Patch pgaudit to disable logging in parallel workers (#12325)

We want to turn logging in parallel workers off to reduce log
amplification in queries which use parallel workers.

Part-of: https://github.com/neondatabase/cloud/issues/28483

Signed-off-by: Tristan Partin <tristan.partin@databricks.com>
This commit is contained in:
Tristan Partin
2025-07-02 14:54:47 -05:00
committed by GitHub
parent a9a51c038b
commit 8b4fbefc29
5 changed files with 575 additions and 1 deletions

View File

@@ -0,0 +1,143 @@
commit 7220bb3a3f23fa27207d77562dcc286f9a123313
Author: Tristan Partin <tristan.partin@databricks.com>
Date: 2025-06-23 02:09:31 +0000
Disable logging in parallel workers
When a query uses parallel workers, pgaudit will log the same query for
every parallel worker. This is undesireable since it can result in log
amplification for queries that use parallel workers.
Signed-off-by: Tristan Partin <tristan.partin@databricks.com>
diff --git a/expected/pgaudit.out b/expected/pgaudit.out
index baa8011..a601375 100644
--- a/expected/pgaudit.out
+++ b/expected/pgaudit.out
@@ -2563,6 +2563,37 @@ COMMIT;
NOTICE: AUDIT: SESSION,12,4,MISC,COMMIT,,,COMMIT;,<not logged>
DROP TABLE part_test;
NOTICE: AUDIT: SESSION,13,1,DDL,DROP TABLE,,,DROP TABLE part_test;,<not logged>
+--
+-- Test logging in parallel workers
+SET pgaudit.log = 'read';
+SET pgaudit.log_client = on;
+SET pgaudit.log_level = 'notice';
+-- Force parallel execution for testing
+SET max_parallel_workers_per_gather = 2;
+SET parallel_tuple_cost = 0;
+SET parallel_setup_cost = 0;
+SET min_parallel_table_scan_size = 0;
+SET min_parallel_index_scan_size = 0;
+-- Create table with enough data to trigger parallel execution
+CREATE TABLE parallel_test (id int, data text);
+INSERT INTO parallel_test SELECT generate_series(1, 1000), 'test data';
+SELECT count(*) FROM parallel_test;
+NOTICE: AUDIT: SESSION,14,1,READ,SELECT,,,SELECT count(*) FROM parallel_test;,<not logged>
+ count
+-------
+ 1000
+(1 row)
+
+-- Cleanup parallel test
+DROP TABLE parallel_test;
+RESET max_parallel_workers_per_gather;
+RESET parallel_tuple_cost;
+RESET parallel_setup_cost;
+RESET min_parallel_table_scan_size;
+RESET min_parallel_index_scan_size;
+RESET pgaudit.log;
+RESET pgaudit.log_client;
+RESET pgaudit.log_level;
-- Cleanup
-- Set client_min_messages up to warning to avoid noise
SET client_min_messages = 'warning';
diff --git a/pgaudit.c b/pgaudit.c
index 5e6fd38..ac9ded2 100644
--- a/pgaudit.c
+++ b/pgaudit.c
@@ -11,6 +11,7 @@
#include "postgres.h"
#include "access/htup_details.h"
+#include "access/parallel.h"
#include "access/sysattr.h"
#include "access/xact.h"
#include "access/relation.h"
@@ -1303,7 +1304,7 @@ pgaudit_ExecutorStart_hook(QueryDesc *queryDesc, int eflags)
{
AuditEventStackItem *stackItem = NULL;
- if (!internalStatement)
+ if (!internalStatement && !IsParallelWorker())
{
/* Push the audit even onto the stack */
stackItem = stack_push();
@@ -1384,7 +1385,7 @@ pgaudit_ExecutorCheckPerms_hook(List *rangeTabls, bool abort)
/* Log DML if the audit role is valid or session logging is enabled */
if ((auditOid != InvalidOid || auditLogBitmap != 0) &&
- !IsAbortedTransactionBlockState())
+ !IsAbortedTransactionBlockState() && !IsParallelWorker())
{
/* If auditLogRows is on, wait for rows processed to be set */
if (auditLogRows && auditEventStack != NULL)
@@ -1438,7 +1439,7 @@ pgaudit_ExecutorRun_hook(QueryDesc *queryDesc, ScanDirection direction, uint64 c
else
standard_ExecutorRun(queryDesc, direction, count, execute_once);
- if (auditLogRows && !internalStatement)
+ if (auditLogRows && !internalStatement && !IsParallelWorker())
{
/* Find an item from the stack by the query memory context */
stackItem = stack_find_context(queryDesc->estate->es_query_cxt);
@@ -1458,7 +1459,7 @@ pgaudit_ExecutorEnd_hook(QueryDesc *queryDesc)
AuditEventStackItem *stackItem = NULL;
AuditEventStackItem *auditEventStackFull = NULL;
- if (auditLogRows && !internalStatement)
+ if (auditLogRows && !internalStatement && !IsParallelWorker())
{
/* Find an item from the stack by the query memory context */
stackItem = stack_find_context(queryDesc->estate->es_query_cxt);
diff --git a/sql/pgaudit.sql b/sql/pgaudit.sql
index cc1374a..1870a60 100644
--- a/sql/pgaudit.sql
+++ b/sql/pgaudit.sql
@@ -1612,6 +1612,36 @@ COMMIT;
DROP TABLE part_test;
+--
+-- Test logging in parallel workers
+SET pgaudit.log = 'read';
+SET pgaudit.log_client = on;
+SET pgaudit.log_level = 'notice';
+
+-- Force parallel execution for testing
+SET max_parallel_workers_per_gather = 2;
+SET parallel_tuple_cost = 0;
+SET parallel_setup_cost = 0;
+SET min_parallel_table_scan_size = 0;
+SET min_parallel_index_scan_size = 0;
+
+-- Create table with enough data to trigger parallel execution
+CREATE TABLE parallel_test (id int, data text);
+INSERT INTO parallel_test SELECT generate_series(1, 1000), 'test data';
+
+SELECT count(*) FROM parallel_test;
+
+-- Cleanup parallel test
+DROP TABLE parallel_test;
+RESET max_parallel_workers_per_gather;
+RESET parallel_tuple_cost;
+RESET parallel_setup_cost;
+RESET min_parallel_table_scan_size;
+RESET min_parallel_index_scan_size;
+RESET pgaudit.log;
+RESET pgaudit.log_client;
+RESET pgaudit.log_level;
+
-- Cleanup
-- Set client_min_messages up to warning to avoid noise
SET client_min_messages = 'warning';

