplan]);
+
+				/* No node is ready yet, return as not-ready */
+				if (node->as_whichplan == stopplan)
+					return;
+
+				/* Try the next node */
+				continue;
+			}
+
+			/* If not async, immediately wait for this subnode */
+			ExecAsyncWaitForNode(subnode);
+		}				
+
+		result = ExecConsumeResult((PlanState *) subnode);
 
 		if (!TupIsNull(result))
 		{
@@ -220,6 +256,31 @@ ExecAppend(AppendState *node)
 			return;
 		}
 
+		if (node->as_async)
+		{
+			node->as_finished[node->as_whichplan] = true;
+			stopplan = node->as_whichplan;
+
+			/* Find the next living subnode */
+			do
+			{
+				node->as_whichplan =
+					(node->as_whichplan + 1) % node->as_nplans;
+			} while (node->as_whichplan != stopplan &&
+					 node->as_finished[node->as_whichplan]);
+
+			if (node->as_whichplan != stopplan)
+			{
+				stopplan = node->as_whichplan;
+				continue;
+			}
+
+			/* All subnodes are exhausted. Finish this node. */
+			ExecReturnTuple(&node->ps,
+							ExecClearTuple(node->ps.ps_ResultTupleSlot));
+			return;
+		}
+
 		/*
 		 * Go on to the "next" subplan in the appropriate direction. If no
 		 * more subplans, return the empty slot set up for us by
@@ -277,6 +338,8 @@ ExecReScanAppend(AppendState *node)
 	{
 		PlanState  *subnode = node->appendplans[i];
 
+		node->as_finished[i] = false;
+
 		/*
 		 * ExecReScan doesn't know about my subplans, so I have to do
 		 * changed-parameter signaling myself.
diff --git a/src/include/nodes/execnodes.h b/src/include/nodes/execnodes.h
index b72decc..b0a86c5 100644
--- a/src/include/nodes/execnodes.h
+++ b/src/include/nodes/execnodes.h
@@ -1177,6 +1177,8 @@ typedef struct AppendState
 {
 	PlanState	ps;				/* its first field is NodeTag */
 	PlanState **appendplans;	/* array of PlanStates for my inputs */
+	bool		as_async;		/* true to allow async execution */
+	bool	   *as_finished;	/* array of the running state of subplans */
 	int			as_nplans;
 	int			as_whichplan;
 } AppendState;
-- 
1.8.3.1


----Next_Part(Thu_Jul_21_18_50_07_2016_597)--
Content-Type: text/plain
Content-Disposition: inline
Content-Transfer-Encoding: 8bit
MIME-Version: 1.0


-- 
Sent via pgsql-hackers mailing list ([email protected])
To make changes to your subscription:
http://www.postgresql.org/mailpref/pgsql-hackers

----Next_Part(Thu_Jul_21_18_50_07_2016_597)----






^ permalink  raw  reply  [nested|flat] 56+ messages in thread

* [PATCH 7/7] Make Append node async-aware.
@ 2016-06-28 09:52 Kyotaro Horiguchi <[email protected]>
  0 siblings, 0 replies; 56+ messages in thread

From: Kyotaro Horiguchi @ 2016-06-28 09:52 UTC (permalink / raw)

Change append node to be capable to handle asynchronous children
properly. As soon as it receives !async_ready from a child, it moves
to the next child and if no child is ready, it sleeps until at least
one of them become ready.
---
 src/backend/executor/nodeAppend.c | 67 +++++++++++++++++++++++++++++++++++++--
 src/include/nodes/execnodes.h     |  2 ++
 2 files changed, 67 insertions(+), 2 deletions(-)

diff --git a/src/backend/executor/nodeAppend.c b/src/backend/executor/nodeAppend.c
index e0ce8c6..9a4063a 100644
--- a/src/backend/executor/nodeAppend.c
+++ b/src/backend/executor/nodeAppend.c
@@ -58,6 +58,7 @@
 #include "postgres.h"
 
 #include "executor/execdebug.h"
+#include "executor/execAsync.h"
 #include "executor/nodeAppend.h"
 
 static bool exec_append_initialize_next(AppendState *appendstate);
@@ -121,6 +122,7 @@ ExecInitAppend(Append *node, EState *estate, int eflags)
 {
 	AppendState *appendstate = makeNode(AppendState);
 	PlanState **appendplanstates;
+	bool	   *finished;
 	int			nplans;
 	int			i;
 	ListCell   *lc;
@@ -134,14 +136,17 @@ ExecInitAppend(Append *node, EState *estate, int eflags)
 	nplans = list_length(node->appendplans);
 
 	appendplanstates = (PlanState **) palloc0(nplans * sizeof(PlanState *));
-
+	finished = (bool *) palloc0(nplans * sizeof(bool));
+	
 	/*
 	 * create new AppendState for our append node
 	 */
 	appendstate->ps.plan = (Plan *) node;
 	appendstate->ps.state = estate;
 	appendstate->appendplans = appendplanstates;
+	appendstate->as_finished = finished;
 	appendstate->as_nplans = nplans;
+	appendstate->as_async = ((eflags & EXEC_FLAG_BACKWARD) == 0);
 
 	/*
 	 * Miscellaneous initialization
@@ -194,6 +199,8 @@ ExecInitAppend(Append *node, EState *estate, int eflags)
 void
 ExecAppend(AppendState *node)
 {
+	int stopplan = node->as_whichplan;
+
 	for (;;)
 	{
 		PlanState  *subnode;
@@ -207,7 +214,36 @@ ExecAppend(AppendState *node)
 		/*
 		 * get a tuple from the subplan
 		 */
-		result = ExecProcNode(subnode);
+		Assert(!node->as_finished[node->as_whichplan]);
+
+		if (!subnode->result_ready)
+			ExecExecuteNode(subnode);
+
+		if (!subnode->result_ready)
+		{
+			if (node->as_async)
+			{
+				/* Move to the next living node */
+				do
+				{
+					node->as_whichplan = 
+						(node->as_whichplan + 1) %  node->as_nplans;
+				} while (node->as_whichplan != stopplan &&
+						 node->as_finished[node->as_whichplan]);
+
+				/* No node is ready yet, return as not-ready */
+				if (node->as_whichplan == stopplan)
+					return;
+
+				/* Try the next node */
+				continue;
+			}
+
+			/* If not async, immediately wait for this subnode */
+			ExecAsyncWaitForNode(subnode);
+		}				
+
+		result = ExecConsumeResult((PlanState *) subnode);
 
 		if (!TupIsNull(result))
 		{
@@ -220,6 +256,31 @@ ExecAppend(AppendState *node)
 			return;
 		}
 
+		if (node->as_async)
+		{
+			node->as_finished[node->as_whichplan] = true;
+			stopplan = node->as_whichplan;
+
+			/* Find the next living subnode */
+			do
+			{
+				node->as_whichplan =
+					(node->as_whichplan + 1) % node->as_nplans;
+			} while (node->as_whichplan != stopplan &&
+					 node->as_finished[node->as_whichplan]);
+
+			if (node->as_whichplan != stopplan)
+			{
+				stopplan = node->as_whichplan;
+				continue;
+			}
+
+			/* All subnodes are exhausted. Finish this node. */
+			ExecReturnTuple(&node->ps,
+							ExecClearTuple(node->ps.ps_ResultTupleSlot));
+			return;
+		}
+
 		/*
 		 * Go on to the "next" subplan in the appropriate direction. If no
 		 * more subplans, return the empty slot set up for us by
@@ -277,6 +338,8 @@ ExecReScanAppend(AppendState *node)
 	{
 		PlanState  *subnode = node->appendplans[i];
 
+		node->as_finished[i] = false;
+
 		/*
 		 * ExecReScan doesn't know about my subplans, so I have to do
 		 * changed-parameter signaling myself.
diff --git a/src/include/nodes/execnodes.h b/src/include/nodes/execnodes.h
index b72decc..b0a86c5 100644
--- a/src/include/nodes/execnodes.h
+++ b/src/include/nodes/execnodes.h
@@ -1177,6 +1177,8 @@ typedef struct AppendState
 {
 	PlanState	ps;				/* its first field is NodeTag */
 	PlanState **appendplans;	/* array of PlanStates for my inputs */
+	bool		as_async;		/* true to allow async execution */
+	bool	   *as_finished;	/* array of the running state of subplans */
 	int			as_nplans;
 	int			as_whichplan;
 } AppendState;
-- 
2.9.2


----Next_Part(Mon_Aug_29_17_08_36_2016_213)--
Content-Type: text/plain
Content-Disposition: inline
Content-Transfer-Encoding: 8bit
MIME-Version: 1.0


-- 
Sent via pgsql-hackers mailing list ([email protected])
To make changes to your subscription:
http://www.postgresql.org/mailpref/pgsql-hackers

----Next_Part(Mon_Aug_29_17_08_36_2016_213)----






^ permalink  raw  reply  [nested|flat] 56+ messages in thread

* [PATCH 7/7] Make Append node async-aware.
@ 2016-06-28 09:52 Kyotaro Horiguchi <[email protected]>
  0 siblings, 0 replies; 56+ messages in thread

From: Kyotaro Horiguchi @ 2016-06-28 09:52 UTC (permalink / raw)

Change append node to be capable to handle asynchronous children
properly. As soon as it receives !async_ready from a child, it moves
to the next child and if no child is ready, it sleeps until at least
one of them become ready.
---
 src/backend/executor/nodeAppend.c | 67 +++++++++++++++++++++++++++++++++++++--
 src/include/nodes/execnodes.h     |  2 ++
 2 files changed, 67 insertions(+), 2 deletions(-)

diff --git a/src/backend/executor/nodeAppend.c b/src/backend/executor/nodeAppend.c
index e0ce8c6..9a4063a 100644
--- a/src/backend/executor/nodeAppend.c
+++ b/src/backend/executor/nodeAppend.c
@@ -58,6 +58,7 @@
 #include "postgres.h"
 
 #include "executor/execdebug.h"
+#include "executor/execAsync.h"
 #include "executor/nodeAppend.h"
 
 static bool exec_append_initialize_next(AppendState *appendstate);
@@ -121,6 +122,7 @@ ExecInitAppend(Append *node, EState *estate, int eflags)
 {
 	AppendState *appendstate = makeNode(AppendState);
 	PlanState **appendplanstates;
+	bool	   *finished;
 	int			nplans;
 	int			i;
 	ListCell   *lc;
@@ -134,14 +136,17 @@ ExecInitAppend(Append *node, EState *estate, int eflags)
 	nplans = list_length(node->appendplans);
 
 	appendplanstates = (PlanState **) palloc0(nplans * sizeof(PlanState *));
-
+	finished = (bool *) palloc0(nplans * sizeof(bool));
+	
 	/*
 	 * create new AppendState for our append node
 	 */
 	appendstate->ps.plan = (Plan *) node;
 	appendstate->ps.state = estate;
 	appendstate->appendplans = appendplanstates;
+	appendstate->as_finished = finished;
 	appendstate->as_nplans = nplans;
+	appendstate->as_async = ((eflags & EXEC_FLAG_BACKWARD) == 0);
 
 	/*
 	 * Miscellaneous initialization
@@ -194,6 +199,8 @@ ExecInitAppend(Append *node, EState *estate, int eflags)
 void
 ExecAppend(AppendState *node)
 {
+	int stopplan = node->as_whichplan;
+
 	for (;;)
 	{
 		PlanState  *subnode;
@@ -207,7 +214,36 @@ ExecAppend(AppendState *node)
 		/*
 		 * get a tuple from the subplan
 		 */
-		result = ExecProcNode(subnode);
+		Assert(!node->as_finished[node->as_whichplan]);
+
+		if (!subnode->result_ready)
+			ExecExecuteNode(subnode);
+
+		if (!subnode->result_ready)
+		{
+			if (node->as_async)
+			{
+				/* Move to the next living node */
+				do
+				{
+					node->as_whichplan = 
+						(node->as_whichplan + 1) %  node->as_nplans;
+				} while (node->as_whichplan != stopplan &&
+						 node->as_finished[node->as_whichplan]);
+
+				/* No node is ready yet, return as not-ready */
+				if (node->as_whichplan == stopplan)
+					return;
+
+				/* Try the next node */
+				continue;
+			}
+
+			/* If not async, immediately wait for this subnode */
+			ExecAsyncWaitForNode(subnode);
+		}				
+
+		result = ExecConsumeResult((PlanState *) subnode);
 
 		if (!TupIsNull(result))
 		{
@@ -220,6 +256,31 @@ ExecAppend(AppendState *node)
 			return;
 		}
 
+		if (node->as_async)
+		{
+			node->as_finished[node->as_whichplan] = true;
+			stopplan = node->as_whichplan;
+
+			/* Find the next living subnode */
+			do
+			{
+				node->as_whichplan =
+					(node->as_whichplan + 1) % node->as_nplans;
+			} while (node->as_whichplan != stopplan &&
+					 node->as_finished[node->as_whichplan]);
+
+			if (node->as_whichplan != stopplan)
+			{
+				stopplan = node->as_whichplan;
+				continue;
+			}
+
+			/* All subnodes are exhausted. Finish this node. */
+			ExecReturnTuple(&node->ps,
+							ExecClearTuple(node->ps.ps_ResultTupleSlot));
+			return;
+		}
+
 		/*
 		 * Go on to the "next" subplan in the appropriate direction. If no
 		 * more subplans, return the empty slot set up for us by
@@ -277,6 +338,8 @@ ExecReScanAppend(AppendState *node)
 	{
 		PlanState  *subnode = node->appendplans[i];
 
+		node->as_finished[i] = false;
+
 		/*
 		 * ExecReScan doesn't know about my subplans, so I have to do
 		 * changed-parameter signaling myself.
diff --git a/src/include/nodes/execnodes.h b/src/include/nodes/execnodes.h
index b72decc..b0a86c5 100644
--- a/src/include/nodes/execnodes.h
+++ b/src/include/nodes/execnodes.h
@@ -1177,6 +1177,8 @@ typedef struct AppendState
 {
 	PlanState	ps;				/* its first field is NodeTag */
 	PlanState **appendplans;	/* array of PlanStates for my inputs */
+	bool		as_async;		/* true to allow async execution */
+	bool	   *as_finished;	/* array of the running state of subplans */
 	int			as_nplans;
 	int			as_whichplan;
 } AppendState;
-- 
1.8.3.1


----Next_Part(Thu_Jul_21_18_50_07_2016_597)--
Content-Type: text/plain
Content-Disposition: inline
Content-Transfer-Encoding: 8bit
MIME-Version: 1.0


-- 
Sent via pgsql-hackers mailing list ([email protected])
To make changes to your subscription:
http://www.postgresql.org/mailpref/pgsql-hackers

----Next_Part(Thu_Jul_21_18_50_07_2016_597)----






^ permalink  raw  reply  [nested|flat] 56+ messages in thread

* [PATCH 7/7] Make Append node async-aware.
@ 2016-06-28 09:52 Kyotaro Horiguchi <[email protected]>
  0 siblings, 0 replies; 56+ messages in thread

From: Kyotaro Horiguchi @ 2016-06-28 09:52 UTC (permalink / raw)

Change append node to be capable to handle asynchronous children
properly. As soon as it receives !async_ready from a child, it moves
to the next child and if no child is ready, it sleeps until at least
one of them become ready.
---
 src/backend/executor/nodeAppend.c | 67 +++++++++++++++++++++++++++++++++++++--
 src/include/nodes/execnodes.h     |  2 ++
 2 files changed, 67 insertions(+), 2 deletions(-)

diff --git a/src/backend/executor/nodeAppend.c b/src/backend/executor/nodeAppend.c
index e0ce8c6..9a4063a 100644
--- a/src/backend/executor/nodeAppend.c
+++ b/src/backend/executor/nodeAppend.c
@@ -58,6 +58,7 @@
 #include "postgres.h"
 
 #include "executor/execdebug.h"
+#include "executor/execAsync.h"
 #include "executor/nodeAppend.h"
 
 static bool exec_append_initialize_next(AppendState *appendstate);
@@ -121,6 +122,7 @@ ExecInitAppend(Append *node, EState *estate, int eflags)
 {
 	AppendState *appendstate = makeNode(AppendState);
 	PlanState **appendplanstates;
+	bool	   *finished;
 	int			nplans;
 	int			i;
 	ListCell   *lc;
@@ -134,14 +136,17 @@ ExecInitAppend(Append *node, EState *estate, int eflags)
 	nplans = list_length(node->appendplans);
 
 	appendplanstates = (PlanState **) palloc0(nplans * sizeof(PlanState *));
-
+	finished = (bool *) palloc0(nplans * sizeof(bool));
+	
 	/*
 	 * create new AppendState for our append node
 	 */
 	appendstate->ps.plan = (Plan *) node;
 	appendstate->ps.state = estate;
 	appendstate->appendplans = appendplanstates;
+	appendstate->as_finished = finished;
 	appendstate->as_nplans = nplans;
+	appendstate->as_async = ((eflags & EXEC_FLAG_BACKWARD) == 0);
 
 	/*
 	 * Miscellaneous initialization
@@ -194,6 +199,8 @@ ExecInitAppend(Append *node, EState *estate, int eflags)
 void
 ExecAppend(AppendState *node)
 {
+	int stopplan = node->as_whichplan;
+
 	for (;;)
 	{
 		PlanState  *subnode;
@@ -207,7 +214,36 @@ ExecAppend(AppendState *node)
 		/*
 		 * get a tuple from the subplan
 		 */
-		result = ExecProcNode(subnode);
+		Assert(!node->as_finished[node->as_whichplan]);
+
+		if (!subnode->result_ready)
+			ExecExecuteNode(subnode);
+
+		if (!subnode->result_ready)
+		{
+			if (node->as_async)
+			{
+				/* Move to the next living node */
+				do
+				{
+					node->as_whichplan = 
+						(node->as_whichplan + 1) %  node->as_nplans;
+				} while (node->as_whichplan != stopplan &&
+						 node->as_finished[node->as_whichplan]);
+
+				/* No node is ready yet, return as not-ready */
+				if (node->as_whichplan == stopplan)
+					return;
+
+				/* Try the next node */
+				continue;
+			}
+
+			/* If not async, immediately wait for this subnode */
+			ExecAsyncWaitForNode(subnode);
+		}				
+
+		result = ExecConsumeResult((PlanState *) subnode);
 
 		if (!TupIsNull(result))
 		{
@@ -220,6 +256,31 @@ ExecAppend(AppendState *node)
 			return;
 		}
 
+		if (node->as_async)
+		{
+			node->as_finished[node->as_whichplan] = true;
+			stopplan = node->as_whichplan;
+
+			/* Find the next living subnode */
+			do
+			{
+				node->as_whichplan =
+					(node->as_whichplan + 1) % node->as_nplans;
+			} while (node->as_whichplan != stopplan &&
+					 node->as_finished[node->as_whichplan]);
+
+			if (node->as_whichplan != stopplan)
+			{
+				stopplan = node->as_whichplan;
+				continue;
+			}
+
+			/* All subnodes are exhausted. Finish this node. */
+			ExecReturnTuple(&node->ps,
+							ExecClearTuple(node->ps.ps_ResultTupleSlot));
+			return;
+		}
+
 		/*
 		 * Go on to the "next" subplan in the appropriate direction. If no
 		 * more subplans, return the empty slot set up for us by
@@ -277,6 +338,8 @@ ExecReScanAppend(AppendState *node)
 	{
 		PlanState  *subnode = node->appendplans[i];
 
+		node->as_finished[i] = false;
+
 		/*
 		 * ExecReScan doesn't know about my subplans, so I have to do
 		 * changed-parameter signaling myself.
diff --git a/src/include/nodes/execnodes.h b/src/include/nodes/execnodes.h
index b72decc..b0a86c5 100644
--- a/src/include/nodes/execnodes.h
+++ b/src/include/nodes/execnodes.h
@@ -1177,6 +1177,8 @@ typedef struct AppendState
 {
 	PlanState	ps;				/* its first field is NodeTag */
 	PlanState **appendplans;	/* array of PlanStates for my inputs */
+	bool		as_async;		/* true to allow async execution */
+	bool	   *as_finished;	/* array of the running state of subplans */
 	int			as_nplans;
 	int			as_whichplan;
 } AppendState;
-- 
2.9.2


----Next_Part(Mon_Aug_29_17_08_36_2016_213)--
Content-Type: text/plain
Content-Disposition: inline
Content-Transfer-Encoding: 8bit
MIME-Version: 1.0


-- 
Sent via pgsql-hackers mailing list ([email protected])
To make changes to your subscription:
http://www.postgresql.org/mailpref/pgsql-hackers

----Next_Part(Mon_Aug_29_17_08_36_2016_213)----






^ permalink  raw  reply  [nested|flat] 56+ messages in thread

* [PATCH 7/7] Make Append node async-aware.
@ 2016-06-28 09:52 Kyotaro Horiguchi <[email protected]>
  0 siblings, 0 replies; 56+ messages in thread

From: Kyotaro Horiguchi @ 2016-06-28 09:52 UTC (permalink / raw)

Change append node to be capable to handle asynchronous children
properly. As soon as it receives !async_ready from a child, it moves
to the next child and if no child is ready, it sleeps until at least
one of them become ready.
---
 src/backend/executor/nodeAppend.c | 67 +++++++++++++++++++++++++++++++++++++--
 src/include/nodes/execnodes.h     |  2 ++
 2 files changed, 67 insertions(+), 2 deletions(-)

diff --git a/src/backend/executor/nodeAppend.c b/src/backend/executor/nodeAppend.c
index e0ce8c6..9a4063a 100644
--- a/src/backend/executor/nodeAppend.c
+++ b/src/backend/executor/nodeAppend.c
@@ -58,6 +58,7 @@
 #include "postgres.h"
 
 #include "executor/execdebug.h"
+#include "executor/execAsync.h"
 #include "executor/nodeAppend.h"
 
 static bool exec_append_initialize_next(AppendState *appendstate);
@@ -121,6 +122,7 @@ ExecInitAppend(Append *node, EState *estate, int eflags)
 {
 	AppendState *appendstate = makeNode(AppendState);
 	PlanState **appendplanstates;
+	bool	   *finished;
 	int			nplans;
 	int			i;
 	ListCell   *lc;
@@ -134,14 +136,17 @@ ExecInitAppend(Append *node, EState *estate, int eflags)
 	nplans = list_length(node->appendplans);
 
 	appendplanstates = (PlanState **) palloc0(nplans * sizeof(PlanState *));
-
+	finished = (bool *) palloc0(nplans * sizeof(bool));
+	
 	/*
 	 * create new AppendState for our append node
 	 */
 	appendstate->ps.plan = (Plan *) node;
 	appendstate->ps.state = estate;
 	appendstate->appendplans = appendplanstates;
+	appendstate->as_finished = finished;
 	appendstate->as_nplans = nplans;
+	appendstate->as_async = ((eflags & EXEC_FLAG_BACKWARD) == 0);
 
 	/*
 	 * Miscellaneous initialization
@@ -194,6 +199,8 @@ ExecInitAppend(Append *node, EState *estate, int eflags)
 void
 ExecAppend(AppendState *node)
 {
+	int stopplan = node->as_whichplan;
+
 	for (;;)
 	{
 		PlanState  *subnode;
@@ -207,7 +214,36 @@ ExecAppend(AppendState *node)
 		/*
 		 * get a tuple from the subplan
 		 */
-		result = ExecProcNode(subnode);
+		Assert(!node->as_finished[node->as_whichplan]);
+
+		if (!subnode->result_ready)
+			ExecExecuteNode(subnode);
+
+		if (!subnode->result_ready)
+		{
+			if (node->as_async)
+			{
+				/* Move to the next living node */
+				do
+				{
+					node->as_whichplan = 
+						(node->as_whichplan + 1) %  node->as_nplans;
+				} while (node->as_whichplan != stopplan &&
+						 node->as_finished[node->as_whichplan]);
+
+				/* No node is ready yet, return as not-ready */
+				if (node->as_whichplan == stopplan)
+					return;
+
+				/* Try the next node */
+				continue;
+			}
+
+			/* If not async, immediately wait for this subnode */
+			ExecAsyncWaitForNode(subnode);
+		}				
+
+		result = ExecConsumeResult((PlanState *) subnode);
 
 		if (!TupIsNull(result))
 		{
@@ -220,6 +256,31 @@ ExecAppend(AppendState *node)
 			return;
 		}
 
+		if (node->as_async)
+		{
+			node->as_finished[node->as_whichplan] = true;
+			stopplan = node->as_whichplan;
+
+			/* Find the next living subnode */
+			do
+			{
+				node->as_whichplan =
+					(node->as_whichplan + 1) % node->as_nplans;
+			} while (node->as_whichplan != stopplan &&
+					 node->as_finished[node->as_whichplan]);
+
+			if (node->as_whichplan != stopplan)
+			{
+				stopplan = node->as_whichplan;
+				continue;
+			}
+
+			/* All subnodes are exhausted. Finish this node. */
+			ExecReturnTuple(&node->ps,
+							ExecClearTuple(node->ps.ps_ResultTupleSlot));
+			return;
+		}
+
 		/*
 		 * Go on to the "next" subplan in the appropriate direction. If no
 		 * more subplans, return the empty slot set up for us by
@@ -277,6 +338,8 @@ ExecReScanAppend(AppendState *node)
 	{
 		PlanState  *subnode = node->appendplans[i];
 
+		node->as_finished[i] = false;
+
 		/*
 		 * ExecReScan doesn't know about my subplans, so I have to do
 		 * changed-parameter signaling myself.
diff --git a/src/include/nodes/execnodes.h b/src/include/nodes/execnodes.h
index b72decc..b0a86c5 100644
--- a/src/include/nodes/execnodes.h
+++ b/src/include/nodes/execnodes.h
@@ -1177,6 +1177,8 @@ typedef struct AppendState
 {
 	PlanState	ps;				/* its first field is NodeTag */
 	PlanState **appendplans;	/* array of PlanStates for my inputs */
+	bool		as_async;		/* true to allow async execution */
+	bool	   *as_finished;	/* array of the running state of subplans */
 	int			as_nplans;
 	int			as_whichplan;
 } AppendState;
-- 
2.9.2


----Next_Part(Mon_Aug_29_17_08_36_2016_213)--
Content-Type: text/plain
Content-Disposition: inline
Content-Transfer-Encoding: 8bit
MIME-Version: 1.0


-- 
Sent via pgsql-hackers mailing list ([email protected])
To make changes to your subscription:
http://www.postgresql.org/mailpref/pgsql-hackers

----Next_Part(Mon_Aug_29_17_08_36_2016_213)----






^ permalink  raw  reply  [nested|flat] 56+ messages in thread

* [PATCH 7/7] Make Append node async-aware.
@ 2016-06-28 09:52 Kyotaro Horiguchi <[email protected]>
  0 siblings, 0 replies; 56+ messages in thread

From: Kyotaro Horiguchi @ 2016-06-28 09:52 UTC (permalink / raw)

Change append node to be capable to handle asynchronous children
properly. As soon as it receives !async_ready from a child, it moves
to the next child and if no child is ready, it sleeps until at least
one of them become ready.
---
 src/backend/executor/nodeAppend.c | 94 ++++++++++++++++++++++++++++++++-------
 src/include/nodes/execnodes.h     |  2 +
 2 files changed, 80 insertions(+), 16 deletions(-)

diff --git a/src/backend/executor/nodeAppend.c b/src/backend/executor/nodeAppend.c
index e0ce8c6..1c0d26e 100644
--- a/src/backend/executor/nodeAppend.c
+++ b/src/backend/executor/nodeAppend.c
@@ -58,6 +58,7 @@
 #include "postgres.h"
 
 #include "executor/execdebug.h"
+#include "executor/execAsync.h"
 #include "executor/nodeAppend.h"
 
 static bool exec_append_initialize_next(AppendState *appendstate);
@@ -121,6 +122,7 @@ ExecInitAppend(Append *node, EState *estate, int eflags)
 {
 	AppendState *appendstate = makeNode(AppendState);
 	PlanState **appendplanstates;
+	bool	   *finished;
 	int			nplans;
 	int			i;
 	ListCell   *lc;
@@ -134,14 +136,17 @@ ExecInitAppend(Append *node, EState *estate, int eflags)
 	nplans = list_length(node->appendplans);
 
 	appendplanstates = (PlanState **) palloc0(nplans * sizeof(PlanState *));
-
+	finished = (bool *) palloc0(nplans * sizeof(bool));
+	
 	/*
 	 * create new AppendState for our append node
 	 */
 	appendstate->ps.plan = (Plan *) node;
 	appendstate->ps.state = estate;
 	appendstate->appendplans = appendplanstates;
+	appendstate->as_finished = finished;
 	appendstate->as_nplans = nplans;
+	appendstate->as_async = ((eflags & EXEC_FLAG_BACKWARD) == 0);
 
 	/*
 	 * Miscellaneous initialization
@@ -194,49 +199,104 @@ ExecInitAppend(Append *node, EState *estate, int eflags)
 void
 ExecAppend(AppendState *node)
 {
-	for (;;)
+	int n_notready = 0;
+	PlanState  *subnode;
+	int i, n;
+
+	n = node->as_whichplan;
+
+	for (i = 0 ; i < node->as_nplans ; i++)
 	{
-		PlanState  *subnode;
 		TupleTableSlot *result;
 
+		if (node->as_async)
+		{
+			if (n >= node->as_nplans)
+				n = 0;
+
+			if (node->as_finished[n])
+			{
+				n++;
+				continue;
+			}
+		}
+
 		/*
 		 * figure out which subplan we are currently processing
 		 */
-		subnode = node->appendplans[node->as_whichplan];
+		subnode = node->appendplans[n];
 
 		/*
-		 * get a tuple from the subplan
+		 * execute the subplan to get a result if it is not ready yet
 		 */
-		result = ExecProcNode(subnode);
+		if (!subnode->result_ready)
+			ExecExecuteNode(subnode);
+
+		/*
+		 * This subnode is not ready yet when asynchrony is not allowed,
+		 * immediately wait for this subnode.
+		 */
+		if (!subnode->result_ready)
+		{
+			if (node->as_async)
+			{
+				n_notready++;
+				n++;
+				continue;
+			}
+			ExecAsyncWaitForNode(subnode);
+		}
+
+		Assert(subnode->result_ready);
+
+		result = ExecConsumeResult((PlanState *)subnode);
 
 		if (!TupIsNull(result))
 		{
 			/*
-			 * If the subplan gave us something then return it as-is. We do
-			 * NOT make use of the result slot that was set up in
-			 * ExecInitAppend; there's no need for it.
+			 * If the subplan gave us something then return it
+			 * as-is. We do NOT make use of the result slot that was
+			 * set up in ExecInitAppend; there's no need for it.
 			 */
+			node->as_whichplan = n;
 			ExecReturnTuple(&node->ps, result);
 			return;
 		}
 
+		/* Tuple has been exhausted on this subnode */
+		if (node->as_async)
+		{
+			node->as_finished[n++] = true;
+			continue;
+		}
+
 		/*
 		 * Go on to the "next" subplan in the appropriate direction. If no
 		 * more subplans, return the empty slot set up for us by
 		 * ExecInitAppend.
 		 */
 		if (ScanDirectionIsForward(node->ps.state->es_direction))
-			node->as_whichplan++;
+		{
+			if (n >= node->as_nplans - 1)
+				break;
+			n++;
+		}
 		else
-			node->as_whichplan--;
-		if (!exec_append_initialize_next(node))
 		{
-			ExecReturnTuple(&node->ps,
-							ExecClearTuple(node->ps.ps_ResultTupleSlot));
-			return;
+			if (n == 0)
+				break;
+			n--;
 		}
+	}
 
-		/* Else loop back and try to get a tuple from the new subplan */
+	/*
+	 * We are finished if reached here and no subnodes are not-ready
+	 */
+	if (n_notready == 0)
+	{
+		node->as_whichplan = n;
+		ExecReturnTuple(&node->ps,
+						ExecClearTuple(node->ps.ps_ResultTupleSlot));
 	}
 }
 
@@ -277,6 +337,8 @@ ExecReScanAppend(AppendState *node)
 	{
 		PlanState  *subnode = node->appendplans[i];
 
+		node->as_finished[i] = false;
+
 		/*
 		 * ExecReScan doesn't know about my subplans, so I have to do
 		 * changed-parameter signaling myself.
diff --git a/src/include/nodes/execnodes.h b/src/include/nodes/execnodes.h
index 9121537..e4f2bb6 100644
--- a/src/include/nodes/execnodes.h
+++ b/src/include/nodes/execnodes.h
@@ -1170,6 +1170,8 @@ typedef struct AppendState
 {
 	PlanState	ps;				/* its first field is NodeTag */
 	PlanState **appendplans;	/* array of PlanStates for my inputs */
+	bool		as_async;		/* true to allow async execution */
+	bool	   *as_finished;	/* array of the running state of subplans */
 	int			as_nplans;
 	int			as_whichplan;
 } AppendState;
-- 
1.8.3.1


----Next_Part(Wed_Jul_06_16_29_14_2016_599)--
Content-Type: Text/Plain; charset=us-ascii
Content-Transfer-Encoding: 7bit
Content-Disposition: inline; filename="gentbl.sql"

CREATE EXTENSION postgres_fdw;
CREATE TABLE pl (a int, b int);
CREATE TABLE cl1 (LIKE pl) INHERITS (pl);
CREATE TABLE cl2 (LIKE pl) INHERITS (pl);
CREATE TABLE cl3 (LIKE pl) INHERITS (pl);
CREATE TABLE cl4 (LIKE pl) INHERITS (pl);
INSERT INTO cl1 (SELECT a, a FROM generate_series(0000000, 0999999) a);
INSERT INTO cl2 (SELECT a, a FROM generate_series(1000000, 1999999) a);
INSERT INTO cl3 (SELECT a, a FROM generate_series(2000000, 2999999) a);
INSERT INTO cl4 (SELECT a, a FROM generate_series(3000000, 3999999) a);

CREATE TABLE t0 (LIKE pl);
INSERT INTO t0 (SELECT a, a FROM generate_series(0000000, 9999999) a);

CREATE SERVER sv0 FOREIGN DATA WRAPPER postgres_fdw OPTIONS (host '/tmp', dbname 'postgres');
CREATE SERVER sv1 FOREIGN DATA WRAPPER postgres_fdw OPTIONS (host '/tmp', dbname 'postgres');
CREATE SERVER sv2 FOREIGN DATA WRAPPER postgres_fdw OPTIONS (host '/tmp', dbname 'postgres');
CREATE SERVER sv3 FOREIGN DATA WRAPPER postgres_fdw OPTIONS (host '/tmp', dbname 'postgres');
CREATE SERVER sv4 FOREIGN DATA WRAPPER postgres_fdw OPTIONS (host '/tmp', dbname 'postgres');
CREATE USER MAPPING FOR public SERVER sv0;
CREATE USER MAPPING FOR public SERVER sv1;
CREATE USER MAPPING FOR public SERVER sv2;
CREATE USER MAPPING FOR public SERVER sv3;
CREATE USER MAPPING FOR public SERVER sv4;

CREATE FOREIGN TABLE ft10 (a int, b int) SERVER sv0 OPTIONS (table_name 'cl1');
CREATE FOREIGN TABLE ft20 (a int, b int) SERVER sv0 OPTIONS (table_name 'cl2');
CREATE FOREIGN TABLE ft30 (a int, b int) SERVER sv0 OPTIONS (table_name 'cl3');
CREATE FOREIGN TABLE ft40 (a int, b int) SERVER sv0 OPTIONS (table_name 'cl4');
CREATE FOREIGN TABLE ft11 (a int, b int) SERVER sv1 OPTIONS (table_name 'cl1');
CREATE FOREIGN TABLE ft22 (a int, b int) SERVER sv2 OPTIONS (table_name 'cl2');
CREATE FOREIGN TABLE ft33 (a int, b int) SERVER sv3 OPTIONS (table_name 'cl3');
CREATE FOREIGN TABLE ft44 (a int, b int) SERVER sv4 OPTIONS (table_name 'cl4');

