Merge pull request #4196 from citusdata/support-explain-analyze-wal

Support EXPLAIN(ANALYZE, WAL)

src/backend/distributed/planner/multi_explain.c

@@ -83,6 +83,7 @@ typedef struct
 	bool verbose;
 	bool costs;
 	bool buffers;
+	bool wal;
 	bool timing;
 	bool summary;
 	ExplainFormat format;
@@ -91,7 +92,7 @@ typedef struct
 
 /* EXPLAIN flags of current distributed explain */
 static ExplainOptions CurrentDistributedQueryExplainOptions = {
-	0, 0, 0, 0, 0, EXPLAIN_FORMAT_TEXT
+	0, 0, 0, 0, 0, 0, EXPLAIN_FORMAT_TEXT
 };
 
 /* Result for a single remote EXPLAIN command */
@@ -904,12 +905,18 @@ BuildRemoteExplainQuery(char *queryString, ExplainState *es)
 
 	appendStringInfo(explainQuery,
 					 "EXPLAIN (ANALYZE %s, VERBOSE %s, "
-					 "COSTS %s, BUFFERS %s, TIMING %s, SUMMARY %s, "
+					 "COSTS %s, BUFFERS %s, "
-					 "FORMAT %s) %s",
+#if PG_VERSION_NUM >= PG_VERSION_13
+					 "WAL %s, "
+#endif
+					 "TIMING %s, SUMMARY %s, FORMAT %s) %s",
 					 es->analyze ? "TRUE" : "FALSE",
 					 es->verbose ? "TRUE" : "FALSE",
 					 es->costs ? "TRUE" : "FALSE",
 					 es->buffers ? "TRUE" : "FALSE",
+#if PG_VERSION_NUM >= PG_VERSION_13
+					 es->wal ? "TRUE" : "FALSE",
+#endif
 					 es->timing ? "TRUE" : "FALSE",
 					 es->summary ? "TRUE" : "FALSE",
 					 formatStr,
@@ -1005,6 +1012,9 @@ worker_save_query_explain_analyze(PG_FUNCTION_ARGS)
 
 	/* use the same defaults as NewExplainState() for following options */
 	es->buffers = ExtractFieldBoolean(explainOptions, "buffers", es->buffers);
+#if PG_VERSION_NUM >= PG_VERSION_13
+	es->wal = ExtractFieldBoolean(explainOptions, "wal", es->wal);
+#endif
 	es->costs = ExtractFieldBoolean(explainOptions, "costs", es->costs);
 	es->summary = ExtractFieldBoolean(explainOptions, "summary", es->summary);
 	es->verbose = ExtractFieldBoolean(explainOptions, "verbose", es->verbose);
@@ -1206,6 +1216,9 @@ CitusExplainOneQuery(Query *query, int cursorOptions, IntoClause *into,
 	/* save the flags of current EXPLAIN command */
 	CurrentDistributedQueryExplainOptions.costs = es->costs;
 	CurrentDistributedQueryExplainOptions.buffers = es->buffers;
+#if PG_VERSION_NUM >= PG_VERSION_13
+	CurrentDistributedQueryExplainOptions.wal = es->wal;
+#endif
 	CurrentDistributedQueryExplainOptions.verbose = es->verbose;
 	CurrentDistributedQueryExplainOptions.summary = es->summary;
 	CurrentDistributedQueryExplainOptions.timing = es->timing;
@@ -1482,11 +1495,18 @@ WrapQueryForExplainAnalyze(const char *queryString, TupleDesc tupleDesc)
 	}
 
 	StringInfo explainOptions = makeStringInfo();
-	appendStringInfo(explainOptions, "{\"verbose\": %s, \"costs\": %s, \"buffers\": %s, "
+	appendStringInfo(explainOptions,
-									 "\"timing\": %s, \"summary\": %s, \"format\": \"%s\"}",
+					 "{\"verbose\": %s, \"costs\": %s, \"buffers\": %s, "
+#if PG_VERSION_NUM >= PG_VERSION_13
+					 "\"wal\": %s, "
+#endif
+					 "\"timing\": %s, \"summary\": %s, \"format\": \"%s\"}",
 					 CurrentDistributedQueryExplainOptions.verbose ? "true" : "false",
 					 CurrentDistributedQueryExplainOptions.costs ? "true" : "false",
 					 CurrentDistributedQueryExplainOptions.buffers ? "true" : "false",
+#if PG_VERSION_NUM >= PG_VERSION_13
+					 CurrentDistributedQueryExplainOptions.wal ? "true" : "false",
+#endif
 					 CurrentDistributedQueryExplainOptions.timing ? "true" : "false",
 					 CurrentDistributedQueryExplainOptions.summary ? "true" : "false",
 					 ExplainFormatStr(CurrentDistributedQueryExplainOptions.format));
@@ -1632,7 +1652,10 @@ ExplainWorkerPlan(PlannedStmt *plannedstmt, DestReceiver *dest, ExplainState *es
 
