Skip to content

Navigation Menu

Sign in
Appearance settings

Search code, repositories, users, issues, pull requests...

Provide feedback

We read every piece of feedback, and take your input very seriously.

Saved searches

Use saved searches to filter your results more quickly
Appearance settings

Commit 1804284

Browse filesBrowse files
anarazelanarazel
committed
Add parallel-aware hash joins.
Introduce parallel-aware hash joins that appear in EXPLAIN plans as Parallel Hash Join with Parallel Hash. While hash joins could already appear in parallel queries, they were previously always parallel-oblivious and had a partial subplan only on the outer side, meaning that the work of the inner subplan was duplicated in every worker. After this commit, the planner will consider using a partial subplan on the inner side too, using the Parallel Hash node to divide the work over the available CPU cores and combine its results in shared memory. If the join needs to be split into multiple batches in order to respect work_mem, then workers process different batches as much as possible and then work together on the remaining batches. The advantages of a parallel-aware hash join over a parallel-oblivious hash join used in a parallel query are that it: * avoids wasting memory on duplicated hash tables * avoids wasting disk space on duplicated batch files * divides the work of building the hash table over the CPUs One disadvantage is that there is some communication between the participating CPUs which might outweigh the benefits of parallelism in the case of small hash tables. This is avoided by the planner's existing reluctance to supply partial plans for small scans, but it may be necessary to estimate synchronization costs in future if that situation changes. Another is that outer batch 0 must be written to disk if multiple batches are required. A potential future advantage of parallel-aware hash joins is that right and full outer joins could be supported, since there is a single set of matched bits for each hashtable, but that is not yet implemented. A new GUC enable_parallel_hash is defined to control the feature, defaulting to on. Author: Thomas Munro Reviewed-By: Andres Freund, Robert Haas Tested-By: Rafia Sabih, Prabhat Sahu Discussion: https://postgr.es/m/CAEepm=2W=cOkiZxcg6qiFQP-dHUe09aqTrEMM7yJDrHMhDv_RA@mail.gmail.com https://postgr.es/m/CAEepm=37HKyJ4U6XOLi=JgfSHM3o6B-GaeO-6hkOmneTDkH+Uw@mail.gmail.com
1 parent f94eec4 commit 1804284
Copy full SHA for 1804284

File tree

Expand file treeCollapse file tree

30 files changed

+3091
-116
lines changed
Filter options
Expand file treeCollapse file tree

30 files changed

+3091
-116
lines changed

‎doc/src/sgml/config.sgml

Copy file name to clipboardExpand all lines: doc/src/sgml/config.sgml
+15Lines changed: 15 additions & 0 deletions
Original file line numberDiff line numberDiff line change
@@ -3647,6 +3647,21 @@ ANY <replaceable class="parameter">num_sync</replaceable> ( <replaceable class="
36473647
</listitem>
36483648
</varlistentry>
36493649

3650+
<varlistentry id="guc-enable-parallel-hash" xreflabel="enable_parallel_hash">
3651+
<term><varname>enable_parallel_hash</varname> (<type>boolean</type>)
3652+
<indexterm>
3653+
<primary><varname>enable_parallel_hash</varname> configuration parameter</primary>
3654+
</indexterm>
3655+
</term>
3656+
<listitem>
3657+
<para>
3658+
Enables or disables the query planner's use of hash-join plan
3659+
types with parallel hash. Has no effect if hash-join plans are not
3660+
also enabled. The default is <literal>on</literal>.
3661+
</para>
3662+
</listitem>
3663+
</varlistentry>
3664+
36503665
<varlistentry id="guc-enable-partition-wise-join" xreflabel="enable_partition_wise_join">
36513666
<term><varname>enable_partition_wise_join</varname> (<type>boolean</type>)
36523667
<indexterm>