CREATE TABLE pf0 (LIKE pl);
ALTER FOREIGN TABLE ft10 INHERIT pf0;
ALTER FOREIGN TABLE ft20 INHERIT pf0;
ALTER FOREIGN TABLE ft30 INHERIT pf0;
ALTER FOREIGN TABLE ft40 INHERIT pf0;

CREATE table pf1 (LIKE pl);
ALTER FOREIGN TABLE ft11 INHERIT pf1;
ALTER FOREIGN TABLE ft22 INHERIT pf1;
ALTER FOREIGN TABLE ft33 INHERIT pf1;
ALTER FOREIGN TABLE ft44 INHERIT pf1;

ANALYZE;

----Next_Part(Wed_Jul_06_16_29_14_2016_599)--
Content-Type: Text/Plain; charset=us-ascii
Content-Transfer-Encoding: 7bit
Content-Disposition: inline; filename="testrun.sh"

#! /bin/bash

function do_test() {
	echo $1
	for i in $(seq 1 10); do psql postgres -c "set log_min_duration_statement = 0; set client_min_messages=log; select sum(a) from $1"; done | grep LOG
}

for i in "t0" "pl" "pf0" "pf1";
do
  do_test $i
done


----Next_Part(Wed_Jul_06_16_29_14_2016_599)--
Content-Type: text/plain
Content-Disposition: inline
Content-Transfer-Encoding: 8bit
MIME-Version: 1.0


-- 
Sent via pgsql-hackers mailing list ([email protected])
To make changes to your subscription:
http://www.postgresql.org/mailpref/pgsql-hackers

----Next_Part(Wed_Jul_06_16_29_14_2016_599)----






^ permalink  raw  reply  [nested|flat] 56+ messages in thread

* [PATCH 7/7] Make Append node async-aware.
@ 2016-06-28 09:52 Kyotaro Horiguchi <[email protected]>
  0 siblings, 0 replies; 56+ messages in thread

From: Kyotaro Horiguchi @ 2016-06-28 09:52 UTC (permalink / raw)

Change append node to be capable to handle asynchronous children
properly. As soon as it receives !async_ready from a child, it moves
to the next child and if no child is ready, it sleeps until at least
one of them become ready.
---
 src/backend/executor/nodeAppend.c | 67 +++++++++++++++++++++++++++++++++++++--
 src/include/nodes/execnodes.h     |  2 ++
 2 files changed, 67 insertions(+), 2 deletions(-)

diff --git a/src/backend/executor/nodeAppend.c b/src/backend/executor/nodeAppend.c
index e0ce8c6..9a4063a 100644
--- a/src/backend/executor/nodeAppend.c
+++ b/src/backend/executor/nodeAppend.c
@@ -58,6 +58,7 @@
 #include "postgres.h"
 
 #include "executor/execdebug.h"
+#include "executor/execAsync.h"
 #include "executor/nodeAppend.h"
 
 static bool exec_append_initialize_next(AppendState *appendstate);
@@ -121,6 +122,7 @@ ExecInitAppend(Append *node, EState *estate, int eflags)
 {
 	AppendState *appendstate = makeNode(AppendState);
 	PlanState **appendplanstates;
+	bool	   *finished;
 	int			nplans;
 	int			i;
 	ListCell   *lc;
@@ -134,14 +136,17 @@ ExecInitAppend(Append *node, EState *estate, int eflags)
 	nplans = list_length(node->appendplans);
 
 	appendplanstates = (PlanState **) palloc0(nplans * sizeof(PlanState *));
-
+	finished = (bool *) palloc0(nplans * sizeof(bool));
+	
 	/*
 	 * create new AppendState for our append node
 	 */
 	appendstate->ps.plan = (Plan *) node;
 	appendstate->ps.state = estate;
 	appendstate->appendplans = appendplanstates;
+	appendstate->as_finished = finished;
 	appendstate->as_nplans = nplans;
+	appendstate->as_async = ((eflags & EXEC_FLAG_BACKWARD) == 0);
 
 	/*
 	 * Miscellaneous initialization
@@ -194,6 +199,8 @@ ExecInitAppend(Append *node, EState *estate, int eflags)
 void
 ExecAppend(AppendState *node)
 {
+	int stopplan = node->as_whichplan;
+
 	for (;;)
 	{
 		PlanState  *subnode;
@@ -207,7 +214,36 @@ ExecAppend(AppendState *node)
 		/*
 		 * get a tuple from the subplan
 		 */
-		result = ExecProcNode(subnode);
+		Assert(!node->as_finished[node->as_whichplan]);
+
+		if (!subnode->result_ready)
+			ExecExecuteNode(subnode);
+
+		if (!subnode->result_ready)
+		{
+			if (node->as_async)
+			{
+				/* Move to the next living node */
+				do
+				{
+					node->as_whichplan = 
+						(node->as_whichplan + 1) %  node->as_nplans;
+				} while (node->as_whichplan != stopplan &&
+						 node->as_finished[node->as_whichplan]);
+
+				/* No node is ready yet, return as not-ready */
+				if (node->as_whichplan == stopplan)
+					return;
+
+				/* Try the next node */
+				continue;
+			}
+
+			/* If not async, immediately wait for this subnode */
+			ExecAsyncWaitForNode(subnode);
+		}				
+
+		result = ExecConsumeResult((PlanState *) subnode);
 
 		if (!TupIsNull(result))
 		{
@@ -220,6 +256,31 @@ ExecAppend(AppendState *node)
 			return;
 		}
 
+		if (node->as_async)
+		{
+			node->as_finished[node->as_whichplan] = true;
+			stopplan = node->as_whichplan;
+
+			/* Find the next living subnode */
+			do
+			{
+				node->as_whichplan =
+					(node->as_whichplan + 1) % node->as_nplans;
+			} while (node->as_whichplan != stopplan &&
+					 node->as_finished[node->as_whichplan]);
+
+			if (node->as_whichplan != stopplan)
+			{
+				stopplan = node->as_whichplan;
+				continue;
+			}
+
+			/* All subnodes are exhausted. Finish this node. */
+			ExecReturnTuple(&node->ps,
+							ExecClearTuple(node->ps.ps_ResultTupleSlot));
+			return;
+		}
+
 		/*
 		 * Go on to the "next" subplan in the appropriate direction. If no
 		 * more subplans, return the empty slot set up for us by
@@ -277,6 +338,8 @@ ExecReScanAppend(AppendState *node)
 	{
 		PlanState  *subnode = node->appendplans[i];
 
+		node->as_finished[i] = false;
+
 		/*
 		 * ExecReScan doesn't know about my subplans, so I have to do
 		 * changed-parameter signaling myself.
diff --git a/src/include/nodes/execnodes.h b/src/include/nodes/execnodes.h
index b72decc..b0a86c5 100644
--- a/src/include/nodes/execnodes.h
+++ b/src/include/nodes/execnodes.h
@@ -1177,6 +1177,8 @@ typedef struct AppendState
 {
 	PlanState	ps;				/* its first field is NodeTag */
 	PlanState **appendplans;	/* array of PlanStates for my inputs */
+	bool		as_async;		/* true to allow async execution */
+	bool	   *as_finished;	/* array of the running state of subplans */
 	int			as_nplans;
 	int			as_whichplan;
 } AppendState;
-- 
1.8.3.1


----Next_Part(Thu_Jul_21_18_50_07_2016_597)--
Content-Type: text/plain
Content-Disposition: inline
Content-Transfer-Encoding: 8bit
MIME-Version: 1.0


-- 
Sent via pgsql-hackers mailing list ([email protected])
To make changes to your subscription:
http://www.postgresql.org/mailpref/pgsql-hackers

----Next_Part(Thu_Jul_21_18_50_07_2016_597)----






^ permalink  raw  reply  [nested|flat] 56+ messages in thread

* [PATCH 7/7] Make Append node async-aware.
@ 2016-06-28 09:52 Kyotaro Horiguchi <[email protected]>
  0 siblings, 0 replies; 56+ messages in thread

From: Kyotaro Horiguchi @ 2016-06-28 09:52 UTC (permalink / raw)

Change append node to be capable to handle asynchronous children
properly. As soon as it receives !async_ready from a child, it moves
to the next child and if no child is ready, it sleeps until at least
one of them become ready.
---
 src/backend/executor/nodeAppend.c | 94 ++++++++++++++++++++++++++++++++-------
 src/include/nodes/execnodes.h     |  2 +
 2 files changed, 80 insertions(+), 16 deletions(-)

diff --git a/src/backend/executor/nodeAppend.c b/src/backend/executor/nodeAppend.c
index e0ce8c6..1c0d26e 100644
--- a/src/backend/executor/nodeAppend.c
+++ b/src/backend/executor/nodeAppend.c
@@ -58,6 +58,7 @@
 #include "postgres.h"
 
 #include "executor/execdebug.h"
+#include "executor/execAsync.h"
 #include "executor/nodeAppend.h"
 
 static bool exec_append_initialize_next(AppendState *appendstate);
@@ -121,6 +122,7 @@ ExecInitAppend(Append *node, EState *estate, int eflags)
 {
 	AppendState *appendstate = makeNode(AppendState);
 	PlanState **appendplanstates;
+	bool	   *finished;
 	int			nplans;
 	int			i;
 	ListCell   *lc;
@@ -134,14 +136,17 @@ ExecInitAppend(Append *node, EState *estate, int eflags)
 	nplans = list_length(node->appendplans);
 
 	appendplanstates = (PlanState **) palloc0(nplans * sizeof(PlanState *));
-
+	finished = (bool *) palloc0(nplans * sizeof(bool));
+	
 	/*
 	 * create new AppendState for our append node
 	 */
 	appendstate->ps.plan = (Plan *) node;
 	appendstate->ps.state = estate;
 	appendstate->appendplans = appendplanstates;
+	appendstate->as_finished = finished;
 	appendstate->as_nplans = nplans;
+	appendstate->as_async = ((eflags & EXEC_FLAG_BACKWARD) == 0);
 
 	/*
 	 * Miscellaneous initialization
@@ -194,49 +199,104 @@ ExecInitAppend(Append *node, EState *estate, int eflags)
 void
 ExecAppend(AppendState *node)
 {
-	for (;;)
+	int n_notready = 0;
+	PlanState  *subnode;
+	int i, n;
+
+	n = node->as_whichplan;
+
+	for (i = 0 ; i < node->as_nplans ; i++)
 	{
-		PlanState  *subnode;
 		TupleTableSlot *result;
 
+		if (node->as_async)
+		{
+			if (n >= node->as_nplans)
+				n = 0;
+
+			if (node->as_finished[n])
+			{
+				n++;
+				continue;
+			}
+		}
+
 		/*
 		 * figure out which subplan we are currently processing
 		 */
-		subnode = node->appendplans[node->as_whichplan];
+		subnode = node->appendplans[n];
 
 		/*
-		 * get a tuple from the subplan
+		 * execute the subplan to get a result if it is not ready yet
 		 */
-		result = ExecProcNode(subnode);
+		if (!subnode->result_ready)
+			ExecExecuteNode(subnode);
+
+		/*
+		 * This subnode is not ready yet when asynchrony is not allowed,
+		 * immediately wait for this subnode.
+		 */
+		if (!subnode->result_ready)
+		{
+			if (node->as_async)
+			{
+				n_notready++;
+				n++;
+				continue;
+			}
+			ExecAsyncWaitForNode(subnode);
+		}
+
+		Assert(subnode->result_ready);
+
+		result = ExecConsumeResult((PlanState *)subnode);
 
 		if (!TupIsNull(result))
 		{
 			/*
-			 * If the subplan gave us something then return it as-is. We do
-			 * NOT make use of the result slot that was set up in
-			 * ExecInitAppend; there's no need for it.
+			 * If the subplan gave us something then return it
+			 * as-is. We do NOT make use of the result slot that was
+			 * set up in ExecInitAppend; there's no need for it.
 			 */
+			node->as_whichplan = n;
 			ExecReturnTuple(&node->ps, result);
 			return;
 		}
 
+		/* Tuple has been exhausted on this subnode */
+		if (node->as_async)
+		{
+			node->as_finished[n++] = true;
+			continue;
+		}
+
 		/*
 		 * Go on to the "next" subplan in the appropriate direction. If no
 		 * more subplans, return the empty slot set up for us by
 		 * ExecInitAppend.
 		 */
 		if (ScanDirectionIsForward(node->ps.state->es_direction))
-			node->as_whichplan++;
+		{
+			if (n >= node->as_nplans - 1)
+				break;
+			n++;
+		}
 		else
-			node->as_whichplan--;
-		if (!exec_append_initialize_next(node))
 		{
-			ExecReturnTuple(&node->ps,
-							ExecClearTuple(node->ps.ps_ResultTupleSlot));
-			return;
+			if (n == 0)
+				break;
+			n--;
 		}
+	}
 
-		/* Else loop back and try to get a tuple from the new subplan */
+	/*
+	 * We are finished if reached here and no subnodes are not-ready
+	 */
+	if (n_notready == 0)
+	{
+		node->as_whichplan = n;
+		ExecReturnTuple(&node->ps,
+						ExecClearTuple(node->ps.ps_ResultTupleSlot));
 	}
 }
 
@@ -277,6 +337,8 @@ ExecReScanAppend(AppendState *node)
 	{
 		PlanState  *subnode = node->appendplans[i];
 
+		node->as_finished[i] = false;
+
 		/*
 		 * ExecReScan doesn't know about my subplans, so I have to do
 		 * changed-parameter signaling myself.
diff --git a/src/include/nodes/execnodes.h b/src/include/nodes/execnodes.h
index 9121537..e4f2bb6 100644
--- a/src/include/nodes/execnodes.h
+++ b/src/include/nodes/execnodes.h
@@ -1170,6 +1170,8 @@ typedef struct AppendState
 {
 	PlanState	ps;				/* its first field is NodeTag */
 	PlanState **appendplans;	/* array of PlanStates for my inputs */
+	bool		as_async;		/* true to allow async execution */
+	bool	   *as_finished;	/* array of the running state of subplans */
 	int			as_nplans;
 	int			as_whichplan;
 } AppendState;
-- 
1.8.3.1


----Next_Part(Wed_Jul_06_16_29_14_2016_599)--
Content-Type: Text/Plain; charset=us-ascii
Content-Transfer-Encoding: 7bit
Content-Disposition: inline; filename="gentbl.sql"

CREATE EXTENSION postgres_fdw;
CREATE TABLE pl (a int, b int);
CREATE TABLE cl1 (LIKE pl) INHERITS (pl);
CREATE TABLE cl2 (LIKE pl) INHERITS (pl);
CREATE TABLE cl3 (LIKE pl) INHERITS (pl);
CREATE TABLE cl4 (LIKE pl) INHERITS (pl);
INSERT INTO cl1 (SELECT a, a FROM generate_series(0000000, 0999999) a);
INSERT INTO cl2 (SELECT a, a FROM generate_series(1000000, 1999999) a);
INSERT INTO cl3 (SELECT a, a FROM generate_series(2000000, 2999999) a);
INSERT INTO cl4 (SELECT a, a FROM generate_series(3000000, 3999999) a);

CREATE TABLE t0 (LIKE pl);
INSERT INTO t0 (SELECT a, a FROM generate_series(0000000, 9999999) a);

CREATE SERVER sv0 FOREIGN DATA WRAPPER postgres_fdw OPTIONS (host '/tmp', dbname 'postgres');
CREATE SERVER sv1 FOREIGN DATA WRAPPER postgres_fdw OPTIONS (host '/tmp', dbname 'postgres');
CREATE SERVER sv2 FOREIGN DATA WRAPPER postgres_fdw OPTIONS (host '/tmp', dbname 'postgres');
CREATE SERVER sv3 FOREIGN DATA WRAPPER postgres_fdw OPTIONS (host '/tmp', dbname 'postgres');
CREATE SERVER sv4 FOREIGN DATA WRAPPER postgres_fdw OPTIONS (host '/tmp', dbname 'postgres');
CREATE USER MAPPING FOR public SERVER sv0;
CREATE USER MAPPING FOR public SERVER sv1;
CREATE USER MAPPING FOR public SERVER sv2;
CREATE USER MAPPING FOR public SERVER sv3;
CREATE USER MAPPING FOR public SERVER sv4;

CREATE FOREIGN TABLE ft10 (a int, b int) SERVER sv0 OPTIONS (table_name 'cl1');
CREATE FOREIGN TABLE ft20 (a int, b int) SERVER sv0 OPTIONS (table_name 'cl2');
CREATE FOREIGN TABLE ft30 (a int, b int) SERVER sv0 OPTIONS (table_name 'cl3');
CREATE FOREIGN TABLE ft40 (a int, b int) SERVER sv0 OPTIONS (table_name 'cl4');
CREATE FOREIGN TABLE ft11 (a int, b int) SERVER sv1 OPTIONS (table_name 'cl1');
CREATE FOREIGN TABLE ft22 (a int, b int) SERVER sv2 OPTIONS (table_name 'cl2');
CREATE FOREIGN TABLE ft33 (a int, b int) SERVER sv3 OPTIONS (table_name 'cl3');
CREATE FOREIGN TABLE ft44 (a int, b int) SERVER sv4 OPTIONS (table_name 'cl4');

CREATE TABLE pf0 (LIKE pl);
ALTER FOREIGN TABLE ft10 INHERIT pf0;
ALTER FOREIGN TABLE ft20 INHERIT pf0;
ALTER FOREIGN TABLE ft30 INHERIT pf0;
ALTER FOREIGN TABLE ft40 INHERIT pf0;

CREATE table pf1 (LIKE pl);
ALTER FOREIGN TABLE ft11 INHERIT pf1;
ALTER FOREIGN TABLE ft22 INHERIT pf1;
ALTER FOREIGN TABLE ft33 INHERIT pf1;
ALTER FOREIGN TABLE ft44 INHERIT pf1;

ANALYZE;

----Next_Part(Wed_Jul_06_16_29_14_2016_599)--
Content-Type: Text/Plain; charset=us-ascii
Content-Transfer-Encoding: 7bit
Content-Disposition: inline; filename="testrun.sh"

#! /bin/bash

function do_test() {
	echo $1
	for i in $(seq 1 10); do psql postgres -c "set log_min_duration_statement = 0; set client_min_messages=log; select sum(a) from $1"; done | grep LOG
}

for i in "t0" "pl" "pf0" "pf1";
do
  do_test $i
done


----Next_Part(Wed_Jul_06_16_29_14_2016_599)--
Content-Type: text/plain
Content-Disposition: inline
Content-Transfer-Encoding: 8bit
MIME-Version: 1.0


-- 
Sent via pgsql-hackers mailing list ([email protected])
To make changes to your subscription:
http://www.postgresql.org/mailpref/pgsql-hackers

----Next_Part(Wed_Jul_06_16_29_14_2016_599)----






^ permalink  raw  reply  [nested|flat] 56+ messages in thread

* [PATCH 7/7] Make Append node async-aware.
@ 2016-06-28 09:52 Kyotaro Horiguchi <[email protected]>
  0 siblings, 0 replies; 56+ messages in thread

From: Kyotaro Horiguchi @ 2016-06-28 09:52 UTC (permalink / raw)

Change append node to be capable to handle asynchronous children
properly. As soon as it receives !async_ready from a child, it moves
to the next child and if no child is ready, it sleeps until at least
one of them become ready.
---
 src/backend/executor/nodeAppend.c | 94 ++++++++++++++++++++++++++++++++-------
 src/include/nodes/execnodes.h     |  2 +
 2 files changed, 80 insertions(+), 16 deletions(-)

diff --git a/src/backend/executor/nodeAppend.c b/src/backend/executor/nodeAppend.c
index e0ce8c6..1c0d26e 100644
--- a/src/backend/executor/nodeAppend.c
+++ b/src/backend/executor/nodeAppend.c
@@ -58,6 +58,7 @@
 #include "postgres.h"
 
 #include "executor/execdebug.h"
+#include "executor/execAsync.h"
 #include "executor/nodeAppend.h"
 
 static bool exec_append_initialize_next(AppendState *appendstate);
@@ -121,6 +122,7 @@ ExecInitAppend(Append *node, EState *estate, int eflags)
 {
 	AppendState *appendstate = makeNode(AppendState);
 	PlanState **appendplanstates;
+	bool	   *finished;
 	int			nplans;
 	int			i;
 	ListCell   *lc;
@@ -134,14 +136,17 @@ ExecInitAppend(Append *node, EState *estate, int eflags)
 	nplans = list_length(node->appendplans);
 
 	appendplanstates = (PlanState **) palloc0(nplans * sizeof(PlanState *));
-
+	finished = (bool *) palloc0(nplans * sizeof(bool));
+	
 	/*
 	 * create new AppendState for our append node
 	 */
 	appendstate->ps.plan = (Plan *) node;
 	appendstate->ps.state = estate;
 	appendstate->appendplans = appendplanstates;
+	appendstate->as_finished = finished;
 	appendstate->as_nplans = nplans;
+	appendstate->as_async = ((eflags & EXEC_FLAG_BACKWARD) == 0);
 
 	/*
 	 * Miscellaneous initialization
@@ -194,49 +199,104 @@ ExecInitAppend(Append *node, EState *estate, int eflags)
 void
 ExecAppend(AppendState *node)
 {
-	for (;;)
+	int n_notready = 0;
+	PlanState  *subnode;
+	int i, n;
+
+	n = node->as_whichplan;
+
+	for (i = 0 ; i < node->as_nplans ; i++)
 	{
-		PlanState  *subnode;
 		TupleTableSlot *result;
 
+		if (node->as_async)
+		{
+			if (n >= node->as_nplans)
+				n = 0;
+
+			if (node->as_finished[n])
+			{
+				n++;
+				continue;
+			}
+		}
+
 		/*
 		 * figure out which subplan we are currently processing
 		 */
-		subnode = node->appendplans[node->as_whichplan];
+		subnode = node->appendplans[n];
 
 		/*
-		 * get a tuple from the subplan
+		 * execute the subplan to get a result if it is not ready yet
 		 */
-		result = ExecProcNode(subnode);
+		if (!subnode->result_ready)
+			ExecExecuteNode(subnode);
+
+		/*
+		 * This subnode is not ready yet when asynchrony is not allowed,
+		 * immediately wait for this subnode.
+		 */
+		if (!subnode->result_ready)
+		{
+			if (node->as_async)
+			{
+				n_notready++;
+				n++;
+				continue;
+			}
+			ExecAsyncWaitForNode(subnode);
+		}
+
+		Assert(subnode->result_ready);
+
+		result = ExecConsumeResult((PlanState *)subnode);
 
 		if (!TupIsNull(result))
 		{
 			/*
-			 * If the subplan gave us something then return it as-is. We do
-			 * NOT make use of the result slot that was set up in
-			 * ExecInitAppend; there's no need for it.
+			 * If the subplan gave us something then return it
+			 * as-is. We do NOT make use of the result slot that was
+			 * set up in ExecInitAppend; there's no need for it.
 			 */
+			node->as_whichplan = n;
 			ExecReturnTuple(&node->ps, result);
 			return;
 		}
 
+		/* Tuple has been exhausted on this subnode */
+		if (node->as_async)
+		{
+			node->as_finished[n++] = true;
+			continue;
+		}
+
 		/*
 		 * Go on to the "next" subplan in the appropriate direction. If no
 		 * more subplans, return the empty slot set up for us by
 		 * ExecInitAppend.
 		 */
 		if (ScanDirectionIsForward(node->ps.state->es_direction))
-			node->as_whichplan++;
+		{
+			if (n >= node->as_nplans - 1)
+				break;
+			n++;
+		}
 		else
-			node->as_whichplan--;
-		if (!exec_append_initialize_next(node))
 		{
-			ExecReturnTuple(&node->ps,
-							ExecClearTuple(node->ps.ps_ResultTupleSlot));
-			return;
+			if (n == 0)
+				break;
+			n--;
 		}
+	}
 
-		/* Else loop back and try to get a tuple from the new subplan */
+	/*
+	 * We are finished if reached here and no subnodes are not-ready
+	 */
+	if (n_notready == 0)
+	{
+		node->as_whichplan = n;
+		ExecReturnTuple(&node->ps,
+						ExecClearTuple(node->ps.ps_ResultTupleSlot));
 	}
 }
 
@@ -277,6 +337,8 @@ ExecReScanAppend(AppendState *node)
 	{
 		PlanState  *subnode = node->appendplans[i];
 
+		node->as_finished[i] = false;
+
 		/*
 		 * ExecReScan doesn't know about my subplans, so I have to do
 		 * changed-parameter signaling myself.
diff --git a/src/include/nodes/execnodes.h b/src/include/nodes/execnodes.h
index 9121537..e4f2bb6 100644
--- a/src/include/nodes/execnodes.h
+++ b/src/include/nodes/execnodes.h
@@ -1170,6 +1170,8 @@ typedef struct AppendState
 {
 	PlanState	ps;				/* its first field is NodeTag */
 	PlanState **appendplans;	/* array of PlanStates for my inputs */
+	bool		as_async;		/* true to allow async execution */
+	bool	   *as_finished;	/* array of the running state of subplans */
 	int			as_nplans;
 	int			as_whichplan;
 } AppendState;
-- 
1.8.3.1


----Next_Part(Wed_Jul_06_16_29_14_2016_599)--
Content-Type: Text/Plain; charset=us-ascii
Content-Transfer-Encoding: 7bit
Content-Disposition: inline; filename="gentbl.sql"

CREATE EXTENSION postgres_fdw;
CREATE TABLE pl (a int, b int);
CREATE TABLE cl1 (LIKE pl) INHERITS (pl);
CREATE TABLE cl2 (LIKE pl) INHERITS (pl);
CREATE TABLE cl3 (LIKE pl) INHERITS (pl);
CREATE TABLE cl4 (LIKE pl) INHERITS (pl);
INSERT INTO cl1 (SELECT a, a FROM generate_series(0000000, 0999999) a);
INSERT INTO cl2 (SELECT a, a FROM generate_series(1000000, 1999999) a);
INSERT INTO cl3 (SELECT a, a FROM generate_series(2000000, 2999999) a);
INSERT INTO cl4 (SELECT a, a FROM generate_series(3000000, 3999999) a);

CREATE TABLE t0 (LIKE pl);
INSERT INTO t0 (SELECT a, a FROM generate_series(0000000, 9999999) a);

CREATE SERVER sv0 FOREIGN DATA WRAPPER postgres_fdw OPTIONS (host '/tmp', dbname 'postgres');
CREATE SERVER sv1 FOREIGN DATA WRAPPER postgres_fdw OPTIONS (host '/tmp', dbname 'postgres');
CREATE SERVER sv2 FOREIGN DATA WRAPPER postgres_fdw OPTIONS (host '/tmp', dbname 'postgres');
CREATE SERVER sv3 FOREIGN DATA WRAPPER postgres_fdw OPTIONS (host '/tmp', dbname 'postgres');
CREATE SERVER sv4 FOREIGN DATA WRAPPER postgres_fdw OPTIONS (host '/tmp', dbname 'postgres');
CREATE USER MAPPING FOR public SERVER sv0;
CREATE USER MAPPING FOR public SERVER sv1;
CREATE USER MAPPING FOR public SERVER sv2;
CREATE USER MAPPING FOR public SERVER sv3;
CREATE USER MAPPING FOR public SERVER sv4;

CREATE FOREIGN TABLE ft10 (a int, b int) SERVER sv0 OPTIONS (table_name 'cl1');
CREATE FOREIGN TABLE ft20 (a int, b int) SERVER sv0 OPTIONS (table_name 'cl2');
CREATE FOREIGN TABLE ft30 (a int, b int) SERVER sv0 OPTIONS (table_name 'cl3');
CREATE FOREIGN TABLE ft40 (a int, b int) SERVER sv0 OPTIONS (table_name 'cl4');
CREATE FOREIGN TABLE ft11 (a int, b int) SERVER sv1 OPTIONS (table_name 'cl1');
CREATE FOREIGN TABLE ft22 (a int, b int) SERVER sv2 OPTIONS (table_name 'cl2');
CREATE FOREIGN TABLE ft33 (a int, b int) SERVER sv3 OPTIONS (table_name 'cl3');
CREATE FOREIGN TABLE ft44 (a int, b int) SERVER sv4 OPTIONS (table_name 'cl4');

CREATE TABLE pf0 (LIKE pl);
ALTER FOREIGN TABLE ft10 INHERIT pf0;
ALTER FOREIGN TABLE ft20 INHERIT pf0;
ALTER FOREIGN TABLE ft30 INHERIT pf0;
ALTER FOREIGN TABLE ft40 INHERIT pf0;

CREATE table pf1 (LIKE pl);
ALTER FOREIGN TABLE ft11 INHERIT pf1;
ALTER FOREIGN TABLE ft22 INHERIT pf1;
ALTER FOREIGN TABLE ft33 INHERIT pf1;
ALTER FOREIGN TABLE ft44 INHERIT pf1;

ANALYZE;

----Next_Part(Wed_Jul_06_16_29_14_2016_599)--
Content-Type: Text/Plain; charset=us-ascii
Content-Transfer-Encoding: 7bit
Content-Disposition: inline; filename="testrun.sh"

#! /bin/bash

function do_test() {
	echo $1
	for i in $(seq 1 10); do psql postgres -c "set log_min_duration_statement = 0; set client_min_messages=log; select sum(a) from $1"; done | grep LOG
}

for i in "t0" "pl" "pf0" "pf1";
do
  do_test $i
done


----Next_Part(Wed_Jul_06_16_29_14_2016_599)--
Content-Type: text/plain
Content-Disposition: inline
Content-Transfer-Encoding: 8bit
MIME-Version: 1.0


-- 
Sent via pgsql-hackers mailing list ([email protected])
To make changes to your subscription:
http://www.postgresql.org/mailpref/pgsql-hackers

----Next_Part(Wed_Jul_06_16_29_14_2016_599)----






^ permalink  raw  reply  [nested|flat] 56+ messages in thread

* [PATCH 7/7] Make Append node async-aware.
@ 2016-06-28 09:52 Kyotaro Horiguchi <[email protected]>
  0 siblings, 0 replies; 56+ messages in thread

From: Kyotaro Horiguchi @ 2016-06-28 09:52 UTC (permalink / raw)

Change append node to be capable to handle asynchronous children
properly. As soon as it receives !async_ready from a child, it moves
to the next child and if no child is ready, it sleeps until at least
one of them become ready.
---
 src/backend/executor/nodeAppend.c | 67 +++++++++++++++++++++++++++++++++++++--
 src/include/nodes/execnodes.h     |  2 ++
 2 files changed, 67 insertions(+), 2 deletions(-)

diff --git a/src/backend/executor/nodeAppend.c b/src/backend/executor/nodeAppend.c
index e0ce8c6..9a4063a 100644
--- a/src/backend/executor/nodeAppend.c
+++ b/src/backend/executor/nodeAppend.c
@@ -58,6 +58,7 @@
 #include "postgres.h"
 
 #include "executor/execdebug.h"
+#include "executor/execAsync.h"
 #include "executor/nodeAppend.h"
 
 static bool exec_append_initialize_next(AppendState *appendstate);
@@ -121,6 +122,7 @@ ExecInitAppend(Append *node, EState *estate, int eflags)
 {
 	AppendState *appendstate = makeNode(AppendState);
 	PlanState **appendplanstates;
+	bool	   *finished;
 	int			nplans;
 	int			i;
 	ListCell   *lc;
@@ -134,14 +136,17 @@ ExecInitAppend(Append *node, EState *estate, int eflags)
 	nplans = list_length(node->appendplans);
 
 	appendplanstates = (PlanState **) palloc0(nplans * sizeof(PlanState *));
-
+	finished = (bool *) palloc0(nplans * sizeof(bool));
+	
 	/*
 	 * create new AppendState for our append node
 	 */
 	appendstate->ps.plan = (Plan *) node;
 	appendstate->ps.state = estate;
 	appendstate->appendplans = appendplanstates;
+	appendstate->as_finished = finished;
 	appendstate->as_nplans = nplans;
+	appendstate->as_async = ((eflags & EXEC_FLAG_BACKWARD) == 0);
 
 	/*
 	 * Miscellaneous initialization
@@ -194,6 +199,8 @@ ExecInitAppend(Append *node, EState *estate, int eflags)
 void
 ExecAppend(AppendState *node)
 {
+	int stopplan = node->as_whichplan;
+
 	for (;;)
 	{
 		PlanState  *subnode;
@@ -207,7 +214,36 @@ ExecAppend(AppendState *node)
 		/*
 		 * get a tuple from the subplan
 		 */
-		result = ExecProcNode(subnode);
+		Assert(!node->as_finished[node->as_whichplan]);
+
+		if (!subnode->result_ready)
+			ExecExecuteNode(subnode);
+
+		if (!subnode->result_ready)
+		{
+			if (node->as_async)
+			{
+				/* Move to the next living node */
+				do
+				{
+					node->as_whichplan = 
+						(node->as_whichplan + 1) %  node->as_nplans;
+				} while (node->as_whichplan != stopplan &&
+						 node->as_finished[node->as_whichplan]);
+
+				/* No node is ready yet, return as not-ready */
+				if (node->as_whichplan == stopplan)
+					return;
+
+				/* Try the next node */
+				continue;
+			}
+
+			/* If not async, immediately wait for this subnode */
+			ExecAsyncWaitForNode(subnode);
+		}				
+
+		result = ExecConsumeResult((PlanState *) subnode);
 
 		if (!TupIsNull(result))
 		{
@@ -220,6 +256,31 @@ ExecAppend(AppendState *node)
 			return;
 		}
 
+		if (node->as_async)
+		{
+			node->as_finished[node->as_whichplan] = true;
+			stopplan = node->as_whichplan;
+
+			/* Find the next living subnode */
+			do
+			{
+				node->as_whichplan =
+					(node->as_whichplan + 1) % node->as_nplans;
+			} while (node->as_whichplan != stopplan &&
+					 node->as_finished[node->as_whichplan]);
+
+			if (node->as_whichplan != stopplan)
+			{
+				stopplan = node->as_whichplan;
+				continue;
+			}
+
+			/* All subnodes are exhausted. Finish this node. */
+			ExecReturnTuple(&node->ps,
+							ExecClearTuple(node->ps.ps_ResultTupleSlot));
+			return;
+		}
+
 		/*
 		 * Go on to the "next" subplan in the appropriate direction. If no
 		 * more subplans, return the empty slot set up for us by
@@ -277,6 +338,8 @@ ExecReScanAppend(AppendState *node)
 	{
 		PlanState  *subnode = node->appendplans[i];
 
+		node->as_finished[i] = false;
+
 		/*
 		 * ExecReScan doesn't know about my subplans, so I have to do
 		 * changed-parameter signaling myself.
diff --git a/src/include/nodes/execnodes.h b/src/include/nodes/execnodes.h
index b72decc..b0a86c5 100644
--- a/src/include/nodes/execnodes.h
+++ b/src/include/nodes/execnodes.h
@@ -1177,6 +1177,8 @@ typedef struct AppendState
 {
 	PlanState	ps;				/* its first field is NodeTag */
 	PlanState **appendplans;	/* array of PlanStates for my inputs */
+	bool		as_async;		/* true to allow async execution */
+	bool	   *as_finished;	/* array of the running state of subplans */
 	int			as_nplans;
 	int			as_whichplan;
 } AppendState;
-- 
2.9.2