 	if (es->buffers)
 		instrument_option |= INSTRUMENT_BUFFERS;
-
+#if PG_VERSION_NUM >= PG_VERSION_13
+	if (es->wal)
+		instrument_option |= INSTRUMENT_WAL;
+#endif
 	/*
 	 * We always collect timing for the entire statement, even when node-level
 	 * timing is off, so we don't look at es->timing here.  (We could skip

src/test/regress/expected/pg13.out

@@ -170,15 +170,97 @@ SELECT create_distributed_table('test_table', 'a');
 -- we currently don't support this
 CREATE INDEX test_table_index ON test_table USING gist (b tsvector_ops(siglen = 100));
 ERROR:  citus currently doesn't support operator class parameters in indexes
+-- testing WAL
+CREATE TABLE test_wal(a int, b int);
+-- test WAL without ANALYZE, this should raise an error
+EXPLAIN (WAL) INSERT INTO test_wal VALUES(1,11);
+ERROR:  EXPLAIN option WAL requires ANALYZE
+-- test WAL working properly for router queries
+EXPLAIN (ANALYZE TRUE, WAL TRUE, COSTS FALSE, SUMMARY FALSE, BUFFERS FALSE, TIMING FALSE)
+INSERT INTO test_wal VALUES(1,11);
+                 QUERY PLAN
+---------------------------------------------------------------------
+ Insert on test_wal (actual rows=0 loops=1)
+   WAL: records=1 bytes=63
+   ->  Result (actual rows=1 loops=1)
+(3 rows)
+
+SELECT create_distributed_table('test_wal', 'a');
+NOTICE:  Copying data from local table...
+NOTICE:  copying the data has completed
+DETAIL:  The local data in the table is no longer visible, but is still on disk.
+HINT:  To remove the local data, run: SELECT truncate_local_data_after_distributing_table($$test_pg13.test_wal$$)
+ create_distributed_table
+---------------------------------------------------------------------
+
+(1 row)
+
+EXPLAIN (ANALYZE TRUE, WAL TRUE, COSTS FALSE, SUMMARY FALSE, BUFFERS FALSE, TIMING FALSE)
+INSERT INTO test_wal VALUES(2,22);
+                          QUERY PLAN
+---------------------------------------------------------------------
+ Custom Scan (Citus Adaptive) (actual rows=0 loops=1)
+   Task Count: 1
+   Tasks Shown: All
+   ->  Task
+         Node: host=localhost port=xxxxx dbname=regression
+         ->  Insert on test_wal_65013 (actual rows=0 loops=1)
+               WAL: records=1 bytes=63
+               ->  Result (actual rows=1 loops=1)
+(8 rows)
+
+-- Test WAL working for multi-shard query
+SET citus.explain_all_tasks TO on;
+EXPLAIN (ANALYZE TRUE, WAL TRUE, COSTS FALSE, SUMMARY FALSE, BUFFERS FALSE, TIMING FALSE)
+INSERT INTO test_wal VALUES(3,33),(4,44),(5,55) RETURNING *;
+                                   QUERY PLAN
+---------------------------------------------------------------------
+ Custom Scan (Citus Adaptive) (actual rows=3 loops=1)
+   Task Count: 1
+   Tuple data received from nodes: 9 bytes
+   Tasks Shown: All
+   ->  Task
+         Tuple data received from node: 9 bytes
+         Node: host=localhost port=xxxxx dbname=regression
+         ->  Insert on test_wal_65012 citus_table_alias (actual rows=3 loops=1)
+               WAL: records=3 bytes=189
+               ->  Values Scan on "*VALUES*" (actual rows=3 loops=1)
+(10 rows)
+
+-- make sure WAL works in distributed subplans
+EXPLAIN (ANALYZE TRUE, WAL TRUE, COSTS FALSE, SUMMARY FALSE, BUFFERS FALSE, TIMING FALSE)
+WITH cte_1 AS (INSERT INTO test_wal VALUES(6,66),(7,77),(8,88) RETURNING *)
+SELECT * FROM cte_1;
+                                            QUERY PLAN
+---------------------------------------------------------------------
+ Custom Scan (Citus Adaptive) (actual rows=3 loops=1)
+   ->  Distributed Subplan XXX_1
+         Intermediate Data Size: 54 bytes
+         Result destination: Write locally
+         ->  Custom Scan (Citus Adaptive) (actual rows=3 loops=1)
+               Task Count: 2
+               Tuple data received from nodes: 9 bytes
+               Tasks Shown: All
+               ->  Task
+                     Tuple data received from node: 6 bytes
+                     Node: host=localhost port=xxxxx dbname=regression
+                     ->  Insert on test_wal_65012 citus_table_alias (actual rows=2 loops=1)
+                           WAL: records=2 bytes=126
+                           ->  Values Scan on "*VALUES*" (actual rows=2 loops=1)
+               ->  Task
+                     Tuple data received from node: 3 bytes
+                     Node: host=localhost port=xxxxx dbname=regression
+                     ->  Insert on test_wal_65013 citus_table_alias (actual rows=1 loops=1)
+                           WAL: records=1 bytes=63
+                           ->  Result (actual rows=1 loops=1)
+   Task Count: 1
+   Tuple data received from nodes: 9 bytes
+   Tasks Shown: All
+   ->  Task
+         Tuple data received from node: 9 bytes
+         Node: host=localhost port=xxxxx dbname=regression
+         ->  Function Scan on read_intermediate_result intermediate_result (actual rows=3 loops=1)
+(27 rows)
+
+SET client_min_messages TO WARNING;
 drop schema test_pg13 cascade;
-NOTICE:  drop cascades to 10 other objects
-DETAIL:  drop cascades to table dist_table
-drop cascades to table generated_col_table
-drop cascades to view v
-drop cascades to table ab
-drop cascades to table text_table
-drop cascades to function myvarcharout(myvarchar)
-drop cascades to type myvarchar
-drop cascades to function myvarcharin(cstring,oid,integer)
-drop cascades to table my_table
-drop cascades to table test_table