‎doc/src/sgml/monitoring.sgml

Copy file name to clipboardExpand all lines: doc/src/sgml/monitoring.sgml
+61-1Lines changed: 61 additions & 1 deletion
Original file line numberDiff line numberDiff line change
@@ -1263,7 +1263,7 @@ postgres 27093 0.0 0.0 30096 2752 ? Ss 11:34 0:00 postgres: ser
12631263
<entry>Waiting in an extension.</entry>
12641264
</row>
12651265
<row>
1266-
<entry morerows="17"><literal>IPC</literal></entry>
1266+
<entry morerows="32"><literal>IPC</literal></entry>
12671267
<entry><literal>BgWorkerShutdown</literal></entry>
12681268
<entry>Waiting for background worker to shut down.</entry>
12691269
</row>
@@ -1279,6 +1279,66 @@ postgres 27093 0.0 0.0 30096 2752 ? Ss 11:34 0:00 postgres: ser
12791279
<entry><literal>ExecuteGather</literal></entry>
12801280
<entry>Waiting for activity from child process when executing <literal>Gather</literal> node.</entry>
12811281
</row>
1282+
<row>
1283+
<entry><literal>Hash/Batch/Allocating</literal></entry>
1284+
<entry>Waiting for an elected Parallel Hash participant to allocate a hash table.</entry>
1285+
</row>
1286+
<row>
1287+
<entry><literal>Hash/Batch/Electing</literal></entry>
1288+
<entry>Electing a Parallel Hash participant to allocate a hash table.</entry>
1289+
</row>
1290+
<row>
1291+
<entry><literal>Hash/Batch/Loading</literal></entry>
1292+
<entry>Waiting for other Parallel Hash participants to finish loading a hash table.</entry>
1293+
</row>
1294+
<row>
1295+
<entry><literal>Hash/Build/Allocating</literal></entry>
1296+
<entry>Waiting for an elected Parallel Hash participant to allocate the initial hash table.</entry>
1297+
</row>
1298+
<row>
1299+
<entry><literal>Hash/Build/Electing</literal></entry>
1300+
<entry>Electing a Parallel Hash participant to allocate the initial hash table.</entry>
1301+
</row>
1302+
<row>
1303+
<entry><literal>Hash/Build/HashingInner</literal></entry>
1304+
<entry>Waiting for other Parallel Hash participants to finish hashing the inner relation.</entry>
1305+
</row>
1306+
<row>
1307+
<entry><literal>Hash/Build/HashingOuter</literal></entry>
1308+
<entry>Waiting for other Parallel Hash participants to finish partitioning the outer relation.</entry>
1309+
</row>
1310+
<row>
1311+
<entry><literal>Hash/GrowBatches/Allocating</literal></entry>
1312+
<entry>Waiting for an elected Parallel Hash participant to allocate more batches.</entry>
1313+
</row>
1314+
<row>
1315+
<entry><literal>Hash/GrowBatches/Deciding</literal></entry>
1316+
<entry>Electing a Parallel Hash participant to decide on future batch growth.</entry>
1317+
</row>
1318+
<row>
1319+
<entry><literal>Hash/GrowBatches/Electing</literal></entry>
1320+
<entry>Electing a Parallel Hash participant to allocate more batches.</entry>
1321+
</row>
1322+
<row>
1323+
<entry><literal>Hash/GrowBatches/Finishing</literal></entry>
1324+
<entry>Waiting for an elected Parallel Hash participant to decide on future batch growth.</entry>
1325+
</row>
1326+
<row>
1327+
<entry><literal>Hash/GrowBatches/Repartitioning</literal></entry>
1328+
<entry>Waiting for other Parallel Hash participants to finishing repartitioning.</entry>
1329+
</row>
1330+
<row>
1331+
<entry><literal>Hash/GrowBuckets/Allocating</literal></entry>
1332+
<entry>Waiting for an elected Parallel Hash participant to finish allocating more buckets.</entry>
1333+
</row>
1334+
<row>
1335+
<entry><literal>Hash/GrowBuckets/Electing</literal></entry>
1336+
<entry>Electing a Parallel Hash participant to allocate more buckets.</entry>
1337+
</row>
1338+
<row>
1339+
<entry><literal>Hash/GrowBuckets/Reinserting</literal></entry>
1340+
<entry>Waiting for other Parallel Hash participants to finish inserting tuples into new buckets.</entry>
1341+
</row>
12821342
<row>
12831343
<entry><literal>LogicalSyncData</literal></entry>
12841344
<entry>Waiting for logical replication remote server to send data for initial table synchronization.</entry>