----Next_Part(Mon_Aug_29_17_08_36_2016_213)--
Content-Type: text/plain
Content-Disposition: inline
Content-Transfer-Encoding: 8bit
MIME-Version: 1.0


-- 
Sent via pgsql-hackers mailing list ([email protected])
To make changes to your subscription:
http://www.postgresql.org/mailpref/pgsql-hackers

----Next_Part(Mon_Aug_29_17_08_36_2016_213)----






^ permalink  raw  reply  [nested|flat] 56+ messages in thread

* [PATCH 7/7] Make Append node async-aware.
@ 2016-06-28 09:52 Kyotaro Horiguchi <[email protected]>
  0 siblings, 0 replies; 56+ messages in thread

From: Kyotaro Horiguchi @ 2016-06-28 09:52 UTC (permalink / raw)

Change append node to be capable to handle asynchronous children
properly. As soon as it receives !async_ready from a child, it moves
to the next child and if no child is ready, it sleeps until at least
one of them become ready.
---
 src/backend/executor/nodeAppend.c | 94 ++++++++++++++++++++++++++++++++-------
 src/include/nodes/execnodes.h     |  2 +
 2 files changed, 80 insertions(+), 16 deletions(-)

diff --git a/src/backend/executor/nodeAppend.c b/src/backend/executor/nodeAppend.c
index e0ce8c6..1c0d26e 100644
--- a/src/backend/executor/nodeAppend.c
+++ b/src/backend/executor/nodeAppend.c
@@ -58,6 +58,7 @@
 #include "postgres.h"
 
 #include "executor/execdebug.h"
+#include "executor/execAsync.h"
 #include "executor/nodeAppend.h"
 
 static bool exec_append_initialize_next(AppendState *appendstate);
@@ -121,6 +122,7 @@ ExecInitAppend(Append *node, EState *estate, int eflags)
 {
 	AppendState *appendstate = makeNode(AppendState);
 	PlanState **appendplanstates;
+	bool	   *finished;
 	int			nplans;
 	int			i;
 	ListCell   *lc;
@@ -134,14 +136,17 @@ ExecInitAppend(Append *node, EState *estate, int eflags)
 	nplans = list_length(node->appendplans);
 
 	appendplanstates = (PlanState **) palloc0(nplans * sizeof(PlanState *));
-
+	finished = (bool *) palloc0(nplans * sizeof(bool));
+	
 	/*
 	 * create new AppendState for our append node
 	 */
 	appendstate->ps.plan = (Plan *) node;
 	appendstate->ps.state = estate;
 	appendstate->appendplans = appendplanstates;
+	appendstate->as_finished = finished;
 	appendstate->as_nplans = nplans;
+	appendstate->as_async = ((eflags & EXEC_FLAG_BACKWARD) == 0);
 
 	/*
 	 * Miscellaneous initialization
@@ -194,49 +199,104 @@ ExecInitAppend(Append *node, EState *estate, int eflags)
 void
 ExecAppend(AppendState *node)
 {
-	for (;;)
+	int n_notready = 0;
+	PlanState  *subnode;
+	int i, n;
+
+	n = node->as_whichplan;
+
+	for (i = 0 ; i < node->as_nplans ; i++)
 	{
-		PlanState  *subnode;
 		TupleTableSlot *result;
 
+		if (node->as_async)
+		{
+			if (n >= node->as_nplans)
+				n = 0;
+
+			if (node->as_finished[n])
+			{
+				n++;
+				continue;
+			}
+		}
+
 		/*
 		 * figure out which subplan we are currently processing
 		 */
-		subnode = node->appendplans[node->as_whichplan];
+		subnode = node->appendplans[n];
 
 		/*
-		 * get a tuple from the subplan
+		 * execute the subplan to get a result if it is not ready yet
 		 */
-		result = ExecProcNode(subnode);
+		if (!subnode->result_ready)
+			ExecExecuteNode(subnode);
+
+		/*
+		 * This subnode is not ready yet when asynchrony is not allowed,
+		 * immediately wait for this subnode.
+		 */
+		if (!subnode->result_ready)
+		{
+			if (node->as_async)
+			{
+				n_notready++;
+				n++;
+				continue;
+			}
+			ExecAsyncWaitForNode(subnode);
+		}
+
+		Assert(subnode->result_ready);
+
+		result = ExecConsumeResult((PlanState *)subnode);
 
 		if (!TupIsNull(result))
 		{
 			/*
-			 * If the subplan gave us something then return it as-is. We do
-			 * NOT make use of the result slot that was set up in
-			 * ExecInitAppend; there's no need for it.
+			 * If the subplan gave us something then return it
+			 * as-is. We do NOT make use of the result slot that was
+			 * set up in ExecInitAppend; there's no need for it.
 			 */
+			node->as_whichplan = n;
 			ExecReturnTuple(&node->ps, result);
 			return;
 		}
 
+		/* Tuple has been exhausted on this subnode */
+		if (node->as_async)
+		{
+			node->as_finished[n++] = true;
+			continue;
+		}
+
 		/*
 		 * Go on to the "next" subplan in the appropriate direction. If no
 		 * more subplans, return the empty slot set up for us by
 		 * ExecInitAppend.
 		 */
 		if (ScanDirectionIsForward(node->ps.state->es_direction))
-			node->as_whichplan++;
+		{
+			if (n >= node->as_nplans - 1)
+				break;
+			n++;
+		}
 		else
-			node->as_whichplan--;
-		if (!exec_append_initialize_next(node))
 		{
-			ExecReturnTuple(&node->ps,
-							ExecClearTuple(node->ps.ps_ResultTupleSlot));
-			return;
+			if (n == 0)
+				break;
+			n--;
 		}
+	}
 
-		/* Else loop back and try to get a tuple from the new subplan */
+	/*
+	 * We are finished if reached here and no subnodes are not-ready
+	 */
+	if (n_notready == 0)
+	{
+		node->as_whichplan = n;
+		ExecReturnTuple(&node->ps,
+						ExecClearTuple(node->ps.ps_ResultTupleSlot));
 	}
 }
 
@@ -277,6 +337,8 @@ ExecReScanAppend(AppendState *node)
 	{
 		PlanState  *subnode = node->appendplans[i];
 
+		node->as_finished[i] = false;
+
 		/*
 		 * ExecReScan doesn't know about my subplans, so I have to do
 		 * changed-parameter signaling myself.
diff --git a/src/include/nodes/execnodes.h b/src/include/nodes/execnodes.h
index 9121537..e4f2bb6 100644
--- a/src/include/nodes/execnodes.h
+++ b/src/include/nodes/execnodes.h
@@ -1170,6 +1170,8 @@ typedef struct AppendState
 {
 	PlanState	ps;				/* its first field is NodeTag */
 	PlanState **appendplans;	/* array of PlanStates for my inputs */
+	bool		as_async;		/* true to allow async execution */
+	bool	   *as_finished;	/* array of the running state of subplans */
 	int			as_nplans;
 	int			as_whichplan;
 } AppendState;
-- 
1.8.3.1


----Next_Part(Wed_Jul_06_16_29_14_2016_599)--
Content-Type: Text/Plain; charset=us-ascii
Content-Transfer-Encoding: 7bit
Content-Disposition: inline; filename="gentbl.sql"

CREATE EXTENSION postgres_fdw;
CREATE TABLE pl (a int, b int);
CREATE TABLE cl1 (LIKE pl) INHERITS (pl);
CREATE TABLE cl2 (LIKE pl) INHERITS (pl);
CREATE TABLE cl3 (LIKE pl) INHERITS (pl);
CREATE TABLE cl4 (LIKE pl) INHERITS (pl);
INSERT INTO cl1 (SELECT a, a FROM generate_series(0000000, 0999999) a);
INSERT INTO cl2 (SELECT a, a FROM generate_series(1000000, 1999999) a);
INSERT INTO cl3 (SELECT a, a FROM generate_series(2000000, 2999999) a);
INSERT INTO cl4 (SELECT a, a FROM generate_series(3000000, 3999999) a);

CREATE TABLE t0 (LIKE pl);
INSERT INTO t0 (SELECT a, a FROM generate_series(0000000, 9999999) a);

CREATE SERVER sv0 FOREIGN DATA WRAPPER postgres_fdw OPTIONS (host '/tmp', dbname 'postgres');
CREATE SERVER sv1 FOREIGN DATA WRAPPER postgres_fdw OPTIONS (host '/tmp', dbname 'postgres');
CREATE SERVER sv2 FOREIGN DATA WRAPPER postgres_fdw OPTIONS (host '/tmp', dbname 'postgres');
CREATE SERVER sv3 FOREIGN DATA WRAPPER postgres_fdw OPTIONS (host '/tmp', dbname 'postgres');
CREATE SERVER sv4 FOREIGN DATA WRAPPER postgres_fdw OPTIONS (host '/tmp', dbname 'postgres');
CREATE USER MAPPING FOR public SERVER sv0;
CREATE USER MAPPING FOR public SERVER sv1;
CREATE USER MAPPING FOR public SERVER sv2;
CREATE USER MAPPING FOR public SERVER sv3;
CREATE USER MAPPING FOR public SERVER sv4;

CREATE FOREIGN TABLE ft10 (a int, b int) SERVER sv0 OPTIONS (table_name 'cl1');
CREATE FOREIGN TABLE ft20 (a int, b int) SERVER sv0 OPTIONS (table_name 'cl2');
CREATE FOREIGN TABLE ft30 (a int, b int) SERVER sv0 OPTIONS (table_name 'cl3');
CREATE FOREIGN TABLE ft40 (a int, b int) SERVER sv0 OPTIONS (table_name 'cl4');
CREATE FOREIGN TABLE ft11 (a int, b int) SERVER sv1 OPTIONS (table_name 'cl1');
CREATE FOREIGN TABLE ft22 (a int, b int) SERVER sv2 OPTIONS (table_name 'cl2');
CREATE FOREIGN TABLE ft33 (a int, b int) SERVER sv3 OPTIONS (table_name 'cl3');
CREATE FOREIGN TABLE ft44 (a int, b int) SERVER sv4 OPTIONS (table_name 'cl4');

CREATE TABLE pf0 (LIKE pl);
ALTER FOREIGN TABLE ft10 INHERIT pf0;
ALTER FOREIGN TABLE ft20 INHERIT pf0;
ALTER FOREIGN TABLE ft30 INHERIT pf0;
ALTER FOREIGN TABLE ft40 INHERIT pf0;

CREATE table pf1 (LIKE pl);
ALTER FOREIGN TABLE ft11 INHERIT pf1;
ALTER FOREIGN TABLE ft22 INHERIT pf1;
ALTER FOREIGN TABLE ft33 INHERIT pf1;
ALTER FOREIGN TABLE ft44 INHERIT pf1;

ANALYZE;

----Next_Part(Wed_Jul_06_16_29_14_2016_599)--
Content-Type: Text/Plain; charset=us-ascii
Content-Transfer-Encoding: 7bit
Content-Disposition: inline; filename="testrun.sh"

#! /bin/bash

function do_test() {
	echo $1
	for i in $(seq 1 10); do psql postgres -c "set log_min_duration_statement = 0; set client_min_messages=log; select sum(a) from $1"; done | grep LOG
}

for i in "t0" "pl" "pf0" "pf1";
do
  do_test $i
done


----Next_Part(Wed_Jul_06_16_29_14_2016_599)--
Content-Type: text/plain
Content-Disposition: inline
Content-Transfer-Encoding: 8bit
MIME-Version: 1.0


-- 
Sent via pgsql-hackers mailing list ([email protected])
To make changes to your subscription:
http://www.postgresql.org/mailpref/pgsql-hackers

----Next_Part(Wed_Jul_06_16_29_14_2016_599)----






^ permalink  raw  reply  [nested|flat] 56+ messages in thread

* [PATCH 7/7] Make Append node async-aware.
@ 2016-06-28 09:52 Kyotaro Horiguchi <[email protected]>
  0 siblings, 0 replies; 56+ messages in thread

From: Kyotaro Horiguchi @ 2016-06-28 09:52 UTC (permalink / raw)

Change append node to be capable to handle asynchronous children
properly. As soon as it receives !async_ready from a child, it moves
to the next child and if no child is ready, it sleeps until at least
one of them become ready.
---
 src/backend/executor/nodeAppend.c | 67 +++++++++++++++++++++++++++++++++++++--
 src/include/nodes/execnodes.h     |  2 ++
 2 files changed, 67 insertions(+), 2 deletions(-)

diff --git a/src/backend/executor/nodeAppend.c b/src/backend/executor/nodeAppend.c
index e0ce8c6..9a4063a 100644
--- a/src/backend/executor/nodeAppend.c
+++ b/src/backend/executor/nodeAppend.c
@@ -58,6 +58,7 @@
 #include "postgres.h"
 
 #include "executor/execdebug.h"
+#include "executor/execAsync.h"
 #include "executor/nodeAppend.h"
 
 static bool exec_append_initialize_next(AppendState *appendstate);
@@ -121,6 +122,7 @@ ExecInitAppend(Append *node, EState *estate, int eflags)
 {
 	AppendState *appendstate = makeNode(AppendState);
 	PlanState **appendplanstates;
+	bool	   *finished;
 	int			nplans;
 	int			i;
 	ListCell   *lc;
@@ -134,14 +136,17 @@ ExecInitAppend(Append *node, EState *estate, int eflags)
 	nplans = list_length(node->appendplans);
 
 	appendplanstates = (PlanState **) palloc0(nplans * sizeof(PlanState *));
-
+	finished = (bool *) palloc0(nplans * sizeof(bool));
+	
 	/*
 	 * create new AppendState for our append node
 	 */
 	appendstate->ps.plan = (Plan *) node;
 	appendstate->ps.state = estate;
 	appendstate->appendplans = appendplanstates;
+	appendstate->as_finished = finished;
 	appendstate->as_nplans = nplans;
+	appendstate->as_async = ((eflags & EXEC_FLAG_BACKWARD) == 0);
 
 	/*
 	 * Miscellaneous initialization
@@ -194,6 +199,8 @@ ExecInitAppend(Append *node, EState *estate, int eflags)
 void
 ExecAppend(AppendState *node)
 {
+	int stopplan = node->as_whichplan;
+
 	for (;;)
 	{
 		PlanState  *subnode;
@@ -207,7 +214,36 @@ ExecAppend(AppendState *node)
 		/*
 		 * get a tuple from the subplan
 		 */
-		result = ExecProcNode(subnode);
+		Assert(!node->as_finished[node->as_whichplan]);
+
+		if (!subnode->result_ready)
+			ExecExecuteNode(subnode);
+
+		if (!subnode->result_ready)
+		{
+			if (node->as_async)
+			{
+				/* Move to the next living node */
+				do
+				{
+					node->as_whichplan = 
+						(node->as_whichplan + 1) %  node->as_nplans;
+				} while (node->as_whichplan != stopplan &&
+						 node->as_finished[node->as_whichplan]);
+
+				/* No node is ready yet, return as not-ready */
+				if (node->as_whichplan == stopplan)
+					return;
+
+				/* Try the next node */
+				continue;
+			}
+
+			/* If not async, immediately wait for this subnode */
+			ExecAsyncWaitForNode(subnode);
+		}				
+
+		result = ExecConsumeResult((PlanState *) subnode);
 
 		if (!TupIsNull(result))
 		{
@@ -220,6 +256,31 @@ ExecAppend(AppendState *node)
 			return;
 		}
 
+		if (node->as_async)
+		{
+			node->as_finished[node->as_whichplan] = true;
+			stopplan = node->as_whichplan;
+
+			/* Find the next living subnode */
+			do
+			{
+				node->as_whichplan =
+					(node->as_whichplan + 1) % node->as_nplans;
+			} while (node->as_whichplan != stopplan &&
+					 node->as_finished[node->as_whichplan]);
+
+			if (node->as_whichplan != stopplan)
+			{
+				stopplan = node->as_whichplan;
+				continue;
+			}
+
+			/* All subnodes are exhausted. Finish this node. */
+			ExecReturnTuple(&node->ps,
+							ExecClearTuple(node->ps.ps_ResultTupleSlot));
+			return;
+		}
+
 		/*
 		 * Go on to the "next" subplan in the appropriate direction. If no
 		 * more subplans, return the empty slot set up for us by
@@ -277,6 +338,8 @@ ExecReScanAppend(AppendState *node)
 	{
 		PlanState  *subnode = node->appendplans[i];
 
+		node->as_finished[i] = false;
+
 		/*
 		 * ExecReScan doesn't know about my subplans, so I have to do
 		 * changed-parameter signaling myself.
diff --git a/src/include/nodes/execnodes.h b/src/include/nodes/execnodes.h
index b72decc..b0a86c5 100644
--- a/src/include/nodes/execnodes.h
+++ b/src/include/nodes/execnodes.h
@@ -1177,6 +1177,8 @@ typedef struct AppendState
 {
 	PlanState	ps;				/* its first field is NodeTag */
 	PlanState **appendplans;	/* array of PlanStates for my inputs */
+	bool		as_async;		/* true to allow async execution */
+	bool	   *as_finished;	/* array of the running state of subplans */
 	int			as_nplans;
 	int			as_whichplan;
 } AppendState;
-- 
1.8.3.1


----Next_Part(Thu_Jul_21_18_50_07_2016_597)--
Content-Type: text/plain
Content-Disposition: inline
Content-Transfer-Encoding: 8bit
MIME-Version: 1.0


-- 
Sent via pgsql-hackers mailing list ([email protected])
To make changes to your subscription:
http://www.postgresql.org/mailpref/pgsql-hackers

----Next_Part(Thu_Jul_21_18_50_07_2016_597)----






^ permalink  raw  reply  [nested|flat] 56+ messages in thread

* [PATCH 7/7] Make Append node async-aware.
@ 2016-06-28 09:52 Kyotaro Horiguchi <[email protected]>
  0 siblings, 0 replies; 56+ messages in thread

From: Kyotaro Horiguchi @ 2016-06-28 09:52 UTC (permalink / raw)

Change append node to be capable to handle asynchronous children
properly. As soon as it receives !async_ready from a child, it moves
to the next child and if no child is ready, it sleeps until at least
one of them become ready.
---
 src/backend/executor/nodeAppend.c | 94 ++++++++++++++++++++++++++++++++-------
 src/include/nodes/execnodes.h     |  2 +
 2 files changed, 80 insertions(+), 16 deletions(-)

diff --git a/src/backend/executor/nodeAppend.c b/src/backend/executor/nodeAppend.c
index e0ce8c6..1c0d26e 100644
--- a/src/backend/executor/nodeAppend.c
+++ b/src/backend/executor/nodeAppend.c
@@ -58,6 +58,7 @@
 #include "postgres.h"
 
 #include "executor/execdebug.h"
+#include "executor/execAsync.h"
 #include "executor/nodeAppend.h"
 
 static bool exec_append_initialize_next(AppendState *appendstate);
@@ -121,6 +122,7 @@ ExecInitAppend(Append *node, EState *estate, int eflags)
 {
 	AppendState *appendstate = makeNode(AppendState);
 	PlanState **appendplanstates;
+	bool	   *finished;
 	int			nplans;
 	int			i;
 	ListCell   *lc;
@@ -134,14 +136,17 @@ ExecInitAppend(Append *node, EState *estate, int eflags)
 	nplans = list_length(node->appendplans);
 
 	appendplanstates = (PlanState **) palloc0(nplans * sizeof(PlanState *));
-
+	finished = (bool *) palloc0(nplans * sizeof(bool));
+	
 	/*
 	 * create new AppendState for our append node
 	 */
 	appendstate->ps.plan = (Plan *) node;
 	appendstate->ps.state = estate;
 	appendstate->appendplans = appendplanstates;
+	appendstate->as_finished = finished;
 	appendstate->as_nplans = nplans;
+	appendstate->as_async = ((eflags & EXEC_FLAG_BACKWARD) == 0);
 
 	/*
 	 * Miscellaneous initialization
@@ -194,49 +199,104 @@ ExecInitAppend(Append *node, EState *estate, int eflags)
 void
 ExecAppend(AppendState *node)
 {
-	for (;;)
+	int n_notready = 0;
+	PlanState  *subnode;
+	int i, n;
+
+	n = node->as_whichplan;
+
+	for (i = 0 ; i < node->as_nplans ; i++)
 	{
-		PlanState  *subnode;
 		TupleTableSlot *result;
 
+		if (node->as_async)
+		{
+			if (n >= node->as_nplans)
+				n = 0;
+
+			if (node->as_finished[n])
+			{
+				n++;
+				continue;
+			}
+		}
+
 		/*
 		 * figure out which subplan we are currently processing
 		 */
-		subnode = node->appendplans[node->as_whichplan];
+		subnode = node->appendplans[n];
 
 		/*
-		 * get a tuple from the subplan
+		 * execute the subplan to get a result if it is not ready yet
 		 */
-		result = ExecProcNode(subnode);
+		if (!subnode->result_ready)
+			ExecExecuteNode(subnode);
+
+		/*
+		 * This subnode is not ready yet when asynchrony is not allowed,
+		 * immediately wait for this subnode.
+		 */
+		if (!subnode->result_ready)
+		{
+			if (node->as_async)
+			{
+				n_notready++;
+				n++;
+				continue;
+			}
+			ExecAsyncWaitForNode(subnode);
+		}
+
+		Assert(subnode->result_ready);
+
+		result = ExecConsumeResult((PlanState *)subnode);
 
 		if (!TupIsNull(result))
 		{
 			/*
-			 * If the subplan gave us something then return it as-is. We do
-			 * NOT make use of the result slot that was set up in
-			 * ExecInitAppend; there's no need for it.
+			 * If the subplan gave us something then return it
+			 * as-is. We do NOT make use of the result slot that was
+			 * set up in ExecInitAppend; there's no need for it.
 			 */
+			node->as_whichplan = n;
 			ExecReturnTuple(&node->ps, result);
 			return;
 		}
 
+		/* Tuple has been exhausted on this subnode */
+		if (node->as_async)
+		{
+			node->as_finished[n++] = true;
+			continue;
+		}
+
 		/*
 		 * Go on to the "next" subplan in the appropriate direction. If no
 		 * more subplans, return the empty slot set up for us by
 		 * ExecInitAppend.
 		 */
 		if (ScanDirectionIsForward(node->ps.state->es_direction))
-			node->as_whichplan++;
+		{
+			if (n >= node->as_nplans - 1)
+				break;
+			n++;
+		}
 		else
-			node->as_whichplan--;
-		if (!exec_append_initialize_next(node))
 		{
-			ExecReturnTuple(&node->ps,
-							ExecClearTuple(node->ps.ps_ResultTupleSlot));
-			return;
+			if (n == 0)
+				break;
+			n--;
 		}
+	}
 
-		/* Else loop back and try to get a tuple from the new subplan */
+	/*
+	 * We are finished if reached here and no subnodes are not-ready
+	 */
+	if (n_notready == 0)
+	{
+		node->as_whichplan = n;
+		ExecReturnTuple(&node->ps,
+						ExecClearTuple(node->ps.ps_ResultTupleSlot));
 	}
 }
 
@@ -277,6 +337,8 @@ ExecReScanAppend(AppendState *node)
 	{
 		PlanState  *subnode = node->appendplans[i];
 
+		node->as_finished[i] = false;
+
 		/*
 		 * ExecReScan doesn't know about my subplans, so I have to do
 		 * changed-parameter signaling myself.
diff --git a/src/include/nodes/execnodes.h b/src/include/nodes/execnodes.h
index 9121537..e4f2bb6 100644
--- a/src/include/nodes/execnodes.h
+++ b/src/include/nodes/execnodes.h
@@ -1170,6 +1170,8 @@ typedef struct AppendState
 {
 	PlanState	ps;				/* its first field is NodeTag */
 	PlanState **appendplans;	/* array of PlanStates for my inputs */
+	bool		as_async;		/* true to allow async execution */
+	bool	   *as_finished;	/* array of the running state of subplans */
 	int			as_nplans;
 	int			as_whichplan;
 } AppendState;
-- 
1.8.3.1


----Next_Part(Wed_Jul_06_16_29_14_2016_599)--
Content-Type: Text/Plain; charset=us-ascii
Content-Transfer-Encoding: 7bit
Content-Disposition: inline; filename="gentbl.sql"

CREATE EXTENSION postgres_fdw;
CREATE TABLE pl (a int, b int);
CREATE TABLE cl1 (LIKE pl) INHERITS (pl);
CREATE TABLE cl2 (LIKE pl) INHERITS (pl);
CREATE TABLE cl3 (LIKE pl) INHERITS (pl);
CREATE TABLE cl4 (LIKE pl) INHERITS (pl);
INSERT INTO cl1 (SELECT a, a FROM generate_series(0000000, 0999999) a);
INSERT INTO cl2 (SELECT a, a FROM generate_series(1000000, 1999999) a);
INSERT INTO cl3 (SELECT a, a FROM generate_series(2000000, 2999999) a);
INSERT INTO cl4 (SELECT a, a FROM generate_series(3000000, 3999999) a);

CREATE TABLE t0 (LIKE pl);
INSERT INTO t0 (SELECT a, a FROM generate_series(0000000, 9999999) a);

CREATE SERVER sv0 FOREIGN DATA WRAPPER postgres_fdw OPTIONS (host '/tmp', dbname 'postgres');
CREATE SERVER sv1 FOREIGN DATA WRAPPER postgres_fdw OPTIONS (host '/tmp', dbname 'postgres');
CREATE SERVER sv2 FOREIGN DATA WRAPPER postgres_fdw OPTIONS (host '/tmp', dbname 'postgres');
CREATE SERVER sv3 FOREIGN DATA WRAPPER postgres_fdw OPTIONS (host '/tmp', dbname 'postgres');
CREATE SERVER sv4 FOREIGN DATA WRAPPER postgres_fdw OPTIONS (host '/tmp', dbname 'postgres');
CREATE USER MAPPING FOR public SERVER sv0;
CREATE USER MAPPING FOR public SERVER sv1;
CREATE USER MAPPING FOR public SERVER sv2;
CREATE USER MAPPING FOR public SERVER sv3;
CREATE USER MAPPING FOR public SERVER sv4;

CREATE FOREIGN TABLE ft10 (a int, b int) SERVER sv0 OPTIONS (table_name 'cl1');
CREATE FOREIGN TABLE ft20 (a int, b int) SERVER sv0 OPTIONS (table_name 'cl2');
CREATE FOREIGN TABLE ft30 (a int, b int) SERVER sv0 OPTIONS (table_name 'cl3');
CREATE FOREIGN TABLE ft40 (a int, b int) SERVER sv0 OPTIONS (table_name 'cl4');
CREATE FOREIGN TABLE ft11 (a int, b int) SERVER sv1 OPTIONS (table_name 'cl1');
CREATE FOREIGN TABLE ft22 (a int, b int) SERVER sv2 OPTIONS (table_name 'cl2');
CREATE FOREIGN TABLE ft33 (a int, b int) SERVER sv3 OPTIONS (table_name 'cl3');
CREATE FOREIGN TABLE ft44 (a int, b int) SERVER sv4 OPTIONS (table_name 'cl4');

CREATE TABLE pf0 (LIKE pl);
ALTER FOREIGN TABLE ft10 INHERIT pf0;
ALTER FOREIGN TABLE ft20 INHERIT pf0;
ALTER FOREIGN TABLE ft30 INHERIT pf0;
ALTER FOREIGN TABLE ft40 INHERIT pf0;

CREATE table pf1 (LIKE pl);
ALTER FOREIGN TABLE ft11 INHERIT pf1;
ALTER FOREIGN TABLE ft22 INHERIT pf1;
ALTER FOREIGN TABLE ft33 INHERIT pf1;
ALTER FOREIGN TABLE ft44 INHERIT pf1;

ANALYZE;

----Next_Part(Wed_Jul_06_16_29_14_2016_599)--
Content-Type: Text/Plain; charset=us-ascii
Content-Transfer-Encoding: 7bit
Content-Disposition: inline; filename="testrun.sh"

#! /bin/bash

function do_test() {
	echo $1
	for i in $(seq 1 10); do psql postgres -c "set log_min_duration_statement = 0; set client_min_messages=log; select sum(a) from $1"; done | grep LOG
}

for i in "t0" "pl" "pf0" "pf1";
do
  do_test $i
done


----Next_Part(Wed_Jul_06_16_29_14_2016_599)--
Content-Type: text/plain
Content-Disposition: inline
Content-Transfer-Encoding: 8bit
MIME-Version: 1.0


-- 
Sent via pgsql-hackers mailing list ([email protected])
To make changes to your subscription:
http://www.postgresql.org/mailpref/pgsql-hackers

----Next_Part(Wed_Jul_06_16_29_14_2016_599)----






^ permalink  raw  reply  [nested|flat] 56+ messages in thread

* [PATCH 7/7] Make Append node async-aware.
@ 2016-06-28 09:52 Kyotaro Horiguchi <[email protected]>
  0 siblings, 0 replies; 56+ messages in thread

From: Kyotaro Horiguchi @ 2016-06-28 09:52 UTC (permalink / raw)

Change append node to be capable to handle asynchronous children
properly. As soon as it receives !async_ready from a child, it moves
to the next child and if no child is ready, it sleeps until at least
one of them become ready.
---
 src/backend/executor/nodeAppend.c | 67 +++++++++++++++++++++++++++++++++++++--
 src/include/nodes/execnodes.h     |  2 ++
 2 files changed, 67 insertions(+), 2 deletions(-)

diff --git a/src/backend/executor/nodeAppend.c b/src/backend/executor/nodeAppend.c
index e0ce8c6..9a4063a 100644
--- a/src/backend/executor/nodeAppend.c
+++ b/src/backend/executor/nodeAppend.c
@@ -58,6 +58,7 @@
 #include "postgres.h"
 
 #include "executor/execdebug.h"
+#include "executor/execAsync.h"
 #include "executor/nodeAppend.h"
 
 static bool exec_append_initialize_next(AppendState *appendstate);
@@ -121,6 +122,7 @@ ExecInitAppend(Append *node, EState *estate, int eflags)
 {
 	AppendState *appendstate = makeNode(AppendState);
 	PlanState **appendplanstates;
+	bool	   *finished;
 	int			nplans;
 	int			i;
 	ListCell   *lc;
@@ -134,14 +136,17 @@ ExecInitAppend(Append *node, EState *estate, int eflags)
 	nplans = list_length(node->appendplans);
 
 	appendplanstates = (PlanState **) palloc0(nplans * sizeof(PlanState *));
-
+	finished = (bool *) palloc0(nplans * sizeof(bool));
+	
 	/*
 	 * create new AppendState for our append node
 	 */
 	appendstate->ps.plan = (Plan *) node;
 	appendstate->ps.state = estate;
 	appendstate->appendplans = appendplanstates;
+	appendstate->as_finished = finished;
 	appendstate->as_nplans = nplans;
+	appendstate->as_async = ((eflags & EXEC_FLAG_BACKWARD) == 0);
 
 	/*
 	 * Miscellaneous initialization
@@ -194,6 +199,8 @@ ExecInitAppend(Append *node, EState *estate, int eflags)
 void
 ExecAppend(AppendState *node)
 {
+	int stopplan = node->as_whichplan;
+
 	for (;;)
 	{
 		PlanState  *subnode;
@@ -207,7 +214,36 @@ ExecAppend(AppendState *node)
 		/*
 		 * get a tuple from the subplan
 		 */
-		result = ExecProcNode(subnode);
+		Assert(!node->as_finished[node->as_whichplan]);
+
+		if (!subnode->result_ready)
+			ExecExecuteNode(subnode);
+
+		if (!subnode->result_ready)
+		{
+			if (node->as_async)
+			{
+				/* Move to the next living node */
+				do
+				{
+					node->as_whichplan = 
+						(node->as_whichplan + 1) %  node->as_nplans;
+				} while (node->as_whichplan != stopplan &&
+						 node->as_finished[node->as_whichplan]);
+
+				/* No node is ready yet, return as not-ready */
+				if (node->as_whichplan == stopplan)
+					return;
+
+				/* Try the next node */
+				continue;
+			}
+
+			/* If not async, immediately wait for this subnode */
+			ExecAsyncWaitForNode(subnode);
+		}				
+
+		result = ExecConsumeResult((PlanState *) subnode);
 
 		if (!TupIsNull(result))
 		{
@@ -220,6 +256,31 @@ ExecAppend(AppendState *node)
 			return;
 		}
 
+		if (node->as_async)
+		{
+			node->as_finished[node->as_whichplan] = true;
+			stopplan = node->as_whichplan;
+
+			/* Find the next living subnode */
+			do
+			{
+				node->as_whichplan =
+					(node->as_whichplan + 1) % node->as_nplans;
+			} while (node->as_whichplan != stopplan &&
+					 node->as_finished[node->as_whichplan]);
+
+			if (node->as_whichplan != stopplan)
+			{
+				stopplan = node->as_whichplan;
+				continue;
+			}
+
+			/* All subnodes are exhausted. Finish this node. */
+			ExecReturnTuple(&node->ps,
+							ExecClearTuple(node->ps.ps_ResultTupleSlot));
+			return;
+		}
+
 		/*
 		 * Go on to the "next" subplan in the appropriate direction. If no
 		 * more subplans, return the empty slot set up for us by
@@ -277,6 +338,8 @@ ExecReScanAppend(AppendState *node)
 	{
 		PlanState  *subnode = node->appendplans[i];
 
+		node->as_finished[i] = false;
+
 		/*
 		 * ExecReScan doesn't know about my subplans, so I have to do
 		 * changed-parameter signaling myself.
diff --git a/src/include/nodes/execnodes.h b/src/include/nodes/execnodes.h
index b72decc..b0a86c5 100644
--- a/src/include/nodes/execnodes.h
+++ b/src/include/nodes/execnodes.h
@@ -1177,6 +1177,8 @@ typedef struct AppendState
 {
 	PlanState	ps;				/* its first field is NodeTag */
 	PlanState **appendplans;	/* array of PlanStates for my inputs */
+	bool		as_async;		/* true to allow async execution */
+	bool	   *as_finished;	/* array of the running state of subplans */
 	int			as_nplans;
 	int			as_whichplan;
 } AppendState;
-- 
1.8.3.1


----Next_Part(Thu_Jul_21_18_50_07_2016_597)--
Content-Type: text/plain
Content-Disposition: inline
Content-Transfer-Encoding: 8bit
MIME-Version: 1.0


-- 
Sent via pgsql-hackers mailing list ([email protected])
To make changes to your subscription:
http://www.postgresql.org/mailpref/pgsql-hackers

----Next_Part(Thu_Jul_21_18_50_07_2016_597)----






^ permalink  raw  reply  [nested|flat] 56+ messages in thread

* [PATCH 7/7] Make Append node async-aware.
@ 2016-06-28 09:52 Kyotaro Horiguchi <[email protected]>
  0 siblings, 0 replies; 56+ messages in thread

From: Kyotaro Horiguchi @ 2016-06-28 09:52 UTC (permalink / raw)

Change append node to be capable to handle asynchronous children
properly. As soon as it receives !async_ready from a child, it moves
to the next child and if no child is ready, it sleeps until at least
one of them become ready.
---
 src/backend/executor/nodeAppend.c | 67 +++++++++++++++++++++++++++++++++++++--
 src/include/nodes/execnodes.h     |  2 ++
 2 files changed, 67 insertions(+), 2 deletions(-)

diff --git a/src/backend/executor/nodeAppend.c b/src/backend/executor/nodeAppend.c
index e0ce8c6..9a4063a 100644
--- a/src/backend/executor/nodeAppend.c
+++ b/src/backend/executor/nodeAppend.c
@@ -58,6 +58,7 @@
 #include "postgres.h"
 