src/test/regress/sql/pg13.sql

@@ -94,4 +94,26 @@ SELECT create_distributed_table('test_table', 'a');
 -- we currently don't support this
 CREATE INDEX test_table_index ON test_table USING gist (b tsvector_ops(siglen = 100));
 
+-- testing WAL
+CREATE TABLE test_wal(a int, b int);
+-- test WAL without ANALYZE, this should raise an error
+EXPLAIN (WAL) INSERT INTO test_wal VALUES(1,11);
+-- test WAL working properly for router queries
+EXPLAIN (ANALYZE TRUE, WAL TRUE, COSTS FALSE, SUMMARY FALSE, BUFFERS FALSE, TIMING FALSE)
+INSERT INTO test_wal VALUES(1,11);
+SELECT create_distributed_table('test_wal', 'a');
+EXPLAIN (ANALYZE TRUE, WAL TRUE, COSTS FALSE, SUMMARY FALSE, BUFFERS FALSE, TIMING FALSE)
+INSERT INTO test_wal VALUES(2,22);
+
+-- Test WAL working for multi-shard query
+SET citus.explain_all_tasks TO on;
+EXPLAIN (ANALYZE TRUE, WAL TRUE, COSTS FALSE, SUMMARY FALSE, BUFFERS FALSE, TIMING FALSE)
+INSERT INTO test_wal VALUES(3,33),(4,44),(5,55) RETURNING *;
+
+-- make sure WAL works in distributed subplans
+EXPLAIN (ANALYZE TRUE, WAL TRUE, COSTS FALSE, SUMMARY FALSE, BUFFERS FALSE, TIMING FALSE)
+WITH cte_1 AS (INSERT INTO test_wal VALUES(6,66),(7,77),(8,88) RETURNING *)
+SELECT * FROM cte_1;
+
+SET client_min_messages TO WARNING;
 drop schema test_pg13 cascade;