‎src/backend/executor/execParallel.c

Copy file name to clipboardExpand all lines: src/backend/executor/execParallel.c
+21Lines changed: 21 additions & 0 deletions
Original file line numberDiff line numberDiff line change
@@ -31,6 +31,7 @@
3131
#include "executor/nodeCustom.h"
3232
#include "executor/nodeForeignscan.h"
3333
#include "executor/nodeHash.h"
34+
#include "executor/nodeHashjoin.h"
3435
#include "executor/nodeIndexscan.h"
3536
#include "executor/nodeIndexonlyscan.h"
3637
#include "executor/nodeSeqscan.h"
@@ -266,6 +267,11 @@ ExecParallelEstimate(PlanState *planstate, ExecParallelEstimateContext *e)
266267
ExecBitmapHeapEstimate((BitmapHeapScanState *) planstate,
267268
e->pcxt);
268269
break;
270+
case T_HashJoinState:
271+
if (planstate->plan->parallel_aware)
272+
ExecHashJoinEstimate((HashJoinState *) planstate,
273+
e->pcxt);
274+
break;
269275
case T_HashState:
270276
/* even when not parallel-aware, for EXPLAIN ANALYZE */
271277
ExecHashEstimate((HashState *) planstate, e->pcxt);
@@ -474,6 +480,11 @@ ExecParallelInitializeDSM(PlanState *planstate,
474480
ExecBitmapHeapInitializeDSM((BitmapHeapScanState *) planstate,
475481
d->pcxt);
476482
break;
483+
case T_HashJoinState:
484+
if (planstate->plan->parallel_aware)
485+
ExecHashJoinInitializeDSM((HashJoinState *) planstate,
486+
d->pcxt);
487+
break;
477488
case T_HashState:
478489
/* even when not parallel-aware, for EXPLAIN ANALYZE */
479490
ExecHashInitializeDSM((HashState *) planstate, d->pcxt);
@@ -898,6 +909,11 @@ ExecParallelReInitializeDSM(PlanState *planstate,
898909
ExecBitmapHeapReInitializeDSM((BitmapHeapScanState *) planstate,
899910
pcxt);
900911
break;
912+
case T_HashJoinState:
913+
if (planstate->plan->parallel_aware)
914+
ExecHashJoinReInitializeDSM((HashJoinState *) planstate,
915+
pcxt);
916+
break;
901917
case T_HashState:
902918
case T_SortState:
903919
/* these nodes have DSM state, but no reinitialization is required */
@@ -1196,6 +1212,11 @@ ExecParallelInitializeWorker(PlanState *planstate, ParallelWorkerContext *pwcxt)
11961212
ExecBitmapHeapInitializeWorker((BitmapHeapScanState *) planstate,
11971213
pwcxt);
11981214
break;
1215+
case T_HashJoinState:
1216+
if (planstate->plan->parallel_aware)
1217+
ExecHashJoinInitializeWorker((HashJoinState *) planstate,
1218+
pwcxt);
1219+
break;
11991220
case T_HashState:
12001221
/* even when not parallel-aware, for EXPLAIN ANALYZE */
12011222
ExecHashInitializeWorker((HashState *) planstate, pwcxt);

‎src/backend/executor/execProcnode.c

Copy file name to clipboardExpand all lines: src/backend/executor/execProcnode.c
+3Lines changed: 3 additions & 0 deletions
Original file line numberDiff line numberDiff line change
@@ -770,6 +770,9 @@ ExecShutdownNode(PlanState *node)
770770
case T_HashState:
771771
ExecShutdownHash((HashState *) node);
772772
break;
773+
case T_HashJoinState:
774+
ExecShutdownHashJoin((HashJoinState *) node);
775+
break;
773776
default:
774777
break;
775778
}

0 commit comments

Comments
0 (0)
Morty Proxy This is a proxified and sanitized view of the page, visit original site.