 #include "executor/execdebug.h"
+#include "executor/execAsync.h"
 #include "executor/nodeAppend.h"
 
 static bool exec_append_initialize_next(AppendState *appendstate);
@@ -121,6 +122,7 @@ ExecInitAppend(Append *node, EState *estate, int eflags)
 {
 	AppendState *appendstate = makeNode(AppendState);
 	PlanState **appendplanstates;
+	bool	   *finished;
 	int			nplans;
 	int			i;
 	ListCell   *lc;
@@ -134,14 +136,17 @@ ExecInitAppend(Append *node, EState *estate, int eflags)
 	nplans = list_length(node->appendplans);
 
 	appendplanstates = (PlanState **) palloc0(nplans * sizeof(PlanState *));
-
+	finished = (bool *) palloc0(nplans * sizeof(bool));
+	
 	/*
 	 * create new AppendState for our append node
 	 */
 	appendstate->ps.plan = (Plan *) node;
 	appendstate->ps.state = estate;
 	appendstate->appendplans = appendplanstates;
+	appendstate->as_finished = finished;
 	appendstate->as_nplans = nplans;
+	appendstate->as_async = ((eflags & EXEC_FLAG_BACKWARD) == 0);
 
 	/*
 	 * Miscellaneous initialization
@@ -194,6 +199,8 @@ ExecInitAppend(Append *node, EState *estate, int eflags)
 void
 ExecAppend(AppendState *node)
 {
+	int stopplan = node->as_whichplan;
+
 	for (;;)
 	{
 		PlanState  *subnode;
@@ -207,7 +214,36 @@ ExecAppend(AppendState *node)
 		/*
 		 * get a tuple from the subplan
 		 */
-		result = ExecProcNode(subnode);
+		Assert(!node->as_finished[node->as_whichplan]);
+
+		if (!subnode->result_ready)
+			ExecExecuteNode(subnode);
+
+		if (!subnode->result_ready)
+		{
+			if (node->as_async)
+			{
+				/* Move to the next living node */
+				do
+				{
+					node->as_whichplan = 
+						(node->as_whichplan + 1) %  node->as_nplans;
+				} while (node->as_whichplan != stopplan &&
+						 node->as_finished[node->as_whichplan]);
+
+				/* No node is ready yet, return as not-ready */
+				if (node->as_whichplan == stopplan)
+					return;
+
+				/* Try the next node */
+				continue;
+			}
+
+			/* If not async, immediately wait for this subnode */
+			ExecAsyncWaitForNode(subnode);
+		}				
+
+		result = ExecConsumeResult((PlanState *) subnode);
 
 		if (!TupIsNull(result))
 		{
@@ -220,6 +256,31 @@ ExecAppend(AppendState *node)
 			return;
 		}
 
+		if (node->as_async)
+		{
+			node->as_finished[node->as_whichplan] = true;
+			stopplan = node->as_whichplan;
+
+			/* Find the next living subnode */
+			do
+			{
+				node->as_whichplan =
+					(node->as_whichplan + 1) % node->as_nplans;
+			} while (node->as_whichplan != stopplan &&
+					 node->as_finished[node->as_whichplan]);
+
+			if (node->as_whichplan != stopplan)
+			{
+				stopplan = node->as_whichplan;
+				continue;
+			}
+
+			/* All subnodes are exhausted. Finish this node. */
+			ExecReturnTuple(&node->ps,
+							ExecClearTuple(node->ps.ps_ResultTupleSlot));
+			return;
+		}
+
 		/*
 		 * Go on to the "next" subplan in the appropriate direction. If no
 		 * more subplans, return the empty slot set up for us by
@@ -277,6 +338,8 @@ ExecReScanAppend(AppendState *node)
 	{
 		PlanState  *subnode = node->appendplans[i];
 
+		node->as_finished[i] = false;
+
 		/*
 		 * ExecReScan doesn't know about my subplans, so I have to do
 		 * changed-parameter signaling myself.
diff --git a/src/include/nodes/execnodes.h b/src/include/nodes/execnodes.h
index b72decc..b0a86c5 100644
--- a/src/include/nodes/execnodes.h
+++ b/src/include/nodes/execnodes.h
@@ -1177,6 +1177,8 @@ typedef struct AppendState
 {
 	PlanState	ps;				/* its first field is NodeTag */
 	PlanState **appendplans;	/* array of PlanStates for my inputs */
+	bool		as_async;		/* true to allow async execution */
+	bool	   *as_finished;	/* array of the running state of subplans */
 	int			as_nplans;
 	int			as_whichplan;
 } AppendState;
-- 
1.8.3.1


----Next_Part(Thu_Jul_21_18_50_07_2016_597)--
Content-Type: text/plain
Content-Disposition: inline
Content-Transfer-Encoding: 8bit
MIME-Version: 1.0


-- 
Sent via pgsql-hackers mailing list ([email protected])
To make changes to your subscription:
http://www.postgresql.org/mailpref/pgsql-hackers

----Next_Part(Thu_Jul_21_18_50_07_2016_597)----






^ permalink  raw  reply  [nested|flat] 56+ messages in thread

* [PATCH 7/7] Make Append node async-aware.
@ 2016-06-28 09:52 Kyotaro Horiguchi <[email protected]>
  0 siblings, 0 replies; 56+ messages in thread

From: Kyotaro Horiguchi @ 2016-06-28 09:52 UTC (permalink / raw)

Change append node to be capable to handle asynchronous children
properly. As soon as it receives !async_ready from a child, it moves
to the next child and if no child is ready, it sleeps until at least
one of them become ready.
---
 src/backend/executor/nodeAppend.c | 94 ++++++++++++++++++++++++++++++++-------
 src/include/nodes/execnodes.h     |  2 +
 2 files changed, 80 insertions(+), 16 deletions(-)

diff --git a/src/backend/executor/nodeAppend.c b/src/backend/executor/nodeAppend.c
index e0ce8c6..1c0d26e 100644
--- a/src/backend/executor/nodeAppend.c
+++ b/src/backend/executor/nodeAppend.c
@@ -58,6 +58,7 @@
 #include "postgres.h"
 
 #include "executor/execdebug.h"
+#include "executor/execAsync.h"
 #include "executor/nodeAppend.h"
 
 static bool exec_append_initialize_next(AppendState *appendstate);
@@ -121,6 +122,7 @@ ExecInitAppend(Append *node, EState *estate, int eflags)
 {
 	AppendState *appendstate = makeNode(AppendState);
 	PlanState **appendplanstates;
+	bool	   *finished;
 	int			nplans;
 	int			i;
 	ListCell   *lc;
@@ -134,14 +136,17 @@ ExecInitAppend(Append *node, EState *estate, int eflags)
 	nplans = list_length(node->appendplans);
 
 	appendplanstates = (PlanState **) palloc0(nplans * sizeof(PlanState *));
-
+	finished = (bool *) palloc0(nplans * sizeof(bool));
+	
 	/*
 	 * create new AppendState for our append node
 	 */
 	appendstate->ps.plan = (Plan *) node;
 	appendstate->ps.state = estate;
 	appendstate->appendplans = appendplanstates;
+	appendstate->as_finished = finished;
 	appendstate->as_nplans = nplans;
+	appendstate->as_async = ((eflags & EXEC_FLAG_BACKWARD) == 0);
 
 	/*
 	 * Miscellaneous initialization
@@ -194,49 +199,104 @@ ExecInitAppend(Append *node, EState *estate, int eflags)
 void
 ExecAppend(AppendState *node)
 {
-	for (;;)
+	int n_notready = 0;
+	PlanState  *subnode;
+	int i, n;
+
+	n = node->as_whichplan;
+
+	for (i = 0 ; i < node->as_nplans ; i++)
 	{
-		PlanState  *subnode;
 		TupleTableSlot *result;
 
+		if (node->as_async)
+		{
+			if (n >= node->as_nplans)
+				n = 0;
+
+			if (node->as_finished[n])
+			{
+				n++;
+				continue;
+			}
+		}
+
 		/*
 		 * figure out which subplan we are currently processing
 		 */
-		subnode = node->appendplans[node->as_whichplan];
+		subnode = node->appendplans[n];
 
 		/*
-		 * get a tuple from the subplan
+		 * execute the subplan to get a result if it is not ready yet
 		 */
-		result = ExecProcNode(subnode);
+		if (!subnode->result_ready)
+			ExecExecuteNode(subnode);
+
+		/*
+		 * This subnode is not ready yet when asynchrony is not allowed,
+		 * immediately wait for this subnode.
+		 */
+		if (!subnode->result_ready)
+		{
+			if (node->as_async)
+			{
+				n_notready++;
+				n++;
+				continue;
+			}
+			ExecAsyncWaitForNode(subnode);
+		}
+
+		Assert(subnode->result_ready);
+
+		result = ExecConsumeResult((PlanState *)subnode);
 
 		if (!TupIsNull(result))
 		{
 			/*
-			 * If the subplan gave us something then return it as-is. We do
-			 * NOT make use of the result slot that was set up in
-			 * ExecInitAppend; there's no need for it.
+			 * If the subplan gave us something then return it
+			 * as-is. We do NOT make use of the result slot that was
+			 * set up in ExecInitAppend; there's no need for it.
 			 */
+			node->as_whichplan = n;
 			ExecReturnTuple(&node->ps, result);
 			return;
 		}
 
+		/* Tuple has been exhausted on this subnode */
+		if (node->as_async)
+		{
+			node->as_finished[n++] = true;
+			continue;
+		}
+
 		/*
 		 * Go on to the "next" subplan in the appropriate direction. If no
 		 * more subplans, return the empty slot set up for us by
 		 * ExecInitAppend.
 		 */
 		if (ScanDirectionIsForward(node->ps.state->es_direction))
-			node->as_whichplan++;
+		{
+			if (n >= node->as_nplans - 1)
+				break;
+			n++;
+		}
 		else
-			node->as_whichplan--;
-		if (!exec_append_initialize_next(node))
 		{
-			ExecReturnTuple(&node->ps,
-							ExecClearTuple(node->ps.ps_ResultTupleSlot));
-			return;
+			if (n == 0)
+				break;
+			n--;
 		}
+	}
 
-		/* Else loop back and try to get a tuple from the new subplan */
+	/*
+	 * We are finished if reached here and no subnodes are not-ready
+	 */
+	if (n_notready == 0)
+	{
+		node->as_whichplan = n;
+		ExecReturnTuple(&node->ps,
+						ExecClearTuple(node->ps.ps_ResultTupleSlot));
 	}
 }
 
@@ -277,6 +337,8 @@ ExecReScanAppend(AppendState *node)
 	{
 		PlanState  *subnode = node->appendplans[i];
 
+		node->as_finished[i] = false;
+
 		/*
 		 * ExecReScan doesn't know about my subplans, so I have to do
 		 * changed-parameter signaling myself.
diff --git a/src/include/nodes/execnodes.h b/src/include/nodes/execnodes.h
index 9121537..e4f2bb6 100644
--- a/src/include/nodes/execnodes.h
+++ b/src/include/nodes/execnodes.h
@@ -1170,6 +1170,8 @@ typedef struct AppendState
 {
 	PlanState	ps;				/* its first field is NodeTag */
 	PlanState **appendplans;	/* array of PlanStates for my inputs */
+	bool		as_async;		/* true to allow async execution */
+	bool	   *as_finished;	/* array of the running state of subplans */
 	int			as_nplans;
 	int			as_whichplan;
 } AppendState;
-- 
1.8.3.1


----Next_Part(Wed_Jul_06_16_29_14_2016_599)--
Content-Type: Text/Plain; charset=us-ascii
Content-Transfer-Encoding: 7bit
Content-Disposition: inline; filename="gentbl.sql"

CREATE EXTENSION postgres_fdw;
CREATE TABLE pl (a int, b int);
CREATE TABLE cl1 (LIKE pl) INHERITS (pl);
CREATE TABLE cl2 (LIKE pl) INHERITS (pl);
CREATE TABLE cl3 (LIKE pl) INHERITS (pl);
CREATE TABLE cl4 (LIKE pl) INHERITS (pl);
INSERT INTO cl1 (SELECT a, a FROM generate_series(0000000, 0999999) a);
INSERT INTO cl2 (SELECT a, a FROM generate_series(1000000, 1999999) a);
INSERT INTO cl3 (SELECT a, a FROM generate_series(2000000, 2999999) a);
INSERT INTO cl4 (SELECT a, a FROM generate_series(3000000, 3999999) a);

CREATE TABLE t0 (LIKE pl);
INSERT INTO t0 (SELECT a, a FROM generate_series(0000000, 9999999) a);

CREATE SERVER sv0 FOREIGN DATA WRAPPER postgres_fdw OPTIONS (host '/tmp', dbname 'postgres');
CREATE SERVER sv1 FOREIGN DATA WRAPPER postgres_fdw OPTIONS (host '/tmp', dbname 'postgres');
CREATE SERVER sv2 FOREIGN DATA WRAPPER postgres_fdw OPTIONS (host '/tmp', dbname 'postgres');
CREATE SERVER sv3 FOREIGN DATA WRAPPER postgres_fdw OPTIONS (host '/tmp', dbname 'postgres');
CREATE SERVER sv4 FOREIGN DATA WRAPPER postgres_fdw OPTIONS (host '/tmp', dbname 'postgres');
CREATE USER MAPPING FOR public SERVER sv0;
CREATE USER MAPPING FOR public SERVER sv1;
CREATE USER MAPPING FOR public SERVER sv2;
CREATE USER MAPPING FOR public SERVER sv3;
CREATE USER MAPPING FOR public SERVER sv4;

CREATE FOREIGN TABLE ft10 (a int, b int) SERVER sv0 OPTIONS (table_name 'cl1');
CREATE FOREIGN TABLE ft20 (a int, b int) SERVER sv0 OPTIONS (table_name 'cl2');
CREATE FOREIGN TABLE ft30 (a int, b int) SERVER sv0 OPTIONS (table_name 'cl3');
CREATE FOREIGN TABLE ft40 (a int, b int) SERVER sv0 OPTIONS (table_name 'cl4');
CREATE FOREIGN TABLE ft11 (a int, b int) SERVER sv1 OPTIONS (table_name 'cl1');
CREATE FOREIGN TABLE ft22 (a int, b int) SERVER sv2 OPTIONS (table_name 'cl2');
CREATE FOREIGN TABLE ft33 (a int, b int) SERVER sv3 OPTIONS (table_name 'cl3');
CREATE FOREIGN TABLE ft44 (a int, b int) SERVER sv4 OPTIONS (table_name 'cl4');

CREATE TABLE pf0 (LIKE pl);
ALTER FOREIGN TABLE ft10 INHERIT pf0;
ALTER FOREIGN TABLE ft20 INHERIT pf0;
ALTER FOREIGN TABLE ft30 INHERIT pf0;
ALTER FOREIGN TABLE ft40 INHERIT pf0;

CREATE table pf1 (LIKE pl);
ALTER FOREIGN TABLE ft11 INHERIT pf1;
ALTER FOREIGN TABLE ft22 INHERIT pf1;
ALTER FOREIGN TABLE ft33 INHERIT pf1;
ALTER FOREIGN TABLE ft44 INHERIT pf1;

ANALYZE;

----Next_Part(Wed_Jul_06_16_29_14_2016_599)--
Content-Type: Text/Plain; charset=us-ascii
Content-Transfer-Encoding: 7bit
Content-Disposition: inline; filename="testrun.sh"

#! /bin/bash

function do_test() {
	echo $1
	for i in $(seq 1 10); do psql postgres -c "set log_min_duration_statement = 0; set client_min_messages=log; select sum(a) from $1"; done | grep LOG
}

for i in "t0" "pl" "pf0" "pf1";
do
  do_test $i
done


----Next_Part(Wed_Jul_06_16_29_14_2016_599)--
Content-Type: text/plain
Content-Disposition: inline
Content-Transfer-Encoding: 8bit
MIME-Version: 1.0


-- 
Sent via pgsql-hackers mailing list ([email protected])
To make changes to your subscription:
http://www.postgresql.org/mailpref/pgsql-hackers

----Next_Part(Wed_Jul_06_16_29_14_2016_599)----






^ permalink  raw  reply  [nested|flat] 56+ messages in thread

* [PATCH 7/7] Make Append node async-aware.
@ 2016-06-28 09:52 Kyotaro Horiguchi <[email protected]>
  0 siblings, 0 replies; 56+ messages in thread

From: Kyotaro Horiguchi @ 2016-06-28 09:52 UTC (permalink / raw)

Change append node to be capable to handle asynchronous children
properly. As soon as it receives !async_ready from a child, it moves
to the next child and if no child is ready, it sleeps until at least
one of them become ready.
---
 src/backend/executor/nodeAppend.c | 67 +++++++++++++++++++++++++++++++++++++--
 src/include/nodes/execnodes.h     |  2 ++
 2 files changed, 67 insertions(+), 2 deletions(-)

diff --git a/src/backend/executor/nodeAppend.c b/src/backend/executor/nodeAppend.c
index e0ce8c6..9a4063a 100644
--- a/src/backend/executor/nodeAppend.c
+++ b/src/backend/executor/nodeAppend.c
@@ -58,6 +58,7 @@
 #include "postgres.h"
 
 #include "executor/execdebug.h"
+#include "executor/execAsync.h"
 #include "executor/nodeAppend.h"
 
 static bool exec_append_initialize_next(AppendState *appendstate);
@@ -121,6 +122,7 @@ ExecInitAppend(Append *node, EState *estate, int eflags)
 {
 	AppendState *appendstate = makeNode(AppendState);
 	PlanState **appendplanstates;
+	bool	   *finished;
 	int			nplans;
 	int			i;
 	ListCell   *lc;
@@ -134,14 +136,17 @@ ExecInitAppend(Append *node, EState *estate, int eflags)
 	nplans = list_length(node->appendplans);
 
 	appendplanstates = (PlanState **) palloc0(nplans * sizeof(PlanState *));
-
+	finished = (bool *) palloc0(nplans * sizeof(bool));
+	
 	/*
 	 * create new AppendState for our append node
 	 */
 	appendstate->ps.plan = (Plan *) node;
 	appendstate->ps.state = estate;
 	appendstate->appendplans = appendplanstates;
+	appendstate->as_finished = finished;
 	appendstate->as_nplans = nplans;
+	appendstate->as_async = ((eflags & EXEC_FLAG_BACKWARD) == 0);
 
 	/*
 	 * Miscellaneous initialization
@@ -194,6 +199,8 @@ ExecInitAppend(Append *node, EState *estate, int eflags)
 void
 ExecAppend(AppendState *node)
 {
+	int stopplan = node->as_whichplan;
+
 	for (;;)
 	{
 		PlanState  *subnode;
@@ -207,7 +214,36 @@ ExecAppend(AppendState *node)
 		/*
 		 * get a tuple from the subplan
 		 */
-		result = ExecProcNode(subnode);
+		Assert(!node->as_finished[node->as_whichplan]);
+
+		if (!subnode->result_ready)
+			ExecExecuteNode(subnode);
+
+		if (!subnode->result_ready)
+		{
+			if (node->as_async)
+			{
+				/* Move to the next living node */
+				do
+				{
+					node->as_whichplan = 
+						(node->as_whichplan + 1) %  node->as_nplans;
+				} while (node->as_whichplan != stopplan &&
+						 node->as_finished[node->as_whichplan]);
+
+				/* No node is ready yet, return as not-ready */
+				if (node->as_whichplan == stopplan)
+					return;
+
+				/* Try the next node */
+				continue;
+			}
+
+			/* If not async, immediately wait for this subnode */
+			ExecAsyncWaitForNode(subnode);
+		}				
+
+		result = ExecConsumeResult((PlanState *) subnode);
 
 		if (!TupIsNull(result))
 		{
@@ -220,6 +256,31 @@ ExecAppend(AppendState *node)
 			return;
 		}
 
+		if (node->as_async)
+		{
+			node->as_finished[node->as_whichplan] = true;
+			stopplan = node->as_whichplan;
+
+			/* Find the next living subnode */
+			do
+			{
+				node->as_whichplan =
+					(node->as_whichplan + 1) % node->as_nplans;
+			} while (node->as_whichplan != stopplan &&
+					 node->as_finished[node->as_whichplan]);
+
+			if (node->as_whichplan != stopplan)
+			{
+				stopplan = node->as_whichplan;
+				continue;
+			}
+
+			/* All subnodes are exhausted. Finish this node. */
+			ExecReturnTuple(&node->ps,
+							ExecClearTuple(node->ps.ps_ResultTupleSlot));
+			return;
+		}
+
 		/*
 		 * Go on to the "next" subplan in the appropriate direction. If no
 		 * more subplans, return the empty slot set up for us by
@@ -277,6 +338,8 @@ ExecReScanAppend(AppendState *node)
 	{
 		PlanState  *subnode = node->appendplans[i];
 
+		node->as_finished[i] = false;
+
 		/*
 		 * ExecReScan doesn't know about my subplans, so I have to do
 		 * changed-parameter signaling myself.
diff --git a/src/include/nodes/execnodes.h b/src/include/nodes/execnodes.h
index b72decc..b0a86c5 100644
--- a/src/include/nodes/execnodes.h
+++ b/src/include/nodes/execnodes.h
@@ -1177,6 +1177,8 @@ typedef struct AppendState
 {
 	PlanState	ps;				/* its first field is NodeTag */
 	PlanState **appendplans;	/* array of PlanStates for my inputs */
+	bool		as_async;		/* true to allow async execution */
+	bool	   *as_finished;	/* array of the running state of subplans */
 	int			as_nplans;
 	int			as_whichplan;
 } AppendState;
-- 
2.9.2


----Next_Part(Mon_Aug_29_17_08_36_2016_213)--
Content-Type: text/plain
Content-Disposition: inline
Content-Transfer-Encoding: 8bit
MIME-Version: 1.0


-- 
Sent via pgsql-hackers mailing list ([email protected])
To make changes to your subscription:
http://www.postgresql.org/mailpref/pgsql-hackers

----Next_Part(Mon_Aug_29_17_08_36_2016_213)----






^ permalink  raw  reply  [nested|flat] 56+ messages in thread

* [PATCH 7/7] Make Append node async-aware.
@ 2016-06-28 09:52 Kyotaro Horiguchi <[email protected]>
  0 siblings, 0 replies; 56+ messages in thread

From: Kyotaro Horiguchi @ 2016-06-28 09:52 UTC (permalink / raw)

Change append node to be capable to handle asynchronous children
properly. As soon as it receives !async_ready from a child, it moves
to the next child and if no child is ready, it sleeps until at least
one of them become ready.
---
 src/backend/executor/nodeAppend.c | 67 +++++++++++++++++++++++++++++++++++++--
 src/include/nodes/execnodes.h     |  2 ++
 2 files changed, 67 insertions(+), 2 deletions(-)

diff --git a/src/backend/executor/nodeAppend.c b/src/backend/executor/nodeAppend.c
index e0ce8c6..9a4063a 100644
--- a/src/backend/executor/nodeAppend.c
+++ b/src/backend/executor/nodeAppend.c
@@ -58,6 +58,7 @@
 #include "postgres.h"
 
 #include "executor/execdebug.h"
+#include "executor/execAsync.h"
 #include "executor/nodeAppend.h"
 
 static bool exec_append_initialize_next(AppendState *appendstate);
@@ -121,6 +122,7 @@ ExecInitAppend(Append *node, EState *estate, int eflags)
 {
 	AppendState *appendstate = makeNode(AppendState);
 	PlanState **appendplanstates;
+	bool	   *finished;
 	int			nplans;
 	int			i;
 	ListCell   *lc;
@@ -134,14 +136,17 @@ ExecInitAppend(Append *node, EState *estate, int eflags)
 	nplans = list_length(node->appendplans);
 
 	appendplanstates = (PlanState **) palloc0(nplans * sizeof(PlanState *));
-
+	finished = (bool *) palloc0(nplans * sizeof(bool));
+	
 	/*
 	 * create new AppendState for our append node
 	 */
 	appendstate->ps.plan = (Plan *) node;
 	appendstate->ps.state = estate;
 	appendstate->appendplans = appendplanstates;
+	appendstate->as_finished = finished;
 	appendstate->as_nplans = nplans;
+	appendstate->as_async = ((eflags & EXEC_FLAG_BACKWARD) == 0);
 
 	/*
 	 * Miscellaneous initialization
@@ -194,6 +199,8 @@ ExecInitAppend(Append *node, EState *estate, int eflags)
 void
 ExecAppend(AppendState *node)
 {
+	int stopplan = node->as_whichplan;
+
 	for (;;)
 	{
 		PlanState  *subnode;
@@ -207,7 +214,36 @@ ExecAppend(AppendState *node)
 		/*
 		 * get a tuple from the subplan
 		 */
-		result = ExecProcNode(subnode);
+		Assert(!node->as_finished[node->as_whichplan]);
+
+		if (!subnode->result_ready)
+			ExecExecuteNode(subnode);
+
+		if (!subnode->result_ready)
+		{
+			if (node->as_async)
+			{
+				/* Move to the next living node */
+				do
+				{
+					node->as_whichplan = 
+						(node->as_whichplan + 1) %  node->as_nplans;
+				} while (node->as_whichplan != stopplan &&
+						 node->as_finished[node->as_whichplan]);
+
+				/* No node is ready yet, return as not-ready */
+				if (node->as_whichplan == stopplan)
+					return;
+
+				/* Try the next node */
+				continue;
+			}
+
+			/* If not async, immediately wait for this subnode */
+			ExecAsyncWaitForNode(subnode);
+		}				
+
+		result = ExecConsumeResult((PlanState *) subnode);
 
 		if (!TupIsNull(result))
 		{
@@ -220,6 +256,31 @@ ExecAppend(AppendState *node)
 			return;
 		}
 
+		if (node->as_async)
+		{
+			node->as_finished[node->as_whichplan] = true;
+			stopplan = node->as_whichplan;
+
+			/* Find the next living subnode */
+			do
+			{
+				node->as_whichplan =
+					(node->as_whichplan + 1) % node->as_nplans;
+			} while (node->as_whichplan != stopplan &&
+					 node->as_finished[node->as_whichplan]);
+
+			if (node->as_whichplan != stopplan)
+			{
+				stopplan = node->as_whichplan;
+				continue;
+			}
+
+			/* All subnodes are exhausted. Finish this node. */
+			ExecReturnTuple(&node->ps,
+							ExecClearTuple(node->ps.ps_ResultTupleSlot));
+			return;
+		}
+
 		/*
 		 * Go on to the "next" subplan in the appropriate direction. If no
 		 * more subplans, return the empty slot set up for us by
@@ -277,6 +338,8 @@ ExecReScanAppend(AppendState *node)
 	{
 		PlanState  *subnode = node->appendplans[i];
 
+		node->as_finished[i] = false;
+
 		/*
 		 * ExecReScan doesn't know about my subplans, so I have to do
 		 * changed-parameter signaling myself.
diff --git a/src/include/nodes/execnodes.h b/src/include/nodes/execnodes.h
index b72decc..b0a86c5 100644
--- a/src/include/nodes/execnodes.h
+++ b/src/include/nodes/execnodes.h
@@ -1177,6 +1177,8 @@ typedef struct AppendState
 {
 	PlanState	ps;				/* its first field is NodeTag */
 	PlanState **appendplans;	/* array of PlanStates for my inputs */
+	bool		as_async;		/* true to allow async execution */
+	bool	   *as_finished;	/* array of the running state of subplans */
 	int			as_nplans;
 	int			as_whichplan;
 } AppendState;
-- 
2.9.2


----Next_Part(Mon_Aug_29_17_08_36_2016_213)--
Content-Type: text/plain
Content-Disposition: inline
Content-Transfer-Encoding: 8bit
MIME-Version: 1.0


-- 
Sent via pgsql-hackers mailing list ([email protected])
To make changes to your subscription:
http://www.postgresql.org/mailpref/pgsql-hackers

----Next_Part(Mon_Aug_29_17_08_36_2016_213)----






^ permalink  raw  reply  [nested|flat] 56+ messages in thread

* [PATCH 7/7] Make Append node async-aware.
@ 2016-06-28 09:52 Kyotaro Horiguchi <[email protected]>
  0 siblings, 0 replies; 56+ messages in thread

From: Kyotaro Horiguchi @ 2016-06-28 09:52 UTC (permalink / raw)

Change append node to be capable to handle asynchronous children
properly. As soon as it receives !async_ready from a child, it moves
to the next child and if no child is ready, it sleeps until at least
one of them become ready.
---
 src/backend/executor/nodeAppend.c | 94 ++++++++++++++++++++++++++++++++-------
 src/include/nodes/execnodes.h     |  2 +
 2 files changed, 80 insertions(+), 16 deletions(-)

diff --git a/src/backend/executor/nodeAppend.c b/src/backend/executor/nodeAppend.c
index e0ce8c6..1c0d26e 100644
--- a/src/backend/executor/nodeAppend.c
+++ b/src/backend/executor/nodeAppend.c
@@ -58,6 +58,7 @@
 #include "postgres.h"
 
 #include "executor/execdebug.h"
+#include "executor/execAsync.h"
 #include "executor/nodeAppend.h"
 
 static bool exec_append_initialize_next(AppendState *appendstate);
@@ -121,6 +122,7 @@ ExecInitAppend(Append *node, EState *estate, int eflags)
 {
 	AppendState *appendstate = makeNode(AppendState);
 	PlanState **appendplanstates;
+	bool	   *finished;
 	int			nplans;
 	int			i;
 	ListCell   *lc;
@@ -134,14 +136,17 @@ ExecInitAppend(Append *node, EState *estate, int eflags)
 	nplans = list_length(node->appendplans);
 
 	appendplanstates = (PlanState **) palloc0(nplans * sizeof(PlanState *));
-
+	finished = (bool *) palloc0(nplans * sizeof(bool));
+	
 	/*
 	 * create new AppendState for our append node
 	 */
 	appendstate->ps.plan = (Plan *) node;
 	appendstate->ps.state = estate;
 	appendstate->appendplans = appendplanstates;
+	appendstate->as_finished = finished;
 	appendstate->as_nplans = nplans;
+	appendstate->as_async = ((eflags & EXEC_FLAG_BACKWARD) == 0);
 
 	/*
 	 * Miscellaneous initialization
@@ -194,49 +199,104 @@ ExecInitAppend(Append *node, EState *estate, int eflags)
 void
 ExecAppend(AppendState *node)
 {
-	for (;;)
+	int n_notready = 0;
+	PlanState  *subnode;
+	int i, n;
+
+	n = node->as_whichplan;
+
+	for (i = 0 ; i < node->as_nplans ; i++)
 	{
-		PlanState  *subnode;
 		TupleTableSlot *result;
 
+		if (node->as_async)
+		{
+			if (n >= node->as_nplans)
+				n = 0;
+
+			if (node->as_finished[n])
+			{
+				n++;
+				continue;
+			}
+		}
+
 		/*
 		 * figure out which subplan we are currently processing
 		 */
-		subnode = node->appendplans[node->as_whichplan];
+		subnode = node->appendplans[n];
 
 		/*
-		 * get a tuple from the subplan
+		 * execute the subplan to get a result if it is not ready yet
 		 */
-		result = ExecProcNode(subnode);
+		if (!subnode->result_ready)
+			ExecExecuteNode(subnode);
+
+		/*
+		 * This subnode is not ready yet when asynchrony is not allowed,
+		 * immediately wait for this subnode.
+		 */
+		if (!subnode->result_ready)
+		{
+			if (node->as_async)
+			{
+				n_notready++;
+				n++;
+				continue;
+			}
+			ExecAsyncWaitForNode(subnode);
+		}
+
+		Assert(subnode->result_ready);
+
+		result = ExecConsumeResult((PlanState *)subnode);
 
 		if (!TupIsNull(result))
 		{
 			/*
-			 * If the subplan gave us something then return it as-is. We do
-			 * NOT make use of the result slot that was set up in
-			 * ExecInitAppend; there's no need for it.
+			 * If the subplan gave us something then return it
+			 * as-is. We do NOT make use of the result slot that was
+			 * set up in ExecInitAppend; there's no need for it.
 			 */
+			node->as_whichplan = n;
 			ExecReturnTuple(&node->ps, result);
 			return;
 		}
 
+		/* Tuple has been exhausted on this subnode */
+		if (node->as_async)
+		{
+			node->as_finished[n++] = true;
+			continue;
+		}
+
 		/*
 		 * Go on to the "next" subplan in the appropriate direction. If no
 		 * more subplans, return the empty slot set up for us by
 		 * ExecInitAppend.
 		 */
 		if (ScanDirectionIsForward(node->ps.state->es_direction))
-			node->as_whichplan++;
+		{
+			if (n >= node->as_nplans - 1)
+				break;
+			n++;
+		}
 		else
-			node->as_whichplan--;
-		if (!exec_append_initialize_next(node))
 		{
-			ExecReturnTuple(&node->ps,
-							ExecClearTuple(node->ps.ps_ResultTupleSlot));
-			return;
+			if (n == 0)
+				break;
+			n--;
 		}
+	}
 
-		/* Else loop back and try to get a tuple from the new subplan */
+	/*
+	 * We are finished if reached here and no subnodes are not-ready
+	 */
+	if (n_notready == 0)
+	{
+		node->as_whichplan = n;
+		ExecReturnTuple(&node->ps,
+						ExecClearTuple(node->ps.ps_ResultTupleSlot));
 	}
 }
 
@@ -277,6 +337,8 @@ ExecReScanAppend(AppendState *node)
 	{
 		PlanState  *subnode = node->appendplans[i];
 
+		node->as_finished[i] = false;
+
 		/*
 		 * ExecReScan doesn't know about my subplans, so I have to do
 		 * changed-parameter signaling myself.
diff --git a/src/include/nodes/execnodes.h b/src/include/nodes/execnodes.h
index 9121537..e4f2bb6 100644
--- a/src/include/nodes/execnodes.h
+++ b/src/include/nodes/execnodes.h
@@ -1170,6 +1170,8 @@ typedef struct AppendState
 {
 	PlanState	ps;				/* its first field is NodeTag */
 	PlanState **appendplans;	/* array of PlanStates for my inputs */
+	bool		as_async;		/* true to allow async execution */
+	bool	   *as_finished;	/* array of the running state of subplans */
 	int			as_nplans;
 	int			as_whichplan;
 } AppendState;
-- 
1.8.3.1


----Next_Part(Wed_Jul_06_16_29_14_2016_599)--
Content-Type: Text/Plain; charset=us-ascii
Content-Transfer-Encoding: 7bit
Content-Disposition: inline; filename="gentbl.sql"

CREATE EXTENSION postgres_fdw;
CREATE TABLE pl (a int, b int);
CREATE TABLE cl1 (LIKE pl) INHERITS (pl);
CREATE TABLE cl2 (LIKE pl) INHERITS (pl);
CREATE TABLE cl3 (LIKE pl) INHERITS (pl);
CREATE TABLE cl4 (LIKE pl) INHERITS (pl);
INSERT INTO cl1 (SELECT a, a FROM generate_series(0000000, 0999999) a);
INSERT INTO cl2 (SELECT a, a FROM generate_series(1000000, 1999999) a);
INSERT INTO cl3 (SELECT a, a FROM generate_series(2000000, 2999999) a);
INSERT INTO cl4 (SELECT a, a FROM generate_series(3000000, 3999999) a);

CREATE TABLE t0 (LIKE pl);
INSERT INTO t0 (SELECT a, a FROM generate_series(0000000, 9999999) a);

CREATE SERVER sv0 FOREIGN DATA WRAPPER postgres_fdw OPTIONS (host '/tmp', dbname 'postgres');
CREATE SERVER sv1 FOREIGN DATA WRAPPER postgres_fdw OPTIONS (host '/tmp', dbname 'postgres');
CREATE SERVER sv2 FOREIGN DATA WRAPPER postgres_fdw OPTIONS (host '/tmp', dbname 'postgres');
CREATE SERVER sv3 FOREIGN DATA WRAPPER postgres_fdw OPTIONS (host '/tmp', dbname 'postgres');
CREATE SERVER sv4 FOREIGN DATA WRAPPER postgres_fdw OPTIONS (host '/tmp', dbname 'postgres');
CREATE USER MAPPING FOR public SERVER sv0;
CREATE USER MAPPING FOR public SERVER sv1;
CREATE USER MAPPING FOR public SERVER sv2;
CREATE USER MAPPING FOR public SERVER sv3;
CREATE USER MAPPING FOR public SERVER sv4;

CREATE FOREIGN TABLE ft10 (a int, b int) SERVER sv0 OPTIONS (table_name 'cl1');
CREATE FOREIGN TABLE ft20 (a int, b int) SERVER sv0 OPTIONS (table_name 'cl2');
CREATE FOREIGN TABLE ft30 (a int, b int) SERVER sv0 OPTIONS (table_name 'cl3');
CREATE FOREIGN TABLE ft40 (a int, b int) SERVER sv0 OPTIONS (table_name 'cl4');
CREATE FOREIGN TABLE ft11 (a int, b int) SERVER sv1 OPTIONS (table_name 'cl1');
CREATE FOREIGN TABLE ft22 (a int, b int) SERVER sv2 OPTIONS (table_name 'cl2');
CREATE FOREIGN TABLE ft33 (a int, b int) SERVER sv3 OPTIONS (table_name 'cl3');
CREATE FOREIGN TABLE ft44 (a int, b int) SERVER sv4 OPTIONS (table_name 'cl4');

CREATE TABLE pf0 (LIKE pl);
ALTER FOREIGN TABLE ft10 INHERIT pf0;
ALTER FOREIGN TABLE ft20 INHERIT pf0;
ALTER FOREIGN TABLE ft30 INHERIT pf0;
ALTER FOREIGN TABLE ft40 INHERIT pf0;

CREATE table pf1 (LIKE pl);
ALTER FOREIGN TABLE ft11 INHERIT pf1;
ALTER FOREIGN TABLE ft22 INHERIT pf1;
ALTER FOREIGN TABLE ft33 INHERIT pf1;
ALTER FOREIGN TABLE ft44 INHERIT pf1;

ANALYZE;

----Next_Part(Wed_Jul_06_16_29_14_2016_599)--
Content-Type: Text/Plain; charset=us-ascii
Content-Transfer-Encoding: 7bit
Content-Disposition: inline; filename="testrun.sh"

#! /bin/bash

function do_test() {
	echo $1
	for i in $(seq 1 10); do psql postgres -c "set log_min_duration_statement = 0; set client_min_messages=log; select sum(a) from $1"; done | grep LOG
}

for i in "t0" "pl" "pf0" "pf1";
do
  do_test $i
done


----Next_Part(Wed_Jul_06_16_29_14_2016_599)--
Content-Type: text/plain
Content-Disposition: inline
Content-Transfer-Encoding: 8bit
MIME-Version: 1.0


-- 
Sent via pgsql-hackers mailing list ([email protected])
To make changes to your subscription:
http://www.postgresql.org/mailpref/pgsql-hackers

----Next_Part(Wed_Jul_06_16_29_14_2016_599)----






^ permalink  raw  reply  [nested|flat] 56+ messages in thread

* [PATCH 7/7] Make Append node async-aware.
@ 2016-06-28 09:52 Kyotaro Horiguchi <[email protected]>
  0 siblings, 0 replies; 56+ messages in thread

From: Kyotaro Horiguchi @ 2016-06-28 09:52 UTC (permalink / raw)

Change append node to be capable to handle asynchronous children
properly. As soon as it receives !async_ready from a child, it moves
to the next child and if no child is ready, it sleeps until at least
one of them become ready.
---
 src/backend/executor/nodeAppend.c | 67 +++++++++++++++++++++++++++++++++++++--
 src/include/nodes/execnodes.h     |  2 ++
 2 files changed, 67 insertions(+), 2 deletions(-)

diff --git a/src/backend/executor/nodeAppend.c b/src/backend/executor/nodeAppend.c
index e0ce8c6..9a4063a 100644
--- a/src/backend/executor/nodeAppend.c
+++ b/src/backend/executor/nodeAppend.c
@@ -58,6 +58,7 @@
 #include "postgres.h"
 
 #include "executor/execdebug.h"
+#include "executor/execAsync.h"
 #include "executor/nodeAppend.h"
 
 static bool exec_append_initialize_next(AppendState *appendstate);
@@ -121,6 +122,7 @@ ExecInitAppend(Append *node, EState *estate, int eflags)
 {
 	AppendState *appendstate = makeNode(AppendState);
 	PlanState **appendplanstates;
+	bool	   *finished;
 	int			nplans;
 	int			i;
 	ListCell   *lc;
@@ -134,14 +136,17 @@ ExecInitAppend(Append *node, EState *estate, int eflags)
 	nplans = list_length(node->appendplans);
 
 	appendplanstates = (PlanState **) palloc0(nplans * sizeof(PlanState *));
-
+	finished = (bool *) palloc0(nplans * sizeof(bool));
+	
 	/*
 	 * create new AppendState for our append node
 	 */
 	appendstate->ps.plan = (Plan *) node;
 	appendstate->ps.state = estate;
 	appendstate->appendplans = appendplanstates;
+	appendstate->as_finished = finished;
 	appendstate->as_nplans = nplans;
+	appendstate->as_async = ((eflags & EXEC_FLAG_BACKWARD) == 0);
 
 	/*
 	 * Miscellaneous initialization
@@ -194,6 +199,8 @@ ExecInitAppend(Append *node, EState *estate, int eflags)
 void
 ExecAppend(AppendState *node)
 {
+	int stopplan = node->as_whichplan;
+
 	for (;;)
 	{
 		PlanState  *subnode;
@@ -207,7 +214,36 @@ ExecAppend(AppendState *node)
 		/*
 		 * get a tuple from the subplan
 		 */
-		result = ExecProcNode(subnode);
+		Assert(!node->as_finished[node->as_whichplan]);
+
+		if (!subnode->result_ready)
+			ExecExecuteNode(subnode);
+
+		if (!subnode->result_ready)
+		{
+			if (node->as_async)
+			{
+				/* Move to the next living node */
+				do
+				{
+					node->as_whichplan = 
+						(node->as_whichplan + 1) %  node->as_nplans;
+				} while (node->as_whichplan != stopplan &&
+						 node->as_finished[node->as_whichplan]);
+
+				/* No node is ready yet, return as not-ready */
+				if (node->as_whichplan == stopplan)
+					return;
+
+				/* Try the next node */
+				continue;
+			}
+
+			/* If not async, immediately wait for this subnode */
+			ExecAsyncWaitForNode(subnode);
+		}				
+
+		result = ExecConsumeResult((PlanState *) subnode);
 
 		if (!TupIsNull(result))
 		{
@@ -220,6 +256,31 @@ ExecAppend(AppendState *node)
 			return;
 		}
 
+		if (node->as_async)
+		{
+			node->as_finished[node->as_whichplan] = true;
+			stopplan = node->as_whichplan;
+
+			/* Find the next living subnode */
+			do
+			{
+				node->as_whichplan =
+					(node->as_whichplan + 1) % node->as_nplans;
+			} while (node->as_whichplan != stopplan &&
+					 node->as_finished[node->as_whichplan]);
+
+			if (node->as_whichplan != stopplan)
+			{
+				stopplan = node->as_whichplan;
+				continue;
+			}
+
+			/* All subnodes are exhausted. Finish this node. */
+			ExecReturnTuple(&node->ps,
+							ExecClearTuple(node->ps.ps_ResultTupleSlot));
+			return;
+		}
+
 		/*
 		 * Go on to the "next" subplan in the appropriate direction. If no
 		 * more subplans, return the empty slot set up for us by
@@ -277,6 +338,8 @@ ExecReScanAppend(AppendState *node)
 	{
 		PlanState  *subnode = node->appendplans[i];
 
+		node->as_finished[i] = false;
+
 		/*
 		 * ExecReScan doesn't know about my subplans, so I have to do
 		 * changed-parameter signaling myself.
diff --git a/src/include/nodes/execnodes.h b/src/include/nodes/execnodes.h
index b72decc..b0a86c5 100644
--- a/src/include/nodes/execnodes.h
+++ b/src/include/nodes/execnodes.h
@@ -1177,6 +1177,8 @@ typedef struct AppendState
 {
 	PlanState	ps;				/* its first field is NodeTag */
 	PlanState **appendplans;	/* array of PlanStates for my inputs */
+	bool		as_async;		/* true to allow async execution */
+	bool	   *as_finished;	/* array of the running state of subplans */
 	int			as_nplans;
 	int			as_whichplan;
 } AppendState;
-- 
1.8.3.1


----Next_Part(Thu_Jul_21_18_50_07_2016_597)--
Content-Type: text/plain
Content-Disposition: inline
Content-Transfer-Encoding: 8bit
MIME-Version: 1.0


-- 
Sent via pgsql-hackers mailing list ([email protected])
To make changes to your subscription:
http://www.postgresql.org/mailpref/pgsql-hackers

----Next_Part(Thu_Jul_21_18_50_07_2016_597)----






^ permalink  raw  reply  [nested|flat] 56+ messages in thread

* [PATCH 7/7] Make Append node async-aware.
@ 2016-06-28 09:52 Kyotaro Horiguchi <[email protected]>
  0 siblings, 0 replies; 56+ messages in thread

From: Kyotaro Horiguchi @ 2016-06-28 09:52 UTC (permalink / raw)

Change append node to be capable to handle asynchronous children
properly. As soon as it receives !async_ready from a child, it moves
to the next child and if no child is ready, it sleeps until at least
one of them become ready.
---
 src/backend/executor/nodeAppend.c | 94 ++++++++++++++++++++++++++++++++-------
 src/include/nodes/execnodes.h     |  2 +
 2 files changed, 80 insertions(+), 16 deletions(-)

diff --git a/src/backend/executor/nodeAppend.c b/src/backend/executor/nodeAppend.c
index e0ce8c6..1c0d26e 100644
--- a/src/backend/executor/nodeAppend.c
+++ b/src/backend/executor/nodeAppend.c
@@ -58,6 +58,7 @@
 #include "postgres.h"
 
 #include "executor/execdebug.h"
+#include "executor/execAsync.h"
 #include "executor/nodeAppend.h"
 
 static bool exec_append_initialize_next(AppendState *appendstate);
@@ -121,6 +122,7 @@ ExecInitAppend(Append *node, EState *estate, int eflags)
 {
 	AppendState *appendstate = makeNode(AppendState);
 	PlanState **appendplanstates;
+	bool	   *finished;
 	int			nplans;
 	int			i;
 	ListCell   *lc;
@@ -134,14 +136,17 @@ ExecInitAppend(Append *node, EState *estate, int eflags)
 	nplans = list_length(node->appendplans);
 
 	appendplanstates = (PlanState **) palloc0(nplans * sizeof(PlanState *));
-
+	finished = (bool *) palloc0(nplans * sizeof(bool));
+	
 	/*
 	 * create new AppendState for our append node
 	 */
 	appendstate->ps.plan = (Plan *) node;
 	appendstate->ps.state = estate;
 	appendstate->appendplans = appendplanstates;
+	appendstate->as_finished = finished;
 	appendstate->as_nplans = nplans;
+	appendstate->as_async = ((eflags & EXEC_FLAG_BACKWARD) == 0);
 
 	/*
 	 * Miscellaneous initialization
@@ -194,49 +199,104 @@ ExecInitAppend(Append *node, EState *estate, int eflags)
 void
 ExecAppend(AppendState *node)
 {
-	for (;;)
+	int n_notready = 0;
+	PlanState  *subnode;
+	int i, n;
+
+	n = node->as_whichplan;
+
+	for (i = 0 ; i < node->as_nplans ; i++)
 	{
-		PlanState  *subnode;
 		TupleTableSlot *result;
 
+		if (node->as_async)
+		{
+			if (n >= node->as_nplans)
+				n = 0;
+
+			if (node->as_finished[n])
+			{
+				n++;
+				continue;
+			}
+		}
+
 		/*
 		 * figure out which subplan we are currently processing
 		 */
-		subnode = node->appendplans[node->as_whichplan];
+		subnode = node->appendplans[n];
 
 		/*
-		 * get a tuple from the subplan
+		 * execute the subplan to get a result if it is not ready yet
 		 */
-		result = ExecProcNode(subnode);
+		if (!subnode->result_ready)
+			ExecExecuteNode(subnode);
+
+		/*
+		 * This subnode is not ready yet when asynchrony is not allowed,
+		 * immediately wait for this subnode.
+		 */
+		if (!subnode->result_ready)
+		{
+			if (node->as_async)
+			{
+				n_notready++;
+				n++;
+				continue;
+			}
+			ExecAsyncWaitForNode(subnode);
+		}
+
+		Assert(subnode->result_ready);
+
+		result = ExecConsumeResult((PlanState *)subnode);
 
 		if (!TupIsNull(result))
 		{
 			/*
-			 * If the subplan gave us something then return it as-is. We do
-			 * NOT make use of the result slot that was set up in
-			 * ExecInitAppend; there's no need for it.
+			 * If the subplan gave us something then return it
+			 * as-is. We do NOT make use of the result slot that was
+			 * set up in ExecInitAppend; there's no need for it.
 			 */
+			node->as_whichplan = n;
 			ExecReturnTuple(&node->ps, result);
 			return;
 		}
 
+		/* Tuple has been exhausted on this subnode */
+		if (node->as_async)
+		{
+			node->as_finished[n++] = true;
+			continue;
+		}
+
 		/*
 		 * Go on to the "next" subplan in the appropriate direction. If no
 		 * more subplans, return the empty slot set up for us by
 		 * ExecInitAppend.
 		 */
 		if (ScanDirectionIsForward(node->ps.state->es_direction))
-			node->as_whichplan++;
+		{
+			if (n >= node->as_nplans - 1)
+				break;
+			n++;
+		}
 		else
-			node->as_whichplan--;
-		if (!exec_append_initialize_next(node))
 		{
-			ExecReturnTuple(&node->ps,
-							ExecClearTuple(node->ps.ps_ResultTupleSlot));
-			return;
+			if (n == 0)
+				break;
+			n--;
 		}
+	}
 
-		/* Else loop back and try to get a tuple from the new subplan */
+	/*
+	 * We are finished if reached here and no subnodes are not-ready
+	 */
+	if (n_notready == 0)
+	{
+		node->as_whichplan = n;
+		ExecReturnTuple(&node->ps,
+						ExecClearTuple(node->ps.ps_ResultTupleSlot));
 	}
 }
 
@@ -277,6 +337,8 @@ ExecReScanAppend(AppendState *node)
 	{
 		PlanState  *subnode = node->appendplans[i];
 
+		node->as_finished[i] = false;
+
 		/*
 		 * ExecReScan doesn't know about my subplans, so I have to do
 		 * changed-parameter signaling myself.
diff --git a/src/include/nodes/execnodes.h b/src/include/nodes/execnodes.h
index 9121537..e4f2bb6 100644
--- a/src/include/nodes/execnodes.h
+++ b/src/include/nodes/execnodes.h
@@ -1170,6 +1170,8 @@ typedef struct AppendState
 {
 	PlanState	ps;				/* its first field is NodeTag */
 	PlanState **appendplans;	/* array of PlanStates for my inputs */
+	bool		as_async;		/* true to allow async execution */
+	bool	   *as_finished;	/* array of the running state of subplans */
 	int			as_nplans;
 	int			as_whichplan;
 } AppendState;
-- 
1.8.3.1


----Next_Part(Wed_Jul_06_16_29_14_2016_599)--
Content-Type: Text/Plain; charset=us-ascii
Content-Transfer-Encoding: 7bit
Content-Disposition: inline; filename="gentbl.sql"

CREATE EXTENSION postgres_fdw;
CREATE TABLE pl (a int, b int);
CREATE TABLE cl1 (LIKE pl) INHERITS (pl);
CREATE TABLE cl2 (LIKE pl) INHERITS (pl);
CREATE TABLE cl3 (LIKE pl) INHERITS (pl);
CREATE TABLE cl4 (LIKE pl) INHERITS (pl);
INSERT INTO cl1 (SELECT a, a FROM generate_series(0000000, 0999999) a);
INSERT INTO cl2 (SELECT a, a FROM generate_series(1000000, 1999999) a);
INSERT INTO cl3 (SELECT a, a FROM generate_series(2000000, 2999999) a);
INSERT INTO cl4 (SELECT a, a FROM generate_series(3000000, 3999999) a);

CREATE TABLE t0 (LIKE pl);
INSERT INTO t0 (SELECT a, a FROM generate_series(0000000, 9999999) a);

CREATE SERVER sv0 FOREIGN DATA WRAPPER postgres_fdw OPTIONS (host '/tmp', dbname 'postgres');
CREATE SERVER sv1 FOREIGN DATA WRAPPER postgres_fdw OPTIONS (host '/tmp', dbname 'postgres');
CREATE SERVER sv2 FOREIGN DATA WRAPPER postgres_fdw OPTIONS (host '/tmp', dbname 'postgres');
CREATE SERVER sv3 FOREIGN DATA WRAPPER postgres_fdw OPTIONS (host '/tmp', dbname 'postgres');
CREATE SERVER sv4 FOREIGN DATA WRAPPER postgres_fdw OPTIONS (host '/tmp', dbname 'postgres');
CREATE USER MAPPING FOR public SERVER sv0;
CREATE USER MAPPING FOR public SERVER sv1;
CREATE USER MAPPING FOR public SERVER sv2;
CREATE USER MAPPING FOR public SERVER sv3;
CREATE USER MAPPING FOR public SERVER sv4;

CREATE FOREIGN TABLE ft10 (a int, b int) SERVER sv0 OPTIONS (table_name 'cl1');
CREATE FOREIGN TABLE ft20 (a int, b int) SERVER sv0 OPTIONS (table_name 'cl2');
CREATE FOREIGN TABLE ft30 (a int, b int) SERVER sv0 OPTIONS (table_name 'cl3');
CREATE FOREIGN TABLE ft40 (a int, b int) SERVER sv0 OPTIONS (table_name 'cl4');
CREATE FOREIGN TABLE ft11 (a int, b int) SERVER sv1 OPTIONS (table_name 'cl1');
CREATE FOREIGN TABLE ft22 (a int, b int) SERVER sv2 OPTIONS (table_name 'cl2');
CREATE FOREIGN TABLE ft33 (a int, b int) SERVER sv3 OPTIONS (table_name 'cl3');
CREATE FOREIGN TABLE ft44 (a int, b int) SERVER sv4 OPTIONS (table_name 'cl4');

CREATE TABLE pf0 (LIKE pl);
ALTER FOREIGN TABLE ft10 INHERIT pf0;
ALTER FOREIGN TABLE ft20 INHERIT pf0;
ALTER FOREIGN TABLE ft30 INHERIT pf0;
ALTER FOREIGN TABLE ft40 INHERIT pf0;

CREATE table pf1 (LIKE pl);
ALTER FOREIGN TABLE ft11 INHERIT pf1;
ALTER FOREIGN TABLE ft22 INHERIT pf1;
ALTER FOREIGN TABLE ft33 INHERIT pf1;
ALTER FOREIGN TABLE ft44 INHERIT pf1;

ANALYZE;

----Next_Part(Wed_Jul_06_16_29_14_2016_599)--
Content-Type: Text/Plain; charset=us-ascii
Content-Transfer-Encoding: 7bit
Content-Disposition: inline; filename="testrun.sh"

#! /bin/bash

function do_test() {
	echo $1
	for i in $(seq 1 10); do psql postgres -c "set log_min_duration_statement = 0; set client_min_messages=log; select sum(a) from $1"; done | grep LOG
}

for i in "t0" "pl" "pf0" "pf1";
do
  do_test $i
done


----Next_Part(Wed_Jul_06_16_29_14_2016_599)--
Content-Type: text/plain
Content-Disposition: inline
Content-Transfer-Encoding: 8bit
MIME-Version: 1.0


-- 
Sent via pgsql-hackers mailing list ([email protected])
To make changes to your subscription:
http://www.postgresql.org/mailpref/pgsql-hackers

----Next_Part(Wed_Jul_06_16_29_14_2016_599)----






^ permalink  raw  reply  [nested|flat] 56+ messages in thread

* [PATCH 7/7] Make Append node async-aware.
@ 2016-06-28 09:52 Kyotaro Horiguchi <[email protected]>
  0 siblings, 0 replies; 56+ messages in thread

From: Kyotaro Horiguchi @ 2016-06-28 09:52 UTC (permalink / raw)

Change append node to be capable to handle asynchronous children
properly. As soon as it receives !async_ready from a child, it moves
to the next child and if no child is ready, it sleeps until at least
one of them become ready.
---
 src/backend/executor/nodeAppend.c | 67 +++++++++++++++++++++++++++++++++++++--
 src/include/nodes/execnodes.h     |  2 ++
 2 files changed, 67 insertions(+), 2 deletions(-)

diff --git a/src/backend/executor/nodeAppend.c b/src/backend/executor/nodeAppend.c
index e0ce8c6..9a4063a 100644
--- a/src/backend/executor/nodeAppend.c
+++ b/src/backend/executor/nodeAppend.c
@@ -58,6 +58,7 @@
 #include "postgres.h"
 
 #include "executor/execdebug.h"
+#include "executor/execAsync.h"
 #include "executor/nodeAppend.h"
 
 static bool exec_append_initialize_next(AppendState *appendstate);
@@ -121,6 +122,7 @@ ExecInitAppend(Append *node, EState *estate, int eflags)
 {
 	AppendState *appendstate = makeNode(AppendState);
 	PlanState **appendplanstates;
+	bool	   *finished;
 	int			nplans;
 	int			i;
 	ListCell   *lc;
@@ -134,14 +136,17 @@ ExecInitAppend(Append *node, EState *estate, int eflags)
 	nplans = list_length(node->appendplans);
 
 	appendplanstates = (PlanState **) palloc0(nplans * sizeof(PlanState *));
-
+	finished = (bool *) palloc0(nplans * sizeof(bool));
+	
 	/*
 	 * create new AppendState for our append node
 	 */
 	appendstate->ps.plan = (Plan *) node;
 	appendstate->ps.state = estate;
 	appendstate->appendplans = appendplanstates;
+	appendstate->as_finished = finished;
 	appendstate->as_nplans = nplans;
+	appendstate->as_async = ((eflags & EXEC_FLAG_BACKWARD) == 0);
 
 	/*
 	 * Miscellaneous initialization
@@ -194,6 +199,8 @@ ExecInitAppend(Append *node, EState *estate, int eflags)
 void
 ExecAppend(AppendState *node)
 {
+	int stopplan = node->as_whichplan;
+
 	for (;;)
 	{
 		PlanState  *subnode;
@@ -207,7 +214,36 @@ ExecAppend(AppendState *node)
 		/*
 		 * get a tuple from the subplan
 		 */
-		result = ExecProcNode(subnode);
+		Assert(!node->as_finished[node->as_whichplan]);
+
+		if (!subnode->result_ready)
+			ExecExecuteNode(subnode);
+
+		if (!subnode->result_ready)
+		{
+			if (node->as_async)
+			{
+				/* Move to the next living node */
+				do
+				{
+					node->as_whichplan = 
+						(node->as_whichplan + 1) %  node->as_nplans;
+				} while (node->as_whichplan != stopplan &&
+						 node->as_finished[node->as_whichplan]);
+
+				/* No node is ready yet, return as not-ready */
+				if (node->as_whichplan == stopplan)
+					return;
+
+				/* Try the next node */
+				continue;
+			}
+
+			/* If not async, immediately wait for this subnode */
+			ExecAsyncWaitForNode(subnode);
+		}				
+
+		result = ExecConsumeResult((PlanState *) subnode);
 
 		if (!TupIsNull(result))
 		{
@@ -220,6 +256,31 @@ ExecAppend(AppendState *node)
 			return;
 		}
 
+		if (node->as_async)
+		{
+			node->as_finished[node->as_whichplan] = true;
+			stopplan = node->as_whichplan;
+
+			/* Find the next living subnode */
+			do
+			{
+				node->as_whichplan =
+					(node->as_whichplan + 1) % node->as_nplans;
+			} while (node->as_whichplan != stopplan &&
+					 node->as_finished[node->as_whichplan]);
+
+			if (node->as_whichplan != stopplan)
+			{
+				stopplan = node->as_whichplan;
+				continue;
+			}
+
+			/* All subnodes are exhausted. Finish this node. */
+			ExecReturnTuple(&node->ps,
+							ExecClearTuple(node->ps.ps_ResultTupleSlot));
+			return;
+		}
+
 		/*
 		 * Go on to the "next" subplan in the appropriate direction. If no
 		 * more subplans, return the empty slot set up for us by
@@ -277,6 +338,8 @@ ExecReScanAppend(AppendState *node)
 	{
 		PlanState  *subnode = node->appendplans[i];
 
+		node->as_finished[i] = false;
+
 		/*
 		 * ExecReScan doesn't know about my subplans, so I have to do
 		 * changed-parameter signaling myself.
diff --git a/src/include/nodes/execnodes.h b/src/include/nodes/execnodes.h
index b72decc..b0a86c5 100644
--- a/src/include/nodes/execnodes.h
+++ b/src/include/nodes/execnodes.h
@@ -1177,6 +1177,8 @@ typedef struct AppendState
 {
 	PlanState	ps;				/* its first field is NodeTag */
 	PlanState **appendplans;	/* array of PlanStates for my inputs */
+	bool		as_async;		/* true to allow async execution */
+	bool	   *as_finished;	/* array of the running state of subplans */
 	int			as_nplans;
 	int			as_whichplan;
 } AppendState;
-- 
2.9.2


----Next_Part(Mon_Aug_29_17_08_36_2016_213)--
Content-Type: text/plain
Content-Disposition: inline
Content-Transfer-Encoding: 8bit
MIME-Version: 1.0


-- 
Sent via pgsql-hackers mailing list ([email protected])
To make changes to your subscription:
http://www.postgresql.org/mailpref/pgsql-hackers

----Next_Part(Mon_Aug_29_17_08_36_2016_213)----






^ permalink  raw  reply  [nested|flat] 56+ messages in thread

* [PATCH 7/7] Make Append node async-aware.
@ 2016-06-28 09:52 Kyotaro Horiguchi <[email protected]>
  0 siblings, 0 replies; 56+ messages in thread

From: Kyotaro Horiguchi @ 2016-06-28 09:52 UTC (permalink / raw)

Change append node to be capable to handle asynchronous children
properly. As soon as it receives !async_ready from a child, it moves
to the next child and if no child is ready, it sleeps until at least
one of them become ready.
---
 src/backend/executor/nodeAppend.c | 67 +++++++++++++++++++++++++++++++++++++--
 src/include/nodes/execnodes.h     |  2 ++
 2 files changed, 67 insertions(+), 2 deletions(-)

diff --git a/src/backend/executor/nodeAppend.c b/src/backend/executor/nodeAppend.c
index e0ce8c6..9a4063a 100644
--- a/src/backend/executor/nodeAppend.c
+++ b/src/backend/executor/nodeAppend.c
@@ -58,6 +58,7 @@
 #include "postgres.h"
 
 #include "executor/execdebug.h"
+#include "executor/execAsync.h"
 #include "executor/nodeAppend.h"
 
 static bool exec_append_initialize_next(AppendState *appendstate);
@@ -121,6 +122,7 @@ ExecInitAppend(Append *node, EState *estate, int eflags)
 {
 	AppendState *appendstate = makeNode(AppendState);
 	PlanState **appendplanstates;
+	bool	   *finished;
 	int			nplans;
 	int			i;
 	ListCell   *lc;
@@ -134,14 +136,17 @@ ExecInitAppend(Append *node, EState *estate, int eflags)
 	nplans = list_length(node->appendplans);
 
 	appendplanstates = (PlanState **) palloc0(nplans * sizeof(PlanState *));
-
+	finished = (bool *) palloc0(nplans * sizeof(bool));
+	
 	/*
 	 * create new AppendState for our append node
 	 */
 	appendstate->ps.plan = (Plan *) node;
 	appendstate->ps.state = estate;
 	appendstate->appendplans = appendplanstates;
+	appendstate->as_finished = finished;
 	appendstate->as_nplans = nplans;
+	appendstate->as_async = ((eflags & EXEC_FLAG_BACKWARD) == 0);
 
 	/*
 	 * Miscellaneous initialization
@@ -194,6 +199,8 @@ ExecInitAppend(Append *node, EState *estate, int eflags)
 void
 ExecAppend(AppendState *node)
 {
+	int stopplan = node->as_whichplan;
+
 	for (;;)
 	{
 		PlanState  *subnode;
@@ -207,7 +214,36 @@ ExecAppend(AppendState *node)
 		/*
 		 * get a tuple from the subplan
 		 */
-		result = ExecProcNode(subnode);
+		Assert(!node->as_finished[node->as_whichplan]);
+
+		if (!subnode->result_ready)
+			ExecExecuteNode(subnode);
+
+		if (!subnode->result_ready)
+		{
+			if (node->as_async)
+			{
+				/* Move to the next living node */
+				do
+				{
+					node->as_whichplan = 
+						(node->as_whichplan + 1) %  node->as_nplans;
+				} while (node->as_whichplan != stopplan &&
+						 node->as_finished[node->as_whichplan]);
+
+				/* No node is ready yet, return as not-ready */
+				if (node->as_whichplan == stopplan)
+					return;
+
+				/* Try the next node */
+				continue;
+			}
+
+			/* If not async, immediately wait for this subnode */
+			ExecAsyncWaitForNode(subnode);
+		}				
+
+		result = ExecConsumeResult((PlanState *) subnode);
 
 		if (!TupIsNull(result))
 		{
@@ -220,6 +256,31 @@ ExecAppend(AppendState *node)
 			return;
 		}
 
+		if (node->as_async)
+		{
+			node->as_finished[node->as_whichplan] = true;
+			stopplan = node->as_whichplan;
+
+			/* Find the next living subnode */
+			do
+			{
+				node->as_whichplan =
+					(node->as_whichplan + 1) % node->as_nplans;
+			} while (node->as_whichplan != stopplan &&
+					 node->as_finished[node->as_whichplan]);
+
+			if (node->as_whichplan != stopplan)
+			{
+				stopplan = node->as_whichplan;
+				continue;
+			}
+
+			/* All subnodes are exhausted. Finish this node. */
+			ExecReturnTuple(&node->ps,
+							ExecClearTuple(node->ps.ps_ResultTupleSlot));
+			return;
+		}
+
 		/*
 		 * Go on to the "next" subplan in the appropriate direction. If no
 		 * more subplans, return the empty slot set up for us by
@@ -277,6 +338,8 @@ ExecReScanAppend(AppendState *node)
 	{
 		PlanState  *subnode = node->appendplans[i];
 
+		node->as_finished[i] = false;
+
 		/*
 		 * ExecReScan doesn't know about my subplans, so I have to do
 		 * changed-parameter signaling myself.
diff --git a/src/include/nodes/execnodes.h b/src/include/nodes/execnodes.h
index b72decc..b0a86c5 100644
--- a/src/include/nodes/execnodes.h
+++ b/src/include/nodes/execnodes.h
@@ -1177,6 +1177,8 @@ typedef struct AppendState
 {
 	PlanState	ps;				/* its first field is NodeTag */
 	PlanState **appendplans;	/* array of PlanStates for my inputs */
+	bool		as_async;		/* true to allow async execution */
+	bool	   *as_finished;	/* array of the running state of subplans */
 	int			as_nplans;
 	int			as_whichplan;
 } AppendState;
-- 
2.9.2


----Next_Part(Mon_Aug_29_17_08_36_2016_213)--
Content-Type: text/plain
Content-Disposition: inline
Content-Transfer-Encoding: 8bit
MIME-Version: 1.0


-- 
Sent via pgsql-hackers mailing list ([email protected])
To make changes to your subscription:
http://www.postgresql.org/mailpref/pgsql-hackers

----Next_Part(Mon_Aug_29_17_08_36_2016_213)----






^ permalink  raw  reply  [nested|flat] 56+ messages in thread

* [PATCH 7/7] Make Append node async-aware.
@ 2016-06-28 09:52 Kyotaro Horiguchi <[email protected]>
  0 siblings, 0 replies; 56+ messages in thread

From: Kyotaro Horiguchi @ 2016-06-28 09:52 UTC (permalink / raw)

Change append node to be capable to handle asynchronous children
properly. As soon as it receives !async_ready from a child, it moves
to the next child and if no child is ready, it sleeps until at least
one of them become ready.
---
 src/backend/executor/nodeAppend.c | 94 ++++++++++++++++++++++++++++++++-------
 src/include/nodes/execnodes.h     |  2 +
 2 files changed, 80 insertions(+), 16 deletions(-)

diff --git a/src/backend/executor/nodeAppend.c b/src/backend/executor/nodeAppend.c
index e0ce8c6..1c0d26e 100644
--- a/src/backend/executor/nodeAppend.c
+++ b/src/backend/executor/nodeAppend.c
@@ -58,6 +58,7 @@
 #include "postgres.h"
 
 #include "executor/execdebug.h"
+#include "executor/execAsync.h"
 #include "executor/nodeAppend.h"
 
 static bool exec_append_initialize_next(AppendState *appendstate);
@@ -121,6 +122,7 @@ ExecInitAppend(Append *node, EState *estate, int eflags)
 {
 	AppendState *appendstate = makeNode(AppendState);
 	PlanState **appendplanstates;
+	bool	   *finished;
 	int			nplans;
 	int			i;
 	ListCell   *lc;
@@ -134,14 +136,17 @@ ExecInitAppend(Append *node, EState *estate, int eflags)
 	nplans = list_length(node->appendplans);
 
 	appendplanstates = (PlanState **) palloc0(nplans * sizeof(PlanState *));
-
+	finished = (bool *) palloc0(nplans * sizeof(bool));
+	
 	/*
 	 * create new AppendState for our append node
 	 */
 	appendstate->ps.plan = (Plan *) node;
 	appendstate->ps.state = estate;
 	appendstate->appendplans = appendplanstates;
+	appendstate->as_finished = finished;
 	appendstate->as_nplans = nplans;
+	appendstate->as_async = ((eflags & EXEC_FLAG_BACKWARD) == 0);
 
 	/*
 	 * Miscellaneous initialization
@@ -194,49 +199,104 @@ ExecInitAppend(Append *node, EState *estate, int eflags)
 void
 ExecAppend(AppendState *node)
 {
-	for (;;)
+	int n_notready = 0;
+	PlanState  *subnode;
+	int i, n;
+
+	n = node->as_whichplan;
+
+	for (i = 0 ; i < node->as_nplans ; i++)
 	{
-		PlanState  *subnode;
 		TupleTableSlot *result;
 
+		if (node->as_async)
+		{
+			if (n >= node->as_nplans)
+				n = 0;
+
+			if (node->as_finished[n])
+			{
+				n++;
+				continue;
+			}
+		}
+
 		/*
 		 * figure out which subplan we are currently processing
 		 */
-		subnode = node->appendplans[node->as_whichplan];
+		subnode = node->appendplans[n];
 
 		/*
-		 * get a tuple from the subplan
+		 * execute the subplan to get a result if it is not ready yet
 		 */
-		result = ExecProcNode(subnode);
+		if (!subnode->result_ready)
+			ExecExecuteNode(subnode);
+
+		/*
+		 * This subnode is not ready yet when asynchrony is not allowed,
+		 * immediately wait for this subnode.
+		 */
+		if (!subnode->result_ready)
+		{
+			if (node->as_async)
+			{
+				n_notready++;
+				n++;
+				continue;
+			}
+			ExecAsyncWaitForNode(subnode);
+		}
+
+		Assert(subnode->result_ready);
+
+		result = ExecConsumeResult((PlanState *)subnode);
 
 		if (!TupIsNull(result))
 		{
 			/*
-			 * If the subplan gave us something then return it as-is. We do
-			 * NOT make use of the result slot that was set up in
-			 * ExecInitAppend; there's no need for it.
+			 * If the subplan gave us something then return it
+			 * as-is. We do NOT make use of the result slot that was
+			 * set up in ExecInitAppend; there's no need for it.
 			 */
+			node->as_whichplan = n;
 			ExecReturnTuple(&node->ps, result);
 			return;
 		}
 
+		/* Tuple has been exhausted on this subnode */
+		if (node->as_async)
+		{
+			node->as_finished[n++] = true;
+			continue;
+		}
+
 		/*
 		 * Go on to the "next" subplan in the appropriate direction. If no
 		 * more subplans, return the empty slot set up for us by
 		 * ExecInitAppend.
 		 */
 		if (ScanDirectionIsForward(node->ps.state->es_direction))
-			node->as_whichplan++;
+		{
+			if (n >= node->as_nplans - 1)
+				break;
+			n++;
+		}
 		else
-			node->as_whichplan--;
-		if (!exec_append_initialize_next(node))
 		{
-			ExecReturnTuple(&node->ps,
-							ExecClearTuple(node->ps.ps_ResultTupleSlot));
-			return;
+			if (n == 0)
+				break;
+			n--;
 		}
+	}
 
-		/* Else loop back and try to get a tuple from the new subplan */
+	/*
+	 * We are finished if reached here and no subnodes are not-ready
+	 */
+	if (n_notready == 0)
+	{
+		node->as_whichplan = n;
+		ExecReturnTuple(&node->ps,
+						ExecClearTuple(node->ps.ps_ResultTupleSlot));
 	}
 }
 
@@ -277,6 +337,8 @@ ExecReScanAppend(AppendState *node)
 	{
 		PlanState  *subnode = node->appendplans[i];
 
+		node->as_finished[i] = false;
+
 		/*
 		 * ExecReScan doesn't know about my subplans, so I have to do
 		 * changed-parameter signaling myself.
diff --git a/src/include/nodes/execnodes.h b/src/include/nodes/execnodes.h
index 9121537..e4f2bb6 100644
--- a/src/include/nodes/execnodes.h
+++ b/src/include/nodes/execnodes.h
@@ -1170,6 +1170,8 @@ typedef struct AppendState
 {
 	PlanState	ps;				/* its first field is NodeTag */
 	PlanState **appendplans;	/* array of PlanStates for my inputs */
+	bool		as_async;		/* true to allow async execution */
+	bool	   *as_finished;	/* array of the running state of subplans */
 	int			as_nplans;
 	int			as_whichplan;
 } AppendState;
-- 
1.8.3.1


----Next_Part(Wed_Jul_06_16_29_14_2016_599)--
Content-Type: Text/Plain; charset=us-ascii
Content-Transfer-Encoding: 7bit
Content-Disposition: inline; filename="gentbl.sql"

CREATE EXTENSION postgres_fdw;
CREATE TABLE pl (a int, b int);
CREATE TABLE cl1 (LIKE pl) INHERITS (pl);
CREATE TABLE cl2 (LIKE pl) INHERITS (pl);
CREATE TABLE cl3 (LIKE pl) INHERITS (pl);
CREATE TABLE cl4 (LIKE pl) INHERITS (pl);
INSERT INTO cl1 (SELECT a, a FROM generate_series(0000000, 0999999) a);
INSERT INTO cl2 (SELECT a, a FROM generate_series(1000000, 1999999) a);
INSERT INTO cl3 (SELECT a, a FROM generate_series(2000000, 2999999) a);
INSERT INTO cl4 (SELECT a, a FROM generate_series(3000000, 3999999) a);

CREATE TABLE t0 (LIKE pl);
INSERT INTO t0 (SELECT a, a FROM generate_series(0000000, 9999999) a);

CREATE SERVER sv0 FOREIGN DATA WRAPPER postgres_fdw OPTIONS (host '/tmp', dbname 'postgres');
CREATE SERVER sv1 FOREIGN DATA WRAPPER postgres_fdw OPTIONS (host '/tmp', dbname 'postgres');
CREATE SERVER sv2 FOREIGN DATA WRAPPER postgres_fdw OPTIONS (host '/tmp', dbname 'postgres');
CREATE SERVER sv3 FOREIGN DATA WRAPPER postgres_fdw OPTIONS (host '/tmp', dbname 'postgres');
CREATE SERVER sv4 FOREIGN DATA WRAPPER postgres_fdw OPTIONS (host '/tmp', dbname 'postgres');
CREATE USER MAPPING FOR public SERVER sv0;
CREATE USER MAPPING FOR public SERVER sv1;
CREATE USER MAPPING FOR public SERVER sv2;
CREATE USER MAPPING FOR public SERVER sv3;
CREATE USER MAPPING FOR public SERVER sv4;

CREATE FOREIGN TABLE ft10 (a int, b int) SERVER sv0 OPTIONS (table_name 'cl1');
CREATE FOREIGN TABLE ft20 (a int, b int) SERVER sv0 OPTIONS (table_name 'cl2');
CREATE FOREIGN TABLE ft30 (a int, b int) SERVER sv0 OPTIONS (table_name 'cl3');
CREATE FOREIGN TABLE ft40 (a int, b int) SERVER sv0 OPTIONS (table_name 'cl4');
CREATE FOREIGN TABLE ft11 (a int, b int) SERVER sv1 OPTIONS (table_name 'cl1');
CREATE FOREIGN TABLE ft22 (a int, b int) SERVER sv2 OPTIONS (table_name 'cl2');
CREATE FOREIGN TABLE ft33 (a int, b int) SERVER sv3 OPTIONS (table_name 'cl3');
CREATE FOREIGN TABLE ft44 (a int, b int) SERVER sv4 OPTIONS (table_name 'cl4');

CREATE TABLE pf0 (LIKE pl);
ALTER FOREIGN TABLE ft10 INHERIT pf0;
ALTER FOREIGN TABLE ft20 INHERIT pf0;
ALTER FOREIGN TABLE ft30 INHERIT pf0;
ALTER FOREIGN TABLE ft40 INHERIT pf0;

CREATE table pf1 (LIKE pl);
ALTER FOREIGN TABLE ft11 INHERIT pf1;
ALTER FOREIGN TABLE ft22 INHERIT pf1;
ALTER FOREIGN TABLE ft33 INHERIT pf1;
ALTER FOREIGN TABLE ft44 INHERIT pf1;

ANALYZE;

----Next_Part(Wed_Jul_06_16_29_14_2016_599)--
Content-Type: Text/Plain; charset=us-ascii
Content-Transfer-Encoding: 7bit
Content-Disposition: inline; filename="testrun.sh"

#! /bin/bash

function do_test() {
	echo $1
	for i in $(seq 1 10); do psql postgres -c "set log_min_duration_statement = 0; set client_min_messages=log; select sum(a) from $1"; done | grep LOG
}

for i in "t0" "pl" "pf0" "pf1";
do
  do_test $i
done


----Next_Part(Wed_Jul_06_16_29_14_2016_599)--
Content-Type: text/plain
Content-Disposition: inline
Content-Transfer-Encoding: 8bit
MIME-Version: 1.0


-- 
Sent via pgsql-hackers mailing list ([email protected])
To make changes to your subscription:
http://www.postgresql.org/mailpref/pgsql-hackers

----Next_Part(Wed_Jul_06_16_29_14_2016_599)----






^ permalink  raw  reply  [nested|flat] 56+ messages in thread

* [PATCH 7/7] Make Append node async-aware.
@ 2016-06-28 09:52 Kyotaro Horiguchi <[email protected]>
  0 siblings, 0 replies; 56+ messages in thread

From: Kyotaro Horiguchi @ 2016-06-28 09:52 UTC (permalink / raw)

Change append node to be capable to handle asynchronous children
properly. As soon as it receives !async_ready from a child, it moves
to the next child and if no child is ready, it sleeps until at least
one of them become ready.
---
 src/backend/executor/nodeAppend.c | 94 ++++++++++++++++++++++++++++++++-------
 src/include/nodes/execnodes.h     |  2 +
 2 files changed, 80 insertions(+), 16 deletions(-)

diff --git a/src/backend/executor/nodeAppend.c b/src/backend/executor/nodeAppend.c
index e0ce8c6..1c0d26e 100644
--- a/src/backend/executor/nodeAppend.c
+++ b/src/backend/executor/nodeAppend.c
@@ -58,6 +58,7 @@
 #include "postgres.h"
 
 #include "executor/execdebug.h"
+#include "executor/execAsync.h"
 #include "executor/nodeAppend.h"
 
 static bool exec_append_initialize_next(AppendState *appendstate);
@@ -121,6 +122,7 @@ ExecInitAppend(Append *node, EState *estate, int eflags)
 {
 	AppendState *appendstate = makeNode(AppendState);
 	PlanState **appendplanstates;
+	bool	   *finished;
 	int			nplans;
 	int			i;
 	ListCell   *lc;
@@ -134,14 +136,17 @@ ExecInitAppend(Append *node, EState *estate, int eflags)
 	nplans = list_length(node->appendplans);
 
 	appendplanstates = (PlanState **) palloc0(nplans * sizeof(PlanState *));
-
+	finished = (bool *) palloc0(nplans * sizeof(bool));
+	
 	/*
 	 * create new AppendState for our append node
 	 */
 	appendstate->ps.plan = (Plan *) node;
 	appendstate->ps.state = estate;
 	appendstate->appendplans = appendplanstates;
+	appendstate->as_finished = finished;
 	appendstate->as_nplans = nplans;
+	appendstate->as_async = ((eflags & EXEC_FLAG_BACKWARD) == 0);
 
 	/*
 	 * Miscellaneous initialization
@@ -194,49 +199,104 @@ ExecInitAppend(Append *node, EState *estate, int eflags)
 void
 ExecAppend(AppendState *node)
 {
-	for (;;)
+	int n_notready = 0;
+	PlanState  *subnode;
+	int i, n;
+
+	n = node->as_whichplan;
+
+	for (i = 0 ; i < node->as_nplans ; i++)
 	{
-		PlanState  *subnode;
 		TupleTableSlot *result;
 
+		if (node->as_async)
+		{
+			if (n >= node->as_nplans)
+				n = 0;
+
+			if (node->as_finished[n])
+			{
+				n++;
+				continue;
+			}
+		}
+
 		/*
 		 * figure out which subplan we are currently processing
 		 */
-		subnode = node->appendplans[node->as_whichplan];
+		subnode = node->appendplans[n];
 
 		/*
-		 * get a tuple from the subplan
+		 * execute the subplan to get a result if it is not ready yet
 		 */
-		result = ExecProcNode(subnode);
+		if (!subnode->result_ready)
+			ExecExecuteNode(subnode);
+
+		/*
+		 * This subnode is not ready yet when asynchrony is not allowed,
+		 * immediately wait for this subnode.
+		 */
+		if (!subnode->result_ready)
+		{
+			if (node->as_async)
+			{
+				n_notready++;
+				n++;
+				continue;
+			}
+			ExecAsyncWaitForNode(subnode);
+		}
+
+		Assert(subnode->result_ready);
+
+		result = ExecConsumeResult((PlanState *)subnode);
 
 		if (!TupIsNull(result))
 		{
 			/*
-			 * If the subplan gave us something then return it as-is. We do
-			 * NOT make use of the result slot that was set up in
-			 * ExecInitAppend; there's no need for it.
+			 * If the subplan gave us something then return it
+			 * as-is. We do NOT make use of the result slot that was
+			 * set up in ExecInitAppend; there's no need for it.
 			 */
+			node->as_whichplan = n;
 			ExecReturnTuple(&node->ps, result);
 			return;
 		}
 
+		/* Tuple has been exhausted on this subnode */
+		if (node->as_async)
+		{
+			node->as_finished[n++] = true;
+			continue;
+		}
+
 		/*
 		 * Go on to the "next" subplan in the appropriate direction. If no
 		 * more subplans, return the empty slot set up for us by
 		 * ExecInitAppend.
 		 */
 		if (ScanDirectionIsForward(node->ps.state->es_direction))
-			node->as_whichplan++;
+		{
+			if (n >= node->as_nplans - 1)
+				break;
+			n++;
+		}
 		else
-			node->as_whichplan--;
-		if (!exec_append_initialize_next(node))
 		{
-			ExecReturnTuple(&node->ps,
-							ExecClearTuple(node->ps.ps_ResultTupleSlot));
-			return;
+			if (n == 0)
+				break;
+			n--;
 		}
+	}
 
-		/* Else loop back and try to get a tuple from the new subplan */
+	/*
+	 * We are finished if reached here and no subnodes are not-ready
+	 */
+	if (n_notready == 0)
+	{
+		node->as_whichplan = n;
+		ExecReturnTuple(&node->ps,
+						ExecClearTuple(node->ps.ps_ResultTupleSlot));
 	}
 }
 
@@ -277,6 +337,8 @@ ExecReScanAppend(AppendState *node)
 	{
 		PlanState  *subnode = node->appendplans[i];
 
+		node->as_finished[i] = false;
+
 		/*
 		 * ExecReScan doesn't know about my subplans, so I have to do
 		 * changed-parameter signaling myself.
diff --git a/src/include/nodes/execnodes.h b/src/include/nodes/execnodes.h
index 9121537..e4f2bb6 100644
--- a/src/include/nodes/execnodes.h
+++ b/src/include/nodes/execnodes.h
@@ -1170,6 +1170,8 @@ typedef struct AppendState
 {
 	PlanState	ps;				/* its first field is NodeTag */
 	PlanState **appendplans;	/* array of PlanStates for my inputs */
+	bool		as_async;		/* true to allow async execution */
+	bool	   *as_finished;	/* array of the running state of subplans */
 	int			as_nplans;
 	int			as_whichplan;
 } AppendState;
-- 
1.8.3.1


----Next_Part(Wed_Jul_06_16_29_14_2016_599)--
Content-Type: Text/Plain; charset=us-ascii
Content-Transfer-Encoding: 7bit
Content-Disposition: inline; filename="gentbl.sql"

CREATE EXTENSION postgres_fdw;
CREATE TABLE pl (a int, b int);
CREATE TABLE cl1 (LIKE pl) INHERITS (pl);
CREATE TABLE cl2 (LIKE pl) INHERITS (pl);
CREATE TABLE cl3 (LIKE pl) INHERITS (pl);
CREATE TABLE cl4 (LIKE pl) INHERITS (pl);
INSERT INTO cl1 (SELECT a, a FROM generate_series(0000000, 0999999) a);
INSERT INTO cl2 (SELECT a, a FROM generate_series(1000000, 1999999) a);
INSERT INTO cl3 (SELECT a, a FROM generate_series(2000000, 2999999) a);
INSERT INTO cl4 (SELECT a, a FROM generate_series(3000000, 3999999) a);

CREATE TABLE t0 (LIKE pl);
INSERT INTO t0 (SELECT a, a FROM generate_series(0000000, 9999999) a);

CREATE SERVER sv0 FOREIGN DATA WRAPPER postgres_fdw OPTIONS (host '/tmp', dbname 'postgres');
CREATE SERVER sv1 FOREIGN DATA WRAPPER postgres_fdw OPTIONS (host '/tmp', dbname 'postgres');
CREATE SERVER sv2 FOREIGN DATA WRAPPER postgres_fdw OPTIONS (host '/tmp', dbname 'postgres');
CREATE SERVER sv3 FOREIGN DATA WRAPPER postgres_fdw OPTIONS (host '/tmp', dbname 'postgres');
CREATE SERVER sv4 FOREIGN DATA WRAPPER postgres_fdw OPTIONS (host '/tmp', dbname 'postgres');
CREATE USER MAPPING FOR public SERVER sv0;
CREATE USER MAPPING FOR public SERVER sv1;
CREATE USER MAPPING FOR public SERVER sv2;
CREATE USER MAPPING FOR public SERVER sv3;
CREATE USER MAPPING FOR public SERVER sv4;

CREATE FOREIGN TABLE ft10 (a int, b int) SERVER sv0 OPTIONS (table_name 'cl1');
CREATE FOREIGN TABLE ft20 (a int, b int) SERVER sv0 OPTIONS (table_name 'cl2');
CREATE FOREIGN TABLE ft30 (a int, b int) SERVER sv0 OPTIONS (table_name 'cl3');
CREATE FOREIGN TABLE ft40 (a int, b int) SERVER sv0 OPTIONS (table_name 'cl4');
CREATE FOREIGN TABLE ft11 (a int, b int) SERVER sv1 OPTIONS (table_name 'cl1');
CREATE FOREIGN TABLE ft22 (a int, b int) SERVER sv2 OPTIONS (table_name 'cl2');
CREATE FOREIGN TABLE ft33 (a int, b int) SERVER sv3 OPTIONS (table_name 'cl3');
CREATE FOREIGN TABLE ft44 (a int, b int) SERVER sv4 OPTIONS (table_name 'cl4');

CREATE TABLE pf0 (LIKE pl);
ALTER FOREIGN TABLE ft10 INHERIT pf0;
ALTER FOREIGN TABLE ft20 INHERIT pf0;
ALTER FOREIGN TABLE ft30 INHERIT pf0;
ALTER FOREIGN TABLE ft40 INHERIT pf0;

CREATE table pf1 (LIKE pl);
ALTER FOREIGN TABLE ft11 INHERIT pf1;
ALTER FOREIGN TABLE ft22 INHERIT pf1;
ALTER FOREIGN TABLE ft33 INHERIT pf1;
ALTER FOREIGN TABLE ft44 INHERIT pf1;

ANALYZE;

----Next_Part(Wed_Jul_06_16_29_14_2016_599)--
Content-Type: Text/Plain; charset=us-ascii
Content-Transfer-Encoding: 7bit
Content-Disposition: inline; filename="testrun.sh"

#! /bin/bash

function do_test() {
	echo $1
	for i in $(seq 1 10); do psql postgres -c "set log_min_duration_statement = 0; set client_min_messages=log; select sum(a) from $1"; done | grep LOG
}

for i in "t0" "pl" "pf0" "pf1";
do
  do_test $i
done


----Next_Part(Wed_Jul_06_16_29_14_2016_599)--
Content-Type: text/plain
Content-Disposition: inline
Content-Transfer-Encoding: 8bit
MIME-Version: 1.0


-- 
Sent via pgsql-hackers mailing list ([email protected])
To make changes to your subscription:
http://www.postgresql.org/mailpref/pgsql-hackers

----Next_Part(Wed_Jul_06_16_29_14_2016_599)----






^ permalink  raw  reply  [nested|flat] 56+ messages in thread

* [PATCH 7/7] Make Append node async-aware.
@ 2016-06-28 09:52 Kyotaro Horiguchi <[email protected]>
  0 siblings, 0 replies; 56+ messages in thread

From: Kyotaro Horiguchi @ 2016-06-28 09:52 UTC (permalink / raw)

Change append node to be capable to handle asynchronous children
properly. As soon as it receives !async_ready from a child, it moves
to the next child and if no child is ready, it sleeps until at least
one of them become ready.
---
 src/backend/executor/nodeAppend.c | 67 +++++++++++++++++++++++++++++++++++++--
 src/include/nodes/execnodes.h     |  2 ++
 2 files changed, 67 insertions(+), 2 deletions(-)

diff --git a/src/backend/executor/nodeAppend.c b/src/backend/executor/nodeAppend.c
index e0ce8c6..9a4063a 100644
--- a/src/backend/executor/nodeAppend.c
+++ b/src/backend/executor/nodeAppend.c
@@ -58,6 +58,7 @@
 #include "postgres.h"
 
 #include "executor/execdebug.h"
+#include "executor/execAsync.h"
 #include "executor/nodeAppend.h"
 
 static bool exec_append_initialize_next(AppendState *appendstate);
@@ -121,6 +122,7 @@ ExecInitAppend(Append *node, EState *estate, int eflags)
 {
 	AppendState *appendstate = makeNode(AppendState);
 	PlanState **appendplanstates;
+	bool	   *finished;
 	int			nplans;
 	int			i;
 	ListCell   *lc;
@@ -134,14 +136,17 @@ ExecInitAppend(Append *node, EState *estate, int eflags)
 	nplans = list_length(node->appendplans);
 
 	appendplanstates = (PlanState **) palloc0(nplans * sizeof(PlanState *));
-
+	finished = (bool *) palloc0(nplans * sizeof(bool));
+	
 	/*
 	 * create new AppendState for our append node
 	 */
 	appendstate->ps.plan = (Plan *) node;
 	appendstate->ps.state = estate;
 	appendstate->appendplans = appendplanstates;
+	appendstate->as_finished = finished;
 	appendstate->as_nplans = nplans;
+	appendstate->as_async = ((eflags & EXEC_FLAG_BACKWARD) == 0);
 
 	/*
 	 * Miscellaneous initialization
@@ -194,6 +199,8 @@ ExecInitAppend(Append *node, EState *estate, int eflags)
 void
 ExecAppend(AppendState *node)
 {
+	int stopplan = node->as_whichplan;
+
 	for (;;)
 	{
 		PlanState  *subnode;
@@ -207,7 +214,36 @@ ExecAppend(AppendState *node)
 		/*
 		 * get a tuple from the subplan
 		 */
-		result = ExecProcNode(subnode);
+		Assert(!node->as_finished[node->as_whichplan]);
+
+		if (!subnode->result_ready)
+			ExecExecuteNode(subnode);
+
+		if (!subnode->result_ready)
+		{
+			if (node->as_async)
+			{
+				/* Move to the next living node */
+				do
+				{
+					node->as_whichplan = 
+						(node->as_whichplan + 1) %  node->as_nplans;
+				} while (node->as_whichplan != stopplan &&
+						 node->as_finished[node->as_whichplan]);
+
+				/* No node is ready yet, return as not-ready */
+				if (node->as_whichplan == stopplan)
+					return;
+
+				/* Try the next node */
+				continue;
+			}
+
+			/* If not async, immediately wait for this subnode */
+			ExecAsyncWaitForNode(subnode);
+		}				
+
+		result = ExecConsumeResult((PlanState *) subnode);
 
 		if (!TupIsNull(result))
 		{
@@ -220,6 +256,31 @@ ExecAppend(AppendState *node)
 			return;
 		}
 
+		if (node->as_async)
+		{
+			node->as_finished[node->as_whichplan] = true;
+			stopplan = node->as_whichplan;
+
+			/* Find the next living subnode */
+			do
+			{
+				node->as_whichplan =
+					(node->as_whichplan + 1) % node->as_nplans;
+			} while (node->as_whichplan != stopplan &&
+					 node->as_finished[node->as_whichplan]);
+
+			if (node->as_whichplan != stopplan)
+			{
+				stopplan = node->as_whichplan;
+				continue;
+			}
+
+			/* All subnodes are exhausted. Finish this node. */
+			ExecReturnTuple(&node->ps,
+							ExecClearTuple(node->ps.ps_ResultTupleSlot));
+			return;
+		}
+
 		/*
 		 * Go on to the "next" subplan in the appropriate direction. If no
 		 * more subplans, return the empty slot set up for us by
@@ -277,6 +338,8 @@ ExecReScanAppend(AppendState *node)
 	{
 		PlanState  *subnode = node->appendplans[i];
 
+		node->as_finished[i] = false;
+
 		/*
 		 * ExecReScan doesn't know about my subplans, so I have to do
 		 * changed-parameter signaling myself.
diff --git a/src/include/nodes/execnodes.h b/src/include/nodes/execnodes.h
index b72decc..b0a86c5 100644
--- a/src/include/nodes/execnodes.h
+++ b/src/include/nodes/execnodes.h
@@ -1177,6 +1177,8 @@ typedef struct AppendState
 {
 	PlanState	ps;				/* its first field is NodeTag */
 	PlanState **appendplans;	/* array of PlanStates for my inputs */
+	bool		as_async;		/* true to allow async execution */
+	bool	   *as_finished;	/* array of the running state of subplans */
 	int			as_nplans;
 	int			as_whichplan;
 } AppendState;
-- 
2.9.2


----Next_Part(Mon_Aug_29_17_08_36_2016_213)--
Content-Type: text/plain
Content-Disposition: inline
Content-Transfer-Encoding: 8bit
MIME-Version: 1.0


-- 
Sent via pgsql-hackers mailing list ([email protected])
To make changes to your subscription:
http://www.postgresql.org/mailpref/pgsql-hackers

----Next_Part(Mon_Aug_29_17_08_36_2016_213)----






^ permalink  raw  reply  [nested|flat] 56+ messages in thread

* [PATCH 7/7] Make Append node async-aware.
@ 2016-06-28 09:52 Kyotaro Horiguchi <[email protected]>
  0 siblings, 0 replies; 56+ messages in thread

From: Kyotaro Horiguchi @ 2016-06-28 09:52 UTC (permalink / raw)

Change append node to be capable to handle asynchronous children
properly. As soon as it receives !async_ready from a child, it moves
to the next child and if no child is ready, it sleeps until at least
one of them become ready.
---
 src/backend/executor/nodeAppend.c | 94 ++++++++++++++++++++++++++++++++-------
 src/include/nodes/execnodes.h     |  2 +
 2 files changed, 80 insertions(+), 16 deletions(-)

diff --git a/src/backend/executor/nodeAppend.c b/src/backend/executor/nodeAppend.c
index e0ce8c6..1c0d26e 100644
--- a/src/backend/executor/nodeAppend.c
+++ b/src/backend/executor/nodeAppend.c
@@ -58,6 +58,7 @@
 #include "postgres.h"
 
 #include "executor/execdebug.h"
+#include "executor/execAsync.h"
 #include "executor/nodeAppend.h"
 
 static bool exec_append_initialize_next(AppendState *appendstate);
@@ -121,6 +122,7 @@ ExecInitAppend(Append *node, EState *estate, int eflags)
 {
 	AppendState *appendstate = makeNode(AppendState);
 	PlanState **appendplanstates;
+	bool	   *finished;
 	int			nplans;
 	int			i;
 	ListCell   *lc;
@@ -134,14 +136,17 @@ ExecInitAppend(Append *node, EState *estate, int eflags)
 	nplans = list_length(node->appendplans);
 
 	appendplanstates = (PlanState **) palloc0(nplans * sizeof(PlanState *));
-
+	finished = (bool *) palloc0(nplans * sizeof(bool));
+	
 	/*
 	 * create new AppendState for our append node
 	 */
 	appendstate->ps.plan = (Plan *) node;
 	appendstate->ps.state = estate;
 	appendstate->appendplans = appendplanstates;
+	appendstate->as_finished = finished;
 	appendstate->as_nplans = nplans;
+	appendstate->as_async = ((eflags & EXEC_FLAG_BACKWARD) == 0);
 
 	/*
 	 * Miscellaneous initialization
@@ -194,49 +199,104 @@ ExecInitAppend(Append *node, EState *estate, int eflags)
 void
 ExecAppend(AppendState *node)
 {
-	for (;;)
+	int n_notready = 0;
+	PlanState  *subnode;
+	int i, n;
+
+	n = node->as_whichplan;
+
+	for (i = 0 ; i < node->as_nplans ; i++)
 	{
-		PlanState  *subnode;
 		TupleTableSlot *result;
 
+		if (node->as_async)
+		{
+			if (n >= node->as_nplans)
+				n = 0;
+
+			if (node->as_finished[n])
+			{
+				n++;
+				continue;
+			}
+		}
+
 		/*
 		 * figure out which subplan we are currently processing
 		 */
-		subnode = node->appendplans[node->as_whichplan];
+		subnode = node->appendplans[n];
 
 		/*
-		 * get a tuple from the subplan
+		 * execute the subplan to get a result if it is not ready yet
 		 */
-		result = ExecProcNode(subnode);
+		if (!subnode->result_ready)
+			ExecExecuteNode(subnode);
+
+		/*
+		 * This subnode is not ready yet when asynchrony is not allowed,
+		 * immediately wait for this subnode.
+		 */
+		if (!subnode->result_ready)
+		{
+			if (node->as_async)
+			{
+				n_notready++;
+				n++;
+				continue;
+			}
+			ExecAsyncWaitForNode(subnode);
+		}
+
+		Assert(subnode->result_ready);
+
+		result = ExecConsumeResult((PlanState *)subnode);
 
 		if (!TupIsNull(result))
 		{
 			/*
-			 * If the subplan gave us something then return it as-is. We do
-			 * NOT make use of the result slot that was set up in
-			 * ExecInitAppend; there's no need for it.
+			 * If the subplan gave us something then return it
+			 * as-is. We do NOT make use of the result slot that was
+			 * set up in ExecInitAppend; there's no need for it.
 			 */
+			node->as_whichplan = n;
 			ExecReturnTuple(&node->ps, result);
 			return;
 		}
 
+		/* Tuple has been exhausted on this subnode */
+		if (node->as_async)
+		{
+			node->as_finished[n++] = true;
+			continue;
+		}
+
 		/*
 		 * Go on to the "next" subplan in the appropriate direction. If no
 		 * more subplans, return the empty slot set up for us by
 		 * ExecInitAppend.
 		 */
 		if (ScanDirectionIsForward(node->ps.state->es_direction))
-			node->as_whichplan++;
+		{
+			if (n >= node->as_nplans - 1)
+				break;
+			n++;
+		}
 		else
-			node->as_whichplan--;
-		if (!exec_append_initialize_next(node))
 		{
-			ExecReturnTuple(&node->ps,
-							ExecClearTuple(node->ps.ps_ResultTupleSlot));
-			return;
+			if (n == 0)
+				break;
+			n--;
 		}
+	}
 
-		/* Else loop back and try to get a tuple from the new subplan */
+	/*
+	 * We are finished if reached here and no subnodes are not-ready
+	 */
+	if (n_notready == 0)
+	{
+		node->as_whichplan = n;
+		ExecReturnTuple(&node->ps,
+						ExecClearTuple(node->ps.ps_ResultTupleSlot));
 	}
 }
 
@@ -277,6 +337,8 @@ ExecReScanAppend(AppendState *node)
 	{
 		PlanState  *subnode = node->appendplans[i];
 
+		node->as_finished[i] = false;
+
 		/*
 		 * ExecReScan doesn't know about my subplans, so I have to do
 		 * changed-parameter signaling myself.
diff --git a/src/include/nodes/execnodes.h b/src/include/nodes/execnodes.h
index 9121537..e4f2bb6 100644
--- a/src/include/nodes/execnodes.h
+++ b/src/include/nodes/execnodes.h
@@ -1170,6 +1170,8 @@ typedef struct AppendState
 {
 	PlanState	ps;				/* its first field is NodeTag */
 	PlanState **appendplans;	/* array of PlanStates for my inputs */
+	bool		as_async;		/* true to allow async execution */
+	bool	   *as_finished;	/* array of the running state of subplans */
 	int			as_nplans;
 	int			as_whichplan;
 } AppendState;
-- 
1.8.3.1


----Next_Part(Wed_Jul_06_16_29_14_2016_599)--
Content-Type: Text/Plain; charset=us-ascii
Content-Transfer-Encoding: 7bit
Content-Disposition: inline; filename="gentbl.sql"

CREATE EXTENSION postgres_fdw;
CREATE TABLE pl (a int, b int);
CREATE TABLE cl1 (LIKE pl) INHERITS (pl);
CREATE TABLE cl2 (LIKE pl) INHERITS (pl);
CREATE TABLE cl3 (LIKE pl) INHERITS (pl);
CREATE TABLE cl4 (LIKE pl) INHERITS (pl);
INSERT INTO cl1 (SELECT a, a FROM generate_series(0000000, 0999999) a);
INSERT INTO cl2 (SELECT a, a FROM generate_series(1000000, 1999999) a);
INSERT INTO cl3 (SELECT a, a FROM generate_series(2000000, 2999999) a);
INSERT INTO cl4 (SELECT a, a FROM generate_series(3000000, 3999999) a);

CREATE TABLE t0 (LIKE pl);
INSERT INTO t0 (SELECT a, a FROM generate_series(0000000, 9999999) a);

CREATE SERVER sv0 FOREIGN DATA WRAPPER postgres_fdw OPTIONS (host '/tmp', dbname 'postgres');
CREATE SERVER sv1 FOREIGN DATA WRAPPER postgres_fdw OPTIONS (host '/tmp', dbname 'postgres');
CREATE SERVER sv2 FOREIGN DATA WRAPPER postgres_fdw OPTIONS (host '/tmp', dbname 'postgres');
CREATE SERVER sv3 FOREIGN DATA WRAPPER postgres_fdw OPTIONS (host '/tmp', dbname 'postgres');
CREATE SERVER sv4 FOREIGN DATA WRAPPER postgres_fdw OPTIONS (host '/tmp', dbname 'postgres');
CREATE USER MAPPING FOR public SERVER sv0;
CREATE USER MAPPING FOR public SERVER sv1;
CREATE USER MAPPING FOR public SERVER sv2;
CREATE USER MAPPING FOR public SERVER sv3;
CREATE USER MAPPING FOR public SERVER sv4;

CREATE FOREIGN TABLE ft10 (a int, b int) SERVER sv0 OPTIONS (table_name 'cl1');
CREATE FOREIGN TABLE ft20 (a int, b int) SERVER sv0 OPTIONS (table_name 'cl2');
CREATE FOREIGN TABLE ft30 (a int, b int) SERVER sv0 OPTIONS (table_name 'cl3');
CREATE FOREIGN TABLE ft40 (a int, b int) SERVER sv0 OPTIONS (table_name 'cl4');
CREATE FOREIGN TABLE ft11 (a int, b int) SERVER sv1 OPTIONS (table_name 'cl1');
CREATE FOREIGN TABLE ft22 (a int, b int) SERVER sv2 OPTIONS (table_name 'cl2');
CREATE FOREIGN TABLE ft33 (a int, b int) SERVER sv3 OPTIONS (table_name 'cl3');
CREATE FOREIGN TABLE ft44 (a int, b int) SERVER sv4 OPTIONS (table_name 'cl4');

CREATE TABLE pf0 (LIKE pl);
ALTER FOREIGN TABLE ft10 INHERIT pf0;
ALTER FOREIGN TABLE ft20 INHERIT pf0;
ALTER FOREIGN TABLE ft30 INHERIT pf0;
ALTER FOREIGN TABLE ft40 INHERIT pf0;

CREATE table pf1 (LIKE pl);
ALTER FOREIGN TABLE ft11 INHERIT pf1;
ALTER FOREIGN TABLE ft22 INHERIT pf1;
ALTER FOREIGN TABLE ft33 INHERIT pf1;
ALTER FOREIGN TABLE ft44 INHERIT pf1;

ANALYZE;

----Next_Part(Wed_Jul_06_16_29_14_2016_599)--
Content-Type: Text/Plain; charset=us-ascii
Content-Transfer-Encoding: 7bit
Content-Disposition: inline; filename="testrun.sh"

#! /bin/bash

function do_test() {
	echo $1
	for i in $(seq 1 10); do psql postgres -c "set log_min_duration_statement = 0; set client_min_messages=log; select sum(a) from $1"; done | grep LOG
}

for i in "t0" "pl" "pf0" "pf1";
do
  do_test $i
done


----Next_Part(Wed_Jul_06_16_29_14_2016_599)--
Content-Type: text/plain
Content-Disposition: inline
Content-Transfer-Encoding: 8bit
MIME-Version: 1.0


-- 
Sent via pgsql-hackers mailing list ([email protected])
To make changes to your subscription:
http://www.postgresql.org/mailpref/pgsql-hackers

----Next_Part(Wed_Jul_06_16_29_14_2016_599)----






^ permalink  raw  reply  [nested|flat] 56+ messages in thread

* [PATCH 7/7] Make Append node async-aware.
@ 2016-06-28 09:52 Kyotaro Horiguchi <[email protected]>
  0 siblings, 0 replies; 56+ messages in thread

From: Kyotaro Horiguchi @ 2016-06-28 09:52 UTC (permalink / raw)

Change append node to be capable to handle asynchronous children
properly. As soon as it receives !async_ready from a child, it moves
to the next child and if no child is ready, it sleeps until at least
one of them become ready.
---
 src/backend/executor/nodeAppend.c | 67 +++++++++++++++++++++++++++++++++++++--
 src/include/nodes/execnodes.h     |  2 ++
 2 files changed, 67 insertions(+), 2 deletions(-)

diff --git a/src/backend/executor/nodeAppend.c b/src/backend/executor/nodeAppend.c
index e0ce8c6..9a4063a 100644
--- a/src/backend/executor/nodeAppend.c
+++ b/src/backend/executor/nodeAppend.c
@@ -58,6 +58,7 @@
 #include "postgres.h"
 
 #include "executor/execdebug.h"
+#include "executor/execAsync.h"
 #include "executor/nodeAppend.h"
 
 static bool exec_append_initialize_next(AppendState *appendstate);
@@ -121,6 +122,7 @@ ExecInitAppend(Append *node, EState *estate, int eflags)
 {
 	AppendState *appendstate = makeNode(AppendState);
 	PlanState **appendplanstates;
+	bool	   *finished;
 	int			nplans;
 	int			i;
 	ListCell   *lc;
@@ -134,14 +136,17 @@ ExecInitAppend(Append *node, EState *estate, int eflags)
 	nplans = list_length(node->appendplans);
 
 	appendplanstates = (PlanState **) palloc0(nplans * sizeof(PlanState *));
-
+	finished = (bool *) palloc0(nplans * sizeof(bool));
+	
 	/*
 	 * create new AppendState for our append node
 	 */
 	appendstate->ps.plan = (Plan *) node;
 	appendstate->ps.state = estate;
 	appendstate->appendplans = appendplanstates;
+	appendstate->as_finished = finished;
 	appendstate->as_nplans = nplans;
+	appendstate->as_async = ((eflags & EXEC_FLAG_BACKWARD) == 0);
 
 	/*
 	 * Miscellaneous initialization
@@ -194,6 +199,8 @@ ExecInitAppend(Append *node, EState *estate, int eflags)
 void
 ExecAppend(AppendState *node)
 {
+	int stopplan = node->as_whichplan;
+
 	for (;;)
 	{
 		PlanState  *subnode;
@@ -207,7 +214,36 @@ ExecAppend(AppendState *node)
 		/*
 		 * get a tuple from the subplan
 		 */
-		result = ExecProcNode(subnode);
+		Assert(!node->as_finished[node->as_whichplan]);
+
+		if (!subnode->result_ready)
+			ExecExecuteNode(subnode);
+
+		if (!subnode->result_ready)
+		{
+			if (node->as_async)
+			{
+				/* Move to the next living node */
+				do
+				{
+					node->as_whichplan = 
+						(node->as_whichplan + 1) %  node->as_nplans;
+				} while (node->as_whichplan != stopplan &&
+						 node->as_finished[node->as_whichplan]);
+
+				/* No node is ready yet, return as not-ready */
+				if (node->as_whichplan == stopplan)
+					return;
+
+				/* Try the next node */
+				continue;
+			}
+
+			/* If not async, immediately wait for this subnode */
+			ExecAsyncWaitForNode(subnode);
+		}				
+
+		result = ExecConsumeResult((PlanState *) subnode);
 
 		if (!TupIsNull(result))
 		{
@@ -220,6 +256,31 @@ ExecAppend(AppendState *node)
 			return;
 		}
 
+		if (node->as_async)
+		{
+			node->as_finished[node->as_whichplan] = true;
+			stopplan = node->as_whichplan;
+
+			/* Find the next living subnode */
+			do
+			{
+				node->as_whichplan =
+					(node->as_whichplan + 1) % node->as_nplans;
+			} while (node->as_whichplan != stopplan &&
+					 node->as_finished[node->as_whichplan]);
+
+			if (node->as_whichplan != stopplan)
+			{
+				stopplan = node->as_whichplan;
+				continue;
+			}
+
+			/* All subnodes are exhausted. Finish this node. */
+			ExecReturnTuple(&node->ps,
+							ExecClearTuple(node->ps.ps_ResultTupleSlot));
+			return;
+		}
+
 		/*
 		 * Go on to the "next" subplan in the appropriate direction. If no
 		 * more subplans, return the empty slot set up for us by
@@ -277,6 +338,8 @@ ExecReScanAppend(AppendState *node)
 	{
 		PlanState  *subnode = node->appendplans[i];
 
+		node->as_finished[i] = false;
+
 		/*
 		 * ExecReScan doesn't know about my subplans, so I have to do
 		 * changed-parameter signaling myself.
diff --git a/src/include/nodes/execnodes.h b/src/include/nodes/execnodes.h
index b72decc..b0a86c5 100644
--- a/src/include/nodes/execnodes.h
+++ b/src/include/nodes/execnodes.h
@@ -1177,6 +1177,8 @@ typedef struct AppendState
 {
 	PlanState	ps;				/* its first field is NodeTag */
 	PlanState **appendplans;	/* array of PlanStates for my inputs */
+	bool		as_async;		/* true to allow async execution */
+	bool	   *as_finished;	/* array of the running state of subplans */
 	int			as_nplans;
 	int			as_whichplan;
 } AppendState;
-- 
1.8.3.1


----Next_Part(Thu_Jul_21_18_50_07_2016_597)--
Content-Type: text/plain
Content-Disposition: inline
Content-Transfer-Encoding: 8bit
MIME-Version: 1.0


-- 
Sent via pgsql-hackers mailing list ([email protected])
To make changes to your subscription:
http://www.postgresql.org/mailpref/pgsql-hackers

----Next_Part(Thu_Jul_21_18_50_07_2016_597)----






^ permalink  raw  reply  [nested|flat] 56+ messages in thread

* [PATCH 7/7] Make Append node async-aware.
@ 2016-06-28 09:52 Kyotaro Horiguchi <[email protected]>
  0 siblings, 0 replies; 56+ messages in thread

From: Kyotaro Horiguchi @ 2016-06-28 09:52 UTC (permalink / raw)

Change append node to be capable to handle asynchronous children
properly. As soon as it receives !async_ready from a child, it moves
to the next child and if no child is ready, it sleeps until at least
one of them become ready.
---
 src/backend/executor/nodeAppend.c | 94 ++++++++++++++++++++++++++++++++-------
 src/include/nodes/execnodes.h     |  2 +
 2 files changed, 80 insertions(+), 16 deletions(-)

diff --git a/src/backend/executor/nodeAppend.c b/src/backend/executor/nodeAppend.c
index e0ce8c6..1c0d26e 100644
--- a/src/backend/executor/nodeAppend.c
+++ b/src/backend/executor/nodeAppend.c
@@ -58,6 +58,7 @@
 #include "postgres.h"
 
 #include "executor/execdebug.h"
+#include "executor/execAsync.h"
 #include "executor/nodeAppend.h"
 
 static bool exec_append_initialize_next(AppendState *appendstate);
@@ -121,6 +122,7 @@ ExecInitAppend(Append *node, EState *estate, int eflags)
 {
 	AppendState *appendstate = makeNode(AppendState);
 	PlanState **appendplanstates;
+	bool	   *finished;
 	int			nplans;
 	int			i;
 	ListCell   *lc;
@@ -134,14 +136,17 @@ ExecInitAppend(Append *node, EState *estate, int eflags)
 	nplans = list_length(node->appendplans);
 
 	appendplanstates = (PlanState **) palloc0(nplans * sizeof(PlanState *));
-
+	finished = (bool *) palloc0(nplans * sizeof(bool));
+	
 	/*
 	 * create new AppendState for our append node
 	 */
 	appendstate->ps.plan = (Plan *) node;
 	appendstate->ps.state = estate;
 	appendstate->appendplans = appendplanstates;
+	appendstate->as_finished = finished;
 	appendstate->as_nplans = nplans;
+	appendstate->as_async = ((eflags & EXEC_FLAG_BACKWARD) == 0);
 
 	/*
 	 * Miscellaneous initialization
@@ -194,49 +199,104 @@ ExecInitAppend(Append *node, EState *estate, int eflags)
 void
 ExecAppend(AppendState *node)
 {
-	for (;;)
+	int n_notready = 0;
+	PlanState  *subnode;
+	int i, n;
+
+	n = node->as_whichplan;
+
+	for (i = 0 ; i < node->as_nplans ; i++)
 	{
-		PlanState  *subnode;
 		TupleTableSlot *result;
 
+		if (node->as_async)
+		{
+			if (n >= node->as_nplans)
+				n = 0;
+
+			if (node->as_finished[n])
+			{
+				n++;
+				continue;
+			}
+		}
+
 		/*
 		 * figure out which subplan we are currently processing
 		 */
-		subnode = node->appendplans[node->as_whichplan];
+		subnode = node->appendplans[n];
 
 		/*
-		 * get a tuple from the subplan
+		 * execute the subplan to get a result if it is not ready yet
 		 */
-		result = ExecProcNode(subnode);
+		if (!subnode->result_ready)
+			ExecExecuteNode(subnode);
+
+		/*
+		 * This subnode is not ready yet when asynchrony is not allowed,
+		 * immediately wait for this subnode.
+		 */
+		if (!subnode->result_ready)
+		{
+			if (node->as_async)
+			{
+				n_notready++;
+				n++;
+				continue;
+			}
+			ExecAsyncWaitForNode(subnode);
+		}
+
+		Assert(subnode->result_ready);
+
+		result = ExecConsumeResult((PlanState *)subnode);
 
 		if (!TupIsNull(result))
 		{
 			/*
-			 * If the subplan gave us something then return it as-is. We do
-			 * NOT make use of the result slot that was set up in
-			 * ExecInitAppend; there's no need for it.
+			 * If the subplan gave us something then return it
+			 * as-is. We do NOT make use of the result slot that was
+			 * set up in ExecInitAppend; there's no need for it.
 			 */
+			node->as_whichplan = n;
 			ExecReturnTuple(&node->ps, result);
 			return;
 		}
 
+		/* Tuple has been exhausted on this subnode */
+		if (node->as_async)
+		{
+			node->as_finished[n++] = true;
+			continue;
+		}
+
 		/*
 		 * Go on to the "next" subplan in the appropriate direction. If no
 		 * more subplans, return the empty slot set up for us by
 		 * ExecInitAppend.
 		 */
 		if (ScanDirectionIsForward(node->ps.state->es_direction))
-			node->as_whichplan++;
+		{
+			if (n >= node->as_nplans - 1)
+				break;
+			n++;
+		}
 		else
-			node->as_whichplan--;
-		if (!exec_append_initialize_next(node))
 		{
-			ExecReturnTuple(&node->ps,
-							ExecClearTuple(node->ps.ps_ResultTupleSlot));
-			return;
+			if (n == 0)
+				break;
+			n--;
 		}
+	}
 
-		/* Else loop back and try to get a tuple from the new subplan */
+	/*
+	 * We are finished if reached here and no subnodes are not-ready
+	 */
+	if (n_notready == 0)
+	{
+		node->as_whichplan = n;
+		ExecReturnTuple(&node->ps,
+						ExecClearTuple(node->ps.ps_ResultTupleSlot));
 	}
 }
 
@@ -277,6 +337,8 @@ ExecReScanAppend(AppendState *node)
 	{
 		PlanState  *subnode = node->appendplans[i];
 
+		node->as_finished[i] = false;
+
 		/*
 		 * ExecReScan doesn't know about my subplans, so I have to do
 		 * changed-parameter signaling myself.
diff --git a/src/include/nodes/execnodes.h b/src/include/nodes/execnodes.h
index 9121537..e4f2bb6 100644
--- a/src/include/nodes/execnodes.h
+++ b/src/include/nodes/execnodes.h
@@ -1170,6 +1170,8 @@ typedef struct AppendState
 {
 	PlanState	ps;				/* its first field is NodeTag */
 	PlanState **appendplans;	/* array of PlanStates for my inputs */
+	bool		as_async;		/* true to allow async execution */
+	bool	   *as_finished;	/* array of the running state of subplans */
 	int			as_nplans;
 	int			as_whichplan;
 } AppendState;
-- 
1.8.3.1


----Next_Part(Wed_Jul_06_16_29_14_2016_599)--
Content-Type: Text/Plain; charset=us-ascii
Content-Transfer-Encoding: 7bit
Content-Disposition: inline; filename="gentbl.sql"

CREATE EXTENSION postgres_fdw;
CREATE TABLE pl (a int, b int);
CREATE TABLE cl1 (LIKE pl) INHERITS (pl);
CREATE TABLE cl2 (LIKE pl) INHERITS (pl);
CREATE TABLE cl3 (LIKE pl) INHERITS (pl);
CREATE TABLE cl4 (LIKE pl) INHERITS (pl);
INSERT INTO cl1 (SELECT a, a FROM generate_series(0000000, 0999999) a);
INSERT INTO cl2 (SELECT a, a FROM generate_series(1000000, 1999999) a);
INSERT INTO cl3 (SELECT a, a FROM generate_series(2000000, 2999999) a);
INSERT INTO cl4 (SELECT a, a FROM generate_series(3000000, 3999999) a);

CREATE TABLE t0 (LIKE pl);
INSERT INTO t0 (SELECT a, a FROM generate_series(0000000, 9999999) a);

CREATE SERVER sv0 FOREIGN DATA WRAPPER postgres_fdw OPTIONS (host '/tmp', dbname 'postgres');
CREATE SERVER sv1 FOREIGN DATA WRAPPER postgres_fdw OPTIONS (host '/tmp', dbname 'postgres');
CREATE SERVER sv2 FOREIGN DATA WRAPPER postgres_fdw OPTIONS (host '/tmp', dbname 'postgres');
CREATE SERVER sv3 FOREIGN DATA WRAPPER postgres_fdw OPTIONS (host '/tmp', dbname 'postgres');
CREATE SERVER sv4 FOREIGN DATA WRAPPER postgres_fdw OPTIONS (host '/tmp', dbname 'postgres');
CREATE USER MAPPING FOR public SERVER sv0;
CREATE USER MAPPING FOR public SERVER sv1;
CREATE USER MAPPING FOR public SERVER sv2;
CREATE USER MAPPING FOR public SERVER sv3;
CREATE USER MAPPING FOR public SERVER sv4;

CREATE FOREIGN TABLE ft10 (a int, b int) SERVER sv0 OPTIONS (table_name 'cl1');
CREATE FOREIGN TABLE ft20 (a int, b int) SERVER sv0 OPTIONS (table_name 'cl2');
CREATE FOREIGN TABLE ft30 (a int, b int) SERVER sv0 OPTIONS (table_name 'cl3');
CREATE FOREIGN TABLE ft40 (a int, b int) SERVER sv0 OPTIONS (table_name 'cl4');
CREATE FOREIGN TABLE ft11 (a int, b int) SERVER sv1 OPTIONS (table_name 'cl1');
CREATE FOREIGN TABLE ft22 (a int, b int) SERVER sv2 OPTIONS (table_name 'cl2');
CREATE FOREIGN TABLE ft33 (a int, b int) SERVER sv3 OPTIONS (table_name 'cl3');
CREATE FOREIGN TABLE ft44 (a int, b int) SERVER sv4 OPTIONS (table_name 'cl4');

CREATE TABLE pf0 (LIKE pl);
ALTER FOREIGN TABLE ft10 INHERIT pf0;
ALTER FOREIGN TABLE ft20 INHERIT pf0;
ALTER FOREIGN TABLE ft30 INHERIT pf0;
ALTER FOREIGN TABLE ft40 INHERIT pf0;

CREATE table pf1 (LIKE pl);
ALTER FOREIGN TABLE ft11 INHERIT pf1;
ALTER FOREIGN TABLE ft22 INHERIT pf1;
ALTER FOREIGN TABLE ft33 INHERIT pf1;
ALTER FOREIGN TABLE ft44 INHERIT pf1;

ANALYZE;

----Next_Part(Wed_Jul_06_16_29_14_2016_599)--
Content-Type: Text/Plain; charset=us-ascii
Content-Transfer-Encoding: 7bit
Content-Disposition: inline; filename="testrun.sh"

#! /bin/bash

function do_test() {
	echo $1
	for i in $(seq 1 10); do psql postgres -c "set log_min_duration_statement = 0; set client_min_messages=log; select sum(a) from $1"; done | grep LOG
}

for i in "t0" "pl" "pf0" "pf1";
do
  do_test $i
done


----Next_Part(Wed_Jul_06_16_29_14_2016_599)--
Content-Type: text/plain
Content-Disposition: inline
Content-Transfer-Encoding: 8bit
MIME-Version: 1.0


-- 
Sent via pgsql-hackers mailing list ([email protected])
To make changes to your subscription:
http://www.postgresql.org/mailpref/pgsql-hackers

----Next_Part(Wed_Jul_06_16_29_14_2016_599)----






^ permalink  raw  reply  [nested|flat] 56+ messages in thread

* [PATCH 7/7] Make Append node async-aware.
@ 2016-06-28 09:52 Kyotaro Horiguchi <[email protected]>
  0 siblings, 0 replies; 56+ messages in thread

From: Kyotaro Horiguchi @ 2016-06-28 09:52 UTC (permalink / raw)

Change append node to be capable to handle asynchronous children
properly. As soon as it receives !async_ready from a child, it moves
to the next child and if no child is ready, it sleeps until at least
one of them become ready.
---
 src/backend/executor/nodeAppend.c | 94 ++++++++++++++++++++++++++++++++-------
 src/include/nodes/execnodes.h     |  2 +
 2 files changed, 80 insertions(+), 16 deletions(-)

diff --git a/src/backend/executor/nodeAppend.c b/src/backend/executor/nodeAppend.c
index e0ce8c6..1c0d26e 100644
--- a/src/backend/executor/nodeAppend.c
+++ b/src/backend/executor/nodeAppend.c
@@ -58,6 +58,7 @@
 #include "postgres.h"
 
 #include "executor/execdebug.h"
+#include "executor/execAsync.h"
 #include "executor/nodeAppend.h"
 
 static bool exec_append_initialize_next(AppendState *appendstate);
@@ -121,6 +122,7 @@ ExecInitAppend(Append *node, EState *estate, int eflags)
 {
 	AppendState *appendstate = makeNode(AppendState);
 	PlanState **appendplanstates;
+	bool	   *finished;
 	int			nplans;
 	int			i;
 	ListCell   *lc;
@@ -134,14 +136,17 @@ ExecInitAppend(Append *node, EState *estate, int eflags)
 	nplans = list_length(node->appendplans);
 
 	appendplanstates = (PlanState **) palloc0(nplans * sizeof(PlanState *));
-
+	finished = (bool *) palloc0(nplans * sizeof(bool));
+	
 	/*
 	 * create new AppendState for our append node
 	 */
 	appendstate->ps.plan = (Plan *) node;
 	appendstate->ps.state = estate;
 	appendstate->appendplans = appendplanstates;
+	appendstate->as_finished = finished;
 	appendstate->as_nplans = nplans;
+	appendstate->as_async = ((eflags & EXEC_FLAG_BACKWARD) == 0);
 
 	/*
 	 * Miscellaneous initialization
@@ -194,49 +199,104 @@ ExecInitAppend(Append *node, EState *estate, int eflags)
 void
 ExecAppend(AppendState *node)
 {
-	for (;;)
+	int n_notready = 0;
+	PlanState  *subnode;
+	int i, n;
+
+	n = node->as_whichplan;
+
+	for (i = 0 ; i < node->as_nplans ; i++)
 	{
-		PlanState  *subnode;
 		TupleTableSlot *result;
 
+		if (node->as_async)
+		{
+			if (n >= node->as_nplans)
+				n = 0;
+
+			if (node->as_finished[n])
+			{
+				n++;
+				continue;
+			}
+		}
+
 		/*
 		 * figure out which subplan we are currently processing
 		 */
-		subnode = node->appendplans[node->as_whichplan];
+		subnode = node->appendplans[n];
 
 		/*
-		 * get a tuple from the subplan
+		 * execute the subplan to get a result if it is not ready yet
 		 */
-		result = ExecProcNode(subnode);
+		if (!subnode->result_ready)
+			ExecExecuteNode(subnode);
+
+		/*
+		 * This subnode is not ready yet when asynchrony is not allowed,
+		 * immediately wait for this subnode.
+		 */
+		if (!subnode->result_ready)
+		{
+			if (node->as_async)
+			{
+				n_notready++;
+				n++;
+				continue;
+			}
+			ExecAsyncWaitForNode(subnode);
+		}
+
+		Assert(subnode->result_ready);
+
+		result = ExecConsumeResult((PlanState *)subnode);
 
 		if (!TupIsNull(result))
 		{
 			/*
-			 * If the subplan gave us something then return it as-is. We do
-			 * NOT make use of the result slot that was set up in
-			 * ExecInitAppend; there's no need for it.
+			 * If the subplan gave us something then return it
+			 * as-is. We do NOT make use of the result slot that was
+			 * set up in ExecInitAppend; there's no need for it.
 			 */
+			node->as_whichplan = n;
 			ExecReturnTuple(&node->ps, result);
 			return;
 		}
 
+		/* Tuple has been exhausted on this subnode */
+		if (node->as_async)
+		{
+			node->as_finished[n++] = true;
+			continue;
+		}
+
 		/*
 		 * Go on to the "next" subplan in the appropriate direction. If no
 		 * more subplans, return the empty slot set up for us by
 		 * ExecInitAppend.
 		 */
 		if (ScanDirectionIsForward(node->ps.state->es_direction))
-			node->as_whichplan++;
+		{
+			if (n >= node->as_nplans - 1)
+				break;
+			n++;
+		}
 		else
-			node->as_whichplan--;
-		if (!exec_append_initialize_next(node))
 		{
-			ExecReturnTuple(&node->ps,
-							ExecClearTuple(node->ps.ps_ResultTupleSlot));
-			return;
+			if (n == 0)
+				break;
+			n--;
 		}
+	}
 
-		/* Else loop back and try to get a tuple from the new subplan */
+	/*
+	 * We are finished if reached here and no subnodes are not-ready
+	 */
+	if (n_notready == 0)
+	{
+		node->as_whichplan = n;
+		ExecReturnTuple(&node->ps,
+						ExecClearTuple(node->ps.ps_ResultTupleSlot));
 	}
 }
 
@@ -277,6 +337,8 @@ ExecReScanAppend(AppendState *node)
 	{
 		PlanState  *subnode = node->appendplans[i];
 
+		node->as_finished[i] = false;
+
 		/*
 		 * ExecReScan doesn't know about my subplans, so I have to do
 		 * changed-parameter signaling myself.
diff --git a/src/include/nodes/execnodes.h b/src/include/nodes/execnodes.h
index 9121537..e4f2bb6 100644
--- a/src/include/nodes/execnodes.h
+++ b/src/include/nodes/execnodes.h
@@ -1170,6 +1170,8 @@ typedef struct AppendState
 {
 	PlanState	ps;				/* its first field is NodeTag */
 	PlanState **appendplans;	/* array of PlanStates for my inputs */
+	bool		as_async;		/* true to allow async execution */
+	bool	   *as_finished;	/* array of the running state of subplans */
 	int			as_nplans;
 	int			as_whichplan;
 } AppendState;
-- 
1.8.3.1


----Next_Part(Wed_Jul_06_16_29_14_2016_599)--
Content-Type: Text/Plain; charset=us-ascii
Content-Transfer-Encoding: 7bit
Content-Disposition: inline; filename="gentbl.sql"

CREATE EXTENSION postgres_fdw;
CREATE TABLE pl (a int, b int);
CREATE TABLE cl1 (LIKE pl) INHERITS (pl);
CREATE TABLE cl2 (LIKE pl) INHERITS (pl);
CREATE TABLE cl3 (LIKE pl) INHERITS (pl);
CREATE TABLE cl4 (LIKE pl) INHERITS (pl);
INSERT INTO cl1 (SELECT a, a FROM generate_series(0000000, 0999999) a);
INSERT INTO cl2 (SELECT a, a FROM generate_series(1000000, 1999999) a);
INSERT INTO cl3 (SELECT a, a FROM generate_series(2000000, 2999999) a);
INSERT INTO cl4 (SELECT a, a FROM generate_series(3000000, 3999999) a);

CREATE TABLE t0 (LIKE pl);
INSERT INTO t0 (SELECT a, a FROM generate_series(0000000, 9999999) a);

CREATE SERVER sv0 FOREIGN DATA WRAPPER postgres_fdw OPTIONS (host '/tmp', dbname 'postgres');
CREATE SERVER sv1 FOREIGN DATA WRAPPER postgres_fdw OPTIONS (host '/tmp', dbname 'postgres');
CREATE SERVER sv2 FOREIGN DATA WRAPPER postgres_fdw OPTIONS (host '/tmp', dbname 'postgres');
CREATE SERVER sv3 FOREIGN DATA WRAPPER postgres_fdw OPTIONS (host '/tmp', dbname 'postgres');
CREATE SERVER sv4 FOREIGN DATA WRAPPER postgres_fdw OPTIONS (host '/tmp', dbname 'postgres');
CREATE USER MAPPING FOR public SERVER sv0;
CREATE USER MAPPING FOR public SERVER sv1;
CREATE USER MAPPING FOR public SERVER sv2;
CREATE USER MAPPING FOR public SERVER sv3;
CREATE USER MAPPING FOR public SERVER sv4;

CREATE FOREIGN TABLE ft10 (a int, b int) SERVER sv0 OPTIONS (table_name 'cl1');
CREATE FOREIGN TABLE ft20 (a int, b int) SERVER sv0 OPTIONS (table_name 'cl2');
CREATE FOREIGN TABLE ft30 (a int, b int) SERVER sv0 OPTIONS (table_name 'cl3');
CREATE FOREIGN TABLE ft40 (a int, b int) SERVER sv0 OPTIONS (table_name 'cl4');
CREATE FOREIGN TABLE ft11 (a int, b int) SERVER sv1 OPTIONS (table_name 'cl1');
CREATE FOREIGN TABLE ft22 (a int, b int) SERVER sv2 OPTIONS (table_name 'cl2');
CREATE FOREIGN TABLE ft33 (a int, b int) SERVER sv3 OPTIONS (table_name 'cl3');
CREATE FOREIGN TABLE ft44 (a int, b int) SERVER sv4 OPTIONS (table_name 'cl4');

CREATE TABLE pf0 (LIKE pl);
ALTER FOREIGN TABLE ft10 INHERIT pf0;
ALTER FOREIGN TABLE ft20 INHERIT pf0;
ALTER FOREIGN TABLE ft30 INHERIT pf0;
ALTER FOREIGN TABLE ft40 INHERIT pf0;

CREATE table pf1 (LIKE pl);
ALTER FOREIGN TABLE ft11 INHERIT pf1;
ALTER FOREIGN TABLE ft22 INHERIT pf1;
ALTER FOREIGN TABLE ft33 INHERIT pf1;
ALTER FOREIGN TABLE ft44 INHERIT pf1;

ANALYZE;

----Next_Part(Wed_Jul_06_16_29_14_2016_599)--
Content-Type: Text/Plain; charset=us-ascii
Content-Transfer-Encoding: 7bit
Content-Disposition: inline; filename="testrun.sh"

#! /bin/bash

function do_test() {
	echo $1
	for i in $(seq 1 10); do psql postgres -c "set log_min_duration_statement = 0; set client_min_messages=log; select sum(a) from $1"; done | grep LOG
}

for i in "t0" "pl" "pf0" "pf1";
do
  do_test $i
done


----Next_Part(Wed_Jul_06_16_29_14_2016_599)--
Content-Type: text/plain
Content-Disposition: inline
Content-Transfer-Encoding: 8bit
MIME-Version: 1.0


-- 
Sent via pgsql-hackers mailing list ([email protected])
To make changes to your subscription:
http://www.postgresql.org/mailpref/pgsql-hackers

----Next_Part(Wed_Jul_06_16_29_14_2016_599)----






^ permalink  raw  reply  [nested|flat] 56+ messages in thread

* [PATCH 7/7] Make Append node async-aware.
@ 2016-06-28 09:52 Kyotaro Horiguchi <[email protected]>
  0 siblings, 0 replies; 56+ messages in thread

From: Kyotaro Horiguchi @ 2016-06-28 09:52 UTC (permalink / raw)

Change append node to be capable to handle asynchronous children
properly. As soon as it receives !async_ready from a child, it moves
to the next child and if no child is ready, it sleeps until at least
one of them become ready.
---
 src/backend/executor/nodeAppend.c | 67 +++++++++++++++++++++++++++++++++++++--
 src/include/nodes/execnodes.h     |  2 ++
 2 files changed, 67 insertions(+), 2 deletions(-)

diff --git a/src/backend/executor/nodeAppend.c b/src/backend/executor/nodeAppend.c
index e0ce8c6..9a4063a 100644
--- a/src/backend/executor/nodeAppend.c
+++ b/src/backend/executor/nodeAppend.c
@@ -58,6 +58,7 @@
 #include "postgres.h"
 
 #include "executor/execdebug.h"
+#include "executor/execAsync.h"
 #include "executor/nodeAppend.h"
 
 static bool exec_append_initialize_next(AppendState *appendstate);
@@ -121,6 +122,7 @@ ExecInitAppend(Append *node, EState *estate, int eflags)
 {
 	AppendState *appendstate = makeNode(AppendState);
 	PlanState **appendplanstates;
+	bool	   *finished;
 	int			nplans;
 	int			i;
 	ListCell   *lc;
@@ -134,14 +136,17 @@ ExecInitAppend(Append *node, EState *estate, int eflags)
 	nplans = list_length(node->appendplans);
 
 	appendplanstates = (PlanState **) palloc0(nplans * sizeof(PlanState *));
-
+	finished = (bool *) palloc0(nplans * sizeof(bool));
+	
 	/*
 	 * create new AppendState for our append node
 	 */
 	appendstate->ps.plan = (Plan *) node;
 	appendstate->ps.state = estate;
 	appendstate->appendplans = appendplanstates;
+	appendstate->as_finished = finished;
 	appendstate->as_nplans = nplans;
+	appendstate->as_async = ((eflags & EXEC_FLAG_BACKWARD) == 0);
 
 	/*
 	 * Miscellaneous initialization
@@ -194,6 +199,8 @@ ExecInitAppend(Append *node, EState *estate, int eflags)
 void
 ExecAppend(AppendState *node)
 {
+	int stopplan = node->as_whichplan;
+
 	for (;;)
 	{
 		PlanState  *subnode;
@@ -207,7 +214,36 @@ ExecAppend(AppendState *node)
 		/*
 		 * get a tuple from the subplan
 		 */
-		result = ExecProcNode(subnode);
+		Assert(!node->as_finished[node->as_whichplan]);
+
+		if (!subnode->result_ready)
+			ExecExecuteNode(subnode);
+
+		if (!subnode->result_ready)
+		{
+			if (node->as_async)
+			{
+				/* Move to the next living node */
+				do
+				{
+					node->as_whichplan = 
+						(node->as_whichplan + 1) %  node->as_nplans;
+				} while (node->as_whichplan != stopplan &&
+						 node->as_finished[node->as_whichplan]);
+
+				/* No node is ready yet, return as not-ready */
+				if (node->as_whichplan == stopplan)
+					return;
+
+				/* Try the next node */
+				continue;
+			}
+
+			/* If not async, immediately wait for this subnode */
+			ExecAsyncWaitForNode(subnode);
+		}				
+
+		result = ExecConsumeResult((PlanState *) subnode);
 
 		if (!TupIsNull(result))
 		{
@@ -220,6 +256,31 @@ ExecAppend(AppendState *node)
 			return;
 		}
 
+		if (node->as_async)
+		{
+			node->as_finished[node->as_whichplan] = true;
+			stopplan = node->as_whichplan;
+
+			/* Find the next living subnode */
+			do
+			{
+				node->as_whichplan =
+					(node->as_whichplan + 1) % node->as_nplans;
+			} while (node->as_whichplan != stopplan &&
+					 node->as_finished[node->as_whichplan]);
+
+			if (node->as_whichplan != stopplan)
+			{
+				stopplan = node->as_whichplan;
+				continue;
+			}
+
+			/* All subnodes are exhausted. Finish this node. */
+			ExecReturnTuple(&node->ps,
+							ExecClearTuple(node->ps.ps_ResultTupleSlot));
+			return;
+		}
+
 		/*
 		 * Go on to the "next" subplan in the appropriate direction. If no
 		 * more subplans, return the empty slot set up for us by
@@ -277,6 +338,8 @@ ExecReScanAppend(AppendState *node)
 	{
 		PlanState  *subnode = node->appendplans[i];
 
+		node->as_finished[i] = false;
+
 		/*
 		 * ExecReScan doesn't know about my subplans, so I have to do
 		 * changed-parameter signaling myself.
diff --git a/src/include/nodes/execnodes.h b/src/include/nodes/execnodes.h
index b72decc..b0a86c5 100644
--- a/src/include/nodes/execnodes.h
+++ b/src/include/nodes/execnodes.h
@@ -1177,6 +1177,8 @@ typedef struct AppendState
 {
 	PlanState	ps;				/* its first field is NodeTag */
 	PlanState **appendplans;	/* array of PlanStates for my inputs */
+	bool		as_async;		/* true to allow async execution */
+	bool	   *as_finished;	/* array of the running state of subplans */
 	int			as_nplans;
 	int			as_whichplan;
 } AppendState;
-- 
2.9.2


----Next_Part(Mon_Aug_29_17_08_36_2016_213)--
Content-Type: text/plain
Content-Disposition: inline
Content-Transfer-Encoding: 8bit
MIME-Version: 1.0


-- 
Sent via pgsql-hackers mailing list ([email protected])
To make changes to your subscription:
http://www.postgresql.org/mailpref/pgsql-hackers

----Next_Part(Mon_Aug_29_17_08_36_2016_213)----






^ permalink  raw  reply  [nested|flat] 56+ messages in thread

* [PATCH 7/7] Make Append node async-aware.
@ 2016-06-28 09:52 Kyotaro Horiguchi <[email protected]>
  0 siblings, 0 replies; 56+ messages in thread

From: Kyotaro Horiguchi @ 2016-06-28 09:52 UTC (permalink / raw)

Change append node to be capable to handle asynchronous children
properly. As soon as it receives !async_ready from a child, it moves
to the next child and if no child is ready, it sleeps until at least
one of them become ready.
---
 src/backend/executor/nodeAppend.c | 67 +++++++++++++++++++++++++++++++++++++--
 src/include/nodes/execnodes.h     |  2 ++
 2 files changed, 67 insertions(+), 2 deletions(-)

diff --git a/src/backend/executor/nodeAppend.c b/src/backend/executor/nodeAppend.c
index e0ce8c6..9a4063a 100644
--- a/src/backend/executor/nodeAppend.c
+++ b/src/backend/executor/nodeAppend.c
@@ -58,6 +58,7 @@
 #include "postgres.h"
 
 #include "executor/execdebug.h"
+#include "executor/execAsync.h"
 #include "executor/nodeAppend.h"
 
 static bool exec_append_initialize_next(AppendState *appendstate);
@@ -121,6 +122,7 @@ ExecInitAppend(Append *node, EState *estate, int eflags)
 {
 	AppendState *appendstate = makeNode(AppendState);
 	PlanState **appendplanstates;
+	bool	   *finished;
 	int			nplans;
 	int			i;
 	ListCell   *lc;
@@ -134,14 +136,17 @@ ExecInitAppend(Append *node, EState *estate, int eflags)
 	nplans = list_length(node->appendplans);
 
 	appendplanstates = (PlanState **) palloc0(nplans * sizeof(PlanState *));
-
+	finished = (bool *) palloc0(nplans * sizeof(bool));
+	
 	/*
 	 * create new AppendState for our append node
 	 */
 	appendstate->ps.plan = (Plan *) node;
 	appendstate->ps.state = estate;
 	appendstate->appendplans = appendplanstates;
+	appendstate->as_finished = finished;
 	appendstate->as_nplans = nplans;
+	appendstate->as_async = ((eflags & EXEC_FLAG_BACKWARD) == 0);
 
 	/*
 	 * Miscellaneous initialization
@@ -194,6 +199,8 @@ ExecInitAppend(Append *node, EState *estate, int eflags)
 void
 ExecAppend(AppendState *node)
 {
+	int stopplan = node->as_whichplan;
+
 	for (;;)
 	{
 		PlanState  *subnode;
@@ -207,7 +214,36 @@ ExecAppend(AppendState *node)
 		/*
 		 * get a tuple from the subplan
 		 */
-		result = ExecProcNode(subnode);
+		Assert(!node->as_finished[node->as_whichplan]);
+
+		if (!subnode->result_ready)
+			ExecExecuteNode(subnode);
+
+		if (!subnode->result_ready)
+		{
+			if (node->as_async)
+			{
+				/* Move to the next living node */
+				do
+				{
+					node->as_whichplan = 
+						(node->as_whichplan + 1) %  node->as_nplans;
+				} while (node->as_whichplan != stopplan &&
+						 node->as_finished[node->as_whichplan]);
+
+				/* No node is ready yet, return as not-ready */
+				if (node->as_whichplan == stopplan)
+					return;
+
+				/* Try the next node */
+				continue;
+			}
+
+			/* If not async, immediately wait for this subnode */
+			ExecAsyncWaitForNode(subnode);
+		}				
+
+		result = ExecConsumeResult((PlanState *) subnode);
 
 		if (!TupIsNull(result))
 		{
@@ -220,6 +256,31 @@ ExecAppend(AppendState *node)
 			return;
 		}
 
+		if (node->as_async)
+		{
+			node->as_finished[node->as_whichplan] = true;
+			stopplan = node->as_whichplan;
+
+			/* Find the next living subnode */
+			do
+			{
+				node->as_whichplan =
+					(node->as_whichplan + 1) % node->as_nplans;
+			} while (node->as_whichplan != stopplan &&
+					 node->as_finished[node->as_whichplan]);
+
+			if (node->as_whichplan != stopplan)
+			{
+				stopplan = node->as_whichplan;
+				continue;
+			}
+
+			/* All subnodes are exhausted. Finish this node. */
+			ExecReturnTuple(&node->ps,
+							ExecClearTuple(node->ps.ps_ResultTupleSlot));
+			return;
+		}
+
 		/*
 		 * Go on to the "next" subplan in the appropriate direction. If no
 		 * more subplans, return the empty slot set up for us by
@@ -277,6 +338,8 @@ ExecReScanAppend(AppendState *node)
 	{
 		PlanState  *subnode = node->appendplans[i];
 
+		node->as_finished[i] = false;
+
 		/*
 		 * ExecReScan doesn't know about my subplans, so I have to do
 		 * changed-parameter signaling myself.
diff --git a/src/include/nodes/execnodes.h b/src/include/nodes/execnodes.h
index b72decc..b0a86c5 100644
--- a/src/include/nodes/execnodes.h
+++ b/src/include/nodes/execnodes.h
@@ -1177,6 +1177,8 @@ typedef struct AppendState
 {
 	PlanState	ps;				/* its first field is NodeTag */
 	PlanState **appendplans;	/* array of PlanStates for my inputs */
+	bool		as_async;		/* true to allow async execution */
+	bool	   *as_finished;	/* array of the running state of subplans */
 	int			as_nplans;
 	int			as_whichplan;
 } AppendState;
-- 
1.8.3.1


----Next_Part(Thu_Jul_21_18_50_07_2016_597)--
Content-Type: text/plain
Content-Disposition: inline
Content-Transfer-Encoding: 8bit
MIME-Version: 1.0


-- 
Sent via pgsql-hackers mailing list ([email protected])
To make changes to your subscription:
http://www.postgresql.org/mailpref/pgsql-hackers

----Next_Part(Thu_Jul_21_18_50_07_2016_597)----






^ permalink  raw  reply  [nested|flat] 56+ messages in thread

* [PATCH 7/7] Make Append node async-aware.
@ 2016-06-28 09:52 Kyotaro Horiguchi <[email protected]>
  0 siblings, 0 replies; 56+ messages in thread

From: Kyotaro Horiguchi @ 2016-06-28 09:52 UTC (permalink / raw)

Change append node to be capable to handle asynchronous children
properly. As soon as it receives !async_ready from a child, it moves
to the next child and if no child is ready, it sleeps until at least
one of them become ready.
---
 src/backend/executor/nodeAppend.c | 94 ++++++++++++++++++++++++++++++++-------
 src/include/nodes/execnodes.h     |  2 +
 2 files changed, 80 insertions(+), 16 deletions(-)

diff --git a/src/backend/executor/nodeAppend.c b/src/backend/executor/nodeAppend.c
index e0ce8c6..1c0d26e 100644
--- a/src/backend/executor/nodeAppend.c
+++ b/src/backend/executor/nodeAppend.c
@@ -58,6 +58,7 @@
 #include "postgres.h"
 
 #include "executor/execdebug.h"
+#include "executor/execAsync.h"
 #include "executor/nodeAppend.h"
 
 static bool exec_append_initialize_next(AppendState *appendstate);
@@ -121,6 +122,7 @@ ExecInitAppend(Append *node, EState *estate, int eflags)
 {
 	AppendState *appendstate = makeNode(AppendState);
 	PlanState **appendplanstates;
+	bool	   *finished;
 	int			nplans;
 	int			i;
 	ListCell   *lc;
@@ -134,14 +136,17 @@ ExecInitAppend(Append *node, EState *estate, int eflags)
 	nplans = list_length(node->appendplans);
 
 	appendplanstates = (PlanState **) palloc0(nplans * sizeof(PlanState *));
-
+	finished = (bool *) palloc0(nplans * sizeof(bool));
+	
 	/*
 	 * create new AppendState for our append node
 	 */
 	appendstate->ps.plan = (Plan *) node;
 	appendstate->ps.state = estate;
 	appendstate->appendplans = appendplanstates;
+	appendstate->as_finished = finished;
 	appendstate->as_nplans = nplans;
+	appendstate->as_async = ((eflags & EXEC_FLAG_BACKWARD) == 0);
 
 	/*
 	 * Miscellaneous initialization
@@ -194,49 +199,104 @@ ExecInitAppend(Append *node, EState *estate, int eflags)
 void
 ExecAppend(AppendState *node)
 {
-	for (;;)
+	int n_notready = 0;
+	PlanState  *subnode;
+	int i, n;
+
+	n = node->as_whichplan;
+
+	for (i = 0 ; i < node->as_nplans ; i++)
 	{
-		PlanState  *subnode;
 		TupleTableSlot *result;
 
+		if (node->as_async)
+		{
+			if (n >= node->as_nplans)
+				n = 0;
+
+			if (node->as_finished[n])
+			{
+				n++;
+				continue;
+			}
+		}
+
 		/*
 		 * figure out which subplan we are currently processing
 		 */
-		subnode = node->appendplans[node->as_whichplan];
+		subnode = node->appendplans[n];
 
 		/*
-		 * get a tuple from the subplan
+		 * execute the subplan to get a result if it is not ready yet
 		 */
-		result = ExecProcNode(subnode);
+		if (!subnode->result_ready)
+			ExecExecuteNode(subnode);
+
+		/*
+		 * This subnode is not ready yet when asynchrony is not allowed,
+		 * immediately wait for this subnode.
+		 */
+		if (!subnode->result_ready)
+		{
+			if (node->as_async)
+			{
+				n_notready++;
+				n++;
+				continue;
+			}
+			ExecAsyncWaitForNode(subnode);
+		}
+
+		Assert(subnode->result_ready);
+
+		result = ExecConsumeResult((PlanState *)subnode);
 
 		if (!TupIsNull(result))
 		{
 			/*
-			 * If the subplan gave us something then return it as-is. We do
-			 * NOT make use of the result slot that was set up in
-			 * ExecInitAppend; there's no need for it.
+			 * If the subplan gave us something then return it
+			 * as-is. We do NOT make use of the result slot that was
+			 * set up in ExecInitAppend; there's no need for it.
 			 */
+			node->as_whichplan = n;
 			ExecReturnTuple(&node->ps, result);
 			return;
 		}
 
+		/* Tuple has been exhausted on this subnode */
+		if (node->as_async)
+		{
+			node->as_finished[n++] = true;
+			continue;
+		}
+
 		/*
 		 * Go on to the "next" subplan in the appropriate direction. If no
 		 * more subplans, return the empty slot set up for us by
 		 * ExecInitAppend.
 		 */
 		if (ScanDirectionIsForward(node->ps.state->es_direction))
-			node->as_whichplan++;
+		{
+			if (n >= node->as_nplans - 1)
+				break;
+			n++;
+		}
 		else
-			node->as_whichplan--;
-		if (!exec_append_initialize_next(node))
 		{
-			ExecReturnTuple(&node->ps,
-							ExecClearTuple(node->ps.ps_ResultTupleSlot));
-			return;
+			if (n == 0)
+				break;
+			n--;
 		}
+	}
 
-		/* Else loop back and try to get a tuple from the new subplan */
+	/*
+	 * We are finished if reached here and no subnodes are not-ready
+	 */
+	if (n_notready == 0)
+	{
+		node->as_whichplan = n;
+		ExecReturnTuple(&node->ps,
+						ExecClearTuple(node->ps.ps_ResultTupleSlot));
 	}
 }
 
@@ -277,6 +337,8 @@ ExecReScanAppend(AppendState *node)
 	{
 		PlanState  *subnode = node->appendplans[i];
 
+		node->as_finished[i] = false;
+
 		/*
 		 * ExecReScan doesn't know about my subplans, so I have to do
 		 * changed-parameter signaling myself.
diff --git a/src/include/nodes/execnodes.h b/src/include/nodes/execnodes.h
index 9121537..e4f2bb6 100644
--- a/src/include/nodes/execnodes.h
+++ b/src/include/nodes/execnodes.h
@@ -1170,6 +1170,8 @@ typedef struct AppendState
 {
 	PlanState	ps;				/* its first field is NodeTag */
 	PlanState **appendplans;	/* array of PlanStates for my inputs */
+	bool		as_async;		/* true to allow async execution */
+	bool	   *as_finished;	/* array of the running state of subplans */
 	int			as_nplans;
 	int			as_whichplan;
 } AppendState;
-- 
1.8.3.1


----Next_Part(Wed_Jul_06_16_29_14_2016_599)--
Content-Type: Text/Plain; charset=us-ascii
Content-Transfer-Encoding: 7bit
Content-Disposition: inline; filename="gentbl.sql"

CREATE EXTENSION postgres_fdw;
CREATE TABLE pl (a int, b int);
CREATE TABLE cl1 (LIKE pl) INHERITS (pl);
CREATE TABLE cl2 (LIKE pl) INHERITS (pl);
CREATE TABLE cl3 (LIKE pl) INHERITS (pl);
CREATE TABLE cl4 (LIKE pl) INHERITS (pl);
INSERT INTO cl1 (SELECT a, a FROM generate_series(0000000, 0999999) a);
INSERT INTO cl2 (SELECT a, a FROM generate_series(1000000, 1999999) a);
INSERT INTO cl3 (SELECT a, a FROM generate_series(2000000, 2999999) a);
INSERT INTO cl4 (SELECT a, a FROM generate_series(3000000, 3999999) a);

CREATE TABLE t0 (LIKE pl);
INSERT INTO t0 (SELECT a, a FROM generate_series(0000000, 9999999) a);

CREATE SERVER sv0 FOREIGN DATA WRAPPER postgres_fdw OPTIONS (host '/tmp', dbname 'postgres');
CREATE SERVER sv1 FOREIGN DATA WRAPPER postgres_fdw OPTIONS (host '/tmp', dbname 'postgres');
CREATE SERVER sv2 FOREIGN DATA WRAPPER postgres_fdw OPTIONS (host '/tmp', dbname 'postgres');
CREATE SERVER sv3 FOREIGN DATA WRAPPER postgres_fdw OPTIONS (host '/tmp', dbname 'postgres');
CREATE SERVER sv4 FOREIGN DATA WRAPPER postgres_fdw OPTIONS (host '/tmp', dbname 'postgres');
CREATE USER MAPPING FOR public SERVER sv0;
CREATE USER MAPPING FOR public SERVER sv1;
CREATE USER MAPPING FOR public SERVER sv2;
CREATE USER MAPPING FOR public SERVER sv3;
CREATE USER MAPPING FOR public SERVER sv4;

CREATE FOREIGN TABLE ft10 (a int, b int) SERVER sv0 OPTIONS (table_name 'cl1');
CREATE FOREIGN TABLE ft20 (a int, b int) SERVER sv0 OPTIONS (table_name 'cl2');
CREATE FOREIGN TABLE ft30 (a int, b int) SERVER sv0 OPTIONS (table_name 'cl3');
CREATE FOREIGN TABLE ft40 (a int, b int) SERVER sv0 OPTIONS (table_name 'cl4');
CREATE FOREIGN TABLE ft11 (a int, b int) SERVER sv1 OPTIONS (table_name 'cl1');
CREATE FOREIGN TABLE ft22 (a int, b int) SERVER sv2 OPTIONS (table_name 'cl2');
CREATE FOREIGN TABLE ft33 (a int, b int) SERVER sv3 OPTIONS (table_name 'cl3');
CREATE FOREIGN TABLE ft44 (a int, b int) SERVER sv4 OPTIONS (table_name 'cl4');

CREATE TABLE pf0 (LIKE pl);
ALTER FOREIGN TABLE ft10 INHERIT pf0;
ALTER FOREIGN TABLE ft20 INHERIT pf0;
ALTER FOREIGN TABLE ft30 INHERIT pf0;
ALTER FOREIGN TABLE ft40 INHERIT pf0;

CREATE table pf1 (LIKE pl);
ALTER FOREIGN TABLE ft11 INHERIT pf1;
ALTER FOREIGN TABLE ft22 INHERIT pf1;
ALTER FOREIGN TABLE ft33 INHERIT pf1;
ALTER FOREIGN TABLE ft44 INHERIT pf1;

ANALYZE;

----Next_Part(Wed_Jul_06_16_29_14_2016_599)--
Content-Type: Text/Plain; charset=us-ascii
Content-Transfer-Encoding: 7bit
Content-Disposition: inline; filename="testrun.sh"

#! /bin/bash

function do_test() {
	echo $1
	for i in $(seq 1 10); do psql postgres -c "set log_min_duration_statement = 0; set client_min_messages=log; select sum(a) from $1"; done | grep LOG
}

for i in "t0" "pl" "pf0" "pf1";
do
  do_test $i
done


----Next_Part(Wed_Jul_06_16_29_14_2016_599)--
Content-Type: text/plain
Content-Disposition: inline
Content-Transfer-Encoding: 8bit
MIME-Version: 1.0


-- 
Sent via pgsql-hackers mailing list ([email protected])
To make changes to your subscription:
http://www.postgresql.org/mailpref/pgsql-hackers

----Next_Part(Wed_Jul_06_16_29_14_2016_599)----






^ permalink  raw  reply  [nested|flat] 56+ messages in thread

* [PATCH 7/7] Make Append node async-aware.
@ 2016-06-28 09:52 Kyotaro Horiguchi <[email protected]>
  0 siblings, 0 replies; 56+ messages in thread

From: Kyotaro Horiguchi @ 2016-06-28 09:52 UTC (permalink / raw)

Change append node to be capable to handle asynchronous children
properly. As soon as it receives !async_ready from a child, it moves
to the next child and if no child is ready, it sleeps until at least
one of them become ready.
---
 src/backend/executor/nodeAppend.c | 67 +++++++++++++++++++++++++++++++++++++--
 src/include/nodes/execnodes.h     |  2 ++
 2 files changed, 67 insertions(+), 2 deletions(-)

diff --git a/src/backend/executor/nodeAppend.c b/src/backend/executor/nodeAppend.c
index e0ce8c6..9a4063a 100644
--- a/src/backend/executor/nodeAppend.c
+++ b/src/backend/executor/nodeAppend.c
@@ -58,6 +58,7 @@
 #include "postgres.h"
 
 #include "executor/execdebug.h"
+#include "executor/execAsync.h"
 #include "executor/nodeAppend.h"
 
 static bool exec_append_initialize_next(AppendState *appendstate);
@@ -121,6 +122,7 @@ ExecInitAppend(Append *node, EState *estate, int eflags)
 {
 	AppendState *appendstate = makeNode(AppendState);
 	PlanState **appendplanstates;
+	bool	   *finished;
 	int			nplans;
 	int			i;
 	ListCell   *lc;
@@ -134,14 +136,17 @@ ExecInitAppend(Append *node, EState *estate, int eflags)
 	nplans = list_length(node->appendplans);
 
 	appendplanstates = (PlanState **) palloc0(nplans * sizeof(PlanState *));
-
+	finished = (bool *) palloc0(nplans * sizeof(bool));
+	
 	/*
 	 * create new AppendState for our append node
 	 */
 	appendstate->ps.plan = (Plan *) node;
 	appendstate->ps.state = estate;
 	appendstate->appendplans = appendplanstates;
+	appendstate->as_finished = finished;
 	appendstate->as_nplans = nplans;
+	appendstate->as_async = ((eflags & EXEC_FLAG_BACKWARD) == 0);
 
 	/*
 	 * Miscellaneous initialization
@@ -194,6 +199,8 @@ ExecInitAppend(Append *node, EState *estate, int eflags)
 void
 ExecAppend(AppendState *node)
 {
+	int stopplan = node->as_whichplan;
+
 	for (;;)
 	{
 		PlanState  *subnode;
@@ -207,7 +214,36 @@ ExecAppend(AppendState *node)
 		/*
 		 * get a tuple from the subplan
 		 */
-		result = ExecProcNode(subnode);
+		Assert(!node->as_finished[node->as_whichplan]);
+
+		if (!subnode->result_ready)
+			ExecExecuteNode(subnode);
+
+		if (!subnode->result_ready)
+		{
+			if (node->as_async)
+			{
+				/* Move to the next living node */
+				do
+				{
+					node->as_whichplan = 
+						(node->as_whichplan + 1) %  node->as_nplans;
+				} while (node->as_whichplan != stopplan &&
+						 node->as_finished[node->as_whichplan]);
+
+				/* No node is ready yet, return as not-ready */
+				if (node->as_whichplan == stopplan)
+					return;
+
+				/* Try the next node */
+				continue;
+			}
+
+			/* If not async, immediately wait for this subnode */
+			ExecAsyncWaitForNode(subnode);
+		}				
+
+		result = ExecConsumeResult((PlanState *) subnode);
 
 		if (!TupIsNull(result))
 		{
@@ -220,6 +256,31 @@ ExecAppend(AppendState *node)
 			return;
 		}
 
+		if (node->as_async)
+		{
+			node->as_finished[node->as_whichplan] = true;
+			stopplan = node->as_whichplan;
+
+			/* Find the next living subnode */
+			do
+			{
+				node->as_whichplan =
+					(node->as_whichplan + 1) % node->as_nplans;
+			} while (node->as_whichplan != stopplan &&
+					 node->as_finished[node->as_whichplan]);
+
+			if (node->as_whichplan != stopplan)
+			{
+				stopplan = node->as_whichplan;
+				continue;
+			}
+
+			/* All subnodes are exhausted. Finish this node. */
+			ExecReturnTuple(&node->ps,
+							ExecClearTuple(node->ps.ps_ResultTupleSlot));
+			return;
+		}
+
 		/*
 		 * Go on to the "next" subplan in the appropriate direction. If no
 		 * more subplans, return the empty slot set up for us by
@@ -277,6 +338,8 @@ ExecReScanAppend(AppendState *node)
 	{
 		PlanState  *subnode = node->appendplans[i];
 
+		node->as_finished[i] = false;
+
 		/*
 		 * ExecReScan doesn't know about my subplans, so I have to do
 		 * changed-parameter signaling myself.
diff --git a/src/include/nodes/execnodes.h b/src/include/nodes/execnodes.h
index b72decc..b0a86c5 100644
--- a/src/include/nodes/execnodes.h
+++ b/src/include/nodes/execnodes.h
@@ -1177,6 +1177,8 @@ typedef struct AppendState
 {
 	PlanState	ps;				/* its first field is NodeTag */
 	PlanState **appendplans;	/* array of PlanStates for my inputs */
+	bool		as_async;		/* true to allow async execution */
+	bool	   *as_finished;	/* array of the running state of subplans */
 	int			as_nplans;
 	int			as_whichplan;
 } AppendState;
-- 
1.8.3.1


----Next_Part(Thu_Jul_21_18_50_07_2016_597)--
Content-Type: text/plain
Content-Disposition: inline
Content-Transfer-Encoding: 8bit
MIME-Version: 1.0


-- 
Sent via pgsql-hackers mailing list ([email protected])
To make changes to your subscription:
http://www.postgresql.org/mailpref/pgsql-hackers

----Next_Part(Thu_Jul_21_18_50_07_2016_597)----






^ permalink  raw  reply  [nested|flat] 56+ messages in thread

* [PATCH 7/7] Make Append node async-aware.
@ 2016-06-28 09:52 Kyotaro Horiguchi <[email protected]>
  0 siblings, 0 replies; 56+ messages in thread

From: Kyotaro Horiguchi @ 2016-06-28 09:52 UTC (permalink / raw)

Change append node to be capable to handle asynchronous children
properly. As soon as it receives !async_ready from a child, it moves
to the next child and if no child is ready, it sleeps until at least
one of them become ready.
---
 src/backend/executor/nodeAppend.c | 67 +++++++++++++++++++++++++++++++++++++--
 src/include/nodes/execnodes.h     |  2 ++
 2 files changed, 67 insertions(+), 2 deletions(-)

diff --git a/src/backend/executor/nodeAppend.c b/src/backend/executor/nodeAppend.c
index e0ce8c6..9a4063a 100644
--- a/src/backend/executor/nodeAppend.c
+++ b/src/backend/executor/nodeAppend.c
@@ -58,6 +58,7 @@
 #include "postgres.h"
 
 #include "executor/execdebug.h"
+#include "executor/execAsync.h"
 #include "executor/nodeAppend.h"
 
 static bool exec_append_initialize_next(AppendState *appendstate);
@@ -121,6 +122,7 @@ ExecInitAppend(Append *node, EState *estate, int eflags)
 {
 	AppendState *appendstate = makeNode(AppendState);
 	PlanState **appendplanstates;
+	bool	   *finished;
 	int			nplans;
 	int			i;
 	ListCell   *lc;
@@ -134,14 +136,17 @@ ExecInitAppend(Append *node, EState *estate, int eflags)
 	nplans = list_length(node->appendplans);
 
 	appendplanstates = (PlanState **) palloc0(nplans * sizeof(PlanState *));
-
+	finished = (bool *) palloc0(nplans * sizeof(bool));
+	
 	/*
 	 * create new AppendState for our append node
 	 */
 	appendstate->ps.plan = (Plan *) node;
 	appendstate->ps.state = estate;
 	appendstate->appendplans = appendplanstates;
+	appendstate->as_finished = finished;
 	appendstate->as_nplans = nplans;
+	appendstate->as_async = ((eflags & EXEC_FLAG_BACKWARD) == 0);
 
 	/*
 	 * Miscellaneous initialization
@@ -194,6 +199,8 @@ ExecInitAppend(Append *node, EState *estate, int eflags)
 void
 ExecAppend(AppendState *node)
 {
+	int stopplan = node->as_whichplan;
+
 	for (;;)
 	{
 		PlanState  *subnode;
@@ -207,7 +214,36 @@ ExecAppend(AppendState *node)
 		/*
 		 * get a tuple from the subplan
 		 */
-		result = ExecProcNode(subnode);
+		Assert(!node->as_finished[node->as_whichplan]);
+
+		if (!subnode->result_ready)
+			ExecExecuteNode(subnode);
+
+		if (!subnode->result_ready)
+		{
+			if (node->as_async)
+			{
+				/* Move to the next living node */
+				do
+				{
+					node->as_whichplan = 
+						(node->as_whichplan + 1) %  node->as_nplans;
+				} while (node->as_whichplan != stopplan &&
+						 node->as_finished[node->as_whichplan]);
+
+				/* No node is ready yet, return as not-ready */
+				if (node->as_whichplan == stopplan)
+					return;
+
+				/* Try the next node */
+				continue;
+			}
+
+			/* If not async, immediately wait for this subnode */
+			ExecAsyncWaitForNode(subnode);
+		}				
+
+		result = ExecConsumeResult((PlanState *) subnode);
 
 		if (!TupIsNull(result))
 		{
@@ -220,6 +256,31 @@ ExecAppend(AppendState *node)
 			return;
 		}
 
+		if (node->as_async)
+		{
+			node->as_finished[node->as_whichplan] = true;
+			stopplan = node->as_whichplan;
+
+			/* Find the next living subnode */
+			do
+			{
+				node->as_whichplan =
+					(node->as_whichplan + 1) % node->as_nplans;
+			} while (node->as_whichplan != stopplan &&
+					 node->as_finished[node->as_whichplan]);
+
+			if (node->as_whichplan != stopplan)
+			{
+				stopplan = node->as_whichplan;
+				continue;
+			}
+
+			/* All subnodes are exhausted. Finish this node. */
+			ExecReturnTuple(&node->ps,
+							ExecClearTuple(node->ps.ps_ResultTupleSlot));
+			return;
+		}
+
 		/*
 		 * Go on to the "next" subplan in the appropriate direction. If no
 		 * more subplans, return the empty slot set up for us by
@@ -277,6 +338,8 @@ ExecReScanAppend(AppendState *node)
 	{
 		PlanState  *subnode = node->appendplans[i];
 
+		node->as_finished[i] = false;
+
 		/*
 		 * ExecReScan doesn't know about my subplans, so I have to do
 		 * changed-parameter signaling myself.
diff --git a/src/include/nodes/execnodes.h b/src/include/nodes/execnodes.h
index b72decc..b0a86c5 100644
--- a/src/include/nodes/execnodes.h
+++ b/src/include/nodes/execnodes.h
@@ -1177,6 +1177,8 @@ typedef struct AppendState
 {
 	PlanState	ps;				/* its first field is NodeTag */
 	PlanState **appendplans;	/* array of PlanStates for my inputs */
+	bool		as_async;		/* true to allow async execution */
+	bool	   *as_finished;	/* array of the running state of subplans */
 	int			as_nplans;
 	int			as_whichplan;
 } AppendState;
-- 
1.8.3.1


----Next_Part(Thu_Jul_21_18_50_07_2016_597)--
Content-Type: text/plain
Content-Disposition: inline
Content-Transfer-Encoding: 8bit
MIME-Version: 1.0


-- 
Sent via pgsql-hackers mailing list ([email protected])
To make changes to your subscription:
http://www.postgresql.org/mailpref/pgsql-hackers

----Next_Part(Thu_Jul_21_18_50_07_2016_597)----






^ permalink  raw  reply  [nested|flat] 56+ messages in thread

end of thread, other threads:[~2016-06-28 09:52 UTC | newest]

Thread overview: 56+ messages (download: mbox mbox.gz follow: Atom feed)
-- links below jump to the message on this page --
2005-02-19 17:30 Re: Data loss, vacuum, transaction wrap-around Jürgen Cappel <[email protected]>
2005-02-20 04:19 ` [email protected]
2016-06-28 09:52 [PATCH 7/7] Make Append node async-aware. Kyotaro Horiguchi <[email protected]>
2016-06-28 09:52 [PATCH 7/7] Make Append node async-aware. Kyotaro Horiguchi <[email protected]>
2016-06-28 09:52 [PATCH 7/7] Make Append node async-aware. Kyotaro Horiguchi <[email protected]>
2016-06-28 09:52 [PATCH 7/7] Make Append node async-aware. Kyotaro Horiguchi <[email protected]>
2016-06-28 09:52 [PATCH 7/7] Make Append node async-aware. Kyotaro Horiguchi <[email protected]>
2016-06-28 09:52 [PATCH 7/7] Make Append node async-aware. Kyotaro Horiguchi <[email protected]>
2016-06-28 09:52 [PATCH 7/7] Make Append node async-aware. Kyotaro Horiguchi <[email protected]>
2016-06-28 09:52 [PATCH 7/7] Make Append node async-aware. Kyotaro Horiguchi <[email protected]>
2016-06-28 09:52 [PATCH 7/7] Make Append node async-aware. Kyotaro Horiguchi <[email protected]>
2016-06-28 09:52 [PATCH 7/7] Make Append node async-aware. Kyotaro Horiguchi <[email protected]>
2016-06-28 09:52 [PATCH 7/7] Make Append node async-aware. Kyotaro Horiguchi <[email protected]>
2016-06-28 09:52 [PATCH 7/7] Make Append node async-aware. Kyotaro Horiguchi <[email protected]>
2016-06-28 09:52 [PATCH 7/7] Make Append node async-aware. Kyotaro Horiguchi <[email protected]>
2016-06-28 09:52 [PATCH 7/7] Make Append node async-aware. Kyotaro Horiguchi <[email protected]>
2016-06-28 09:52 [PATCH 7/7] Make Append node async-aware. Kyotaro Horiguchi <[email protected]>
2016-06-28 09:52 [PATCH 7/7] Make Append node async-aware. Kyotaro Horiguchi <[email protected]>
2016-06-28 09:52 [PATCH 7/7] Make Append node async-aware. Kyotaro Horiguchi <[email protected]>
2016-06-28 09:52 [PATCH 7/7] Make Append node async-aware. Kyotaro Horiguchi <[email protected]>
2016-06-28 09:52 [PATCH 7/7] Make Append node async-aware. Kyotaro Horiguchi <[email protected]>
2016-06-28 09:52 [PATCH 7/7] Make Append node async-aware. Kyotaro Horiguchi <[email protected]>
2016-06-28 09:52 [PATCH 7/7] Make Append node async-aware. Kyotaro Horiguchi <[email protected]>
2016-06-28 09:52 [PATCH 7/7] Make Append node async-aware. Kyotaro Horiguchi <[email protected]>
2016-06-28 09:52 [PATCH 7/7] Make Append node async-aware. Kyotaro Horiguchi <[email protected]>
2016-06-28 09:52 [PATCH 7/7] Make Append node async-aware. Kyotaro Horiguchi <[email protected]>
2016-06-28 09:52 [PATCH 7/7] Make Append node async-aware. Kyotaro Horiguchi <[email protected]>
2016-06-28 09:52 [PATCH 7/7] Make Append node async-aware. Kyotaro Horiguchi <[email protected]>
2016-06-28 09:52 [PATCH 7/7] Make Append node async-aware. Kyotaro Horiguchi <[email protected]>
2016-06-28 09:52 [PATCH 7/7] Make Append node async-aware. Kyotaro Horiguchi <[email protected]>
2016-06-28 09:52 [PATCH 7/7] Make Append node async-aware. Kyotaro Horiguchi <[email protected]>
2016-06-28 09:52 [PATCH 7/7] Make Append node async-aware. Kyotaro Horiguchi <[email protected]>
2016-06-28 09:52 [PATCH 7/7] Make Append node async-aware. Kyotaro Horiguchi <[email protected]>
2016-06-28 09:52 [PATCH 7/7] Make Append node async-aware. Kyotaro Horiguchi <[email protected]>
2016-06-28 09:52 [PATCH 7/7] Make Append node async-aware. Kyotaro Horiguchi <[email protected]>
2016-06-28 09:52 [PATCH 7/7] Make Append node async-aware. Kyotaro Horiguchi <[email protected]>
2016-06-28 09:52 [PATCH 7/7] Make Append node async-aware. Kyotaro Horiguchi <[email protected]>
2016-06-28 09:52 [PATCH 7/7] Make Append node async-aware. Kyotaro Horiguchi <[email protected]>
2016-06-28 09:52 [PATCH 7/7] Make Append node async-aware. Kyotaro Horiguchi <[email protected]>
2016-06-28 09:52 [PATCH 7/7] Make Append node async-aware. Kyotaro Horiguchi <[email protected]>
2016-06-28 09:52 [PATCH 7/7] Make Append node async-aware. Kyotaro Horiguchi <[email protected]>
2016-06-28 09:52 [PATCH 7/7] Make Append node async-aware. Kyotaro Horiguchi <[email protected]>
2016-06-28 09:52 [PATCH 7/7] Make Append node async-aware. Kyotaro Horiguchi <[email protected]>
2016-06-28 09:52 [PATCH 7/7] Make Append node async-aware. Kyotaro Horiguchi <[email protected]>
2016-06-28 09:52 [PATCH 7/7] Make Append node async-aware. Kyotaro Horiguchi <[email protected]>
2016-06-28 09:52 [PATCH 7/7] Make Append node async-aware. Kyotaro Horiguchi <[email protected]>
2016-06-28 09:52 [PATCH 7/7] Make Append node async-aware. Kyotaro Horiguchi <[email protected]>
2016-06-28 09:52 [PATCH 7/7] Make Append node async-aware. Kyotaro Horiguchi <[email protected]>
2016-06-28 09:52 [PATCH 7/7] Make Append node async-aware. Kyotaro Horiguchi <[email protected]>
2016-06-28 09:52 [PATCH 7/7] Make Append node async-aware. Kyotaro Horiguchi <[email protected]>
2016-06-28 09:52 [PATCH 7/7] Make Append node async-aware. Kyotaro Horiguchi <[email protected]>
2016-06-28 09:52 [PATCH 7/7] Make Append node async-aware. Kyotaro Horiguchi <[email protected]>
2016-06-28 09:52 [PATCH 7/7] Make Append node async-aware. Kyotaro Horiguchi <[email protected]>
2016-06-28 09:52 [PATCH 7/7] Make Append node async-aware. Kyotaro Horiguchi <[email protected]>
2016-06-28 09:52 [PATCH 7/7] Make Append node async-aware. Kyotaro Horiguchi <[email protected]>
2016-06-28 09:52 [PATCH 7/7] Make Append node async-aware. Kyotaro Horiguchi <[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