View File

@@ -0,0 +1,143 @@
commit 29dc2847f6255541992f18faf8a815dfab79631a
Author: Tristan Partin <tristan.partin@databricks.com>
Date: 2025-06-23 02:09:31 +0000
Disable logging in parallel workers
When a query uses parallel workers, pgaudit will log the same query for
every parallel worker. This is undesireable since it can result in log
amplification for queries that use parallel workers.
Signed-off-by: Tristan Partin <tristan.partin@databricks.com>
diff --git a/expected/pgaudit.out b/expected/pgaudit.out
index b22560b..73f0327 100644
--- a/expected/pgaudit.out
+++ b/expected/pgaudit.out
@@ -2563,6 +2563,37 @@ COMMIT;
NOTICE: AUDIT: SESSION,12,4,MISC,COMMIT,,,COMMIT;,<not logged>
DROP TABLE part_test;
NOTICE: AUDIT: SESSION,13,1,DDL,DROP TABLE,,,DROP TABLE part_test;,<not logged>
+--
+-- Test logging in parallel workers
+SET pgaudit.log = 'read';
+SET pgaudit.log_client = on;
+SET pgaudit.log_level = 'notice';
+-- Force parallel execution for testing
+SET max_parallel_workers_per_gather = 2;
+SET parallel_tuple_cost = 0;
+SET parallel_setup_cost = 0;
+SET min_parallel_table_scan_size = 0;
+SET min_parallel_index_scan_size = 0;
+-- Create table with enough data to trigger parallel execution
+CREATE TABLE parallel_test (id int, data text);
+INSERT INTO parallel_test SELECT generate_series(1, 1000), 'test data';
+SELECT count(*) FROM parallel_test;
+NOTICE: AUDIT: SESSION,14,1,READ,SELECT,,,SELECT count(*) FROM parallel_test;,<not logged>
+ count
+-------
+ 1000
+(1 row)
+
+-- Cleanup parallel test
+DROP TABLE parallel_test;
+RESET max_parallel_workers_per_gather;
+RESET parallel_tuple_cost;
+RESET parallel_setup_cost;
+RESET min_parallel_table_scan_size;
+RESET min_parallel_index_scan_size;
+RESET pgaudit.log;
+RESET pgaudit.log_client;
+RESET pgaudit.log_level;
-- Cleanup
-- Set client_min_messages up to warning to avoid noise
SET client_min_messages = 'warning';
diff --git a/pgaudit.c b/pgaudit.c
index 5e6fd38..ac9ded2 100644
--- a/pgaudit.c
+++ b/pgaudit.c
@@ -11,6 +11,7 @@
#include "postgres.h"
#include "access/htup_details.h"
+#include "access/parallel.h"
#include "access/sysattr.h"
#include "access/xact.h"
#include "access/relation.h"
@@ -1303,7 +1304,7 @@ pgaudit_ExecutorStart_hook(QueryDesc *queryDesc, int eflags)
{
AuditEventStackItem *stackItem = NULL;
- if (!internalStatement)
+ if (!internalStatement && !IsParallelWorker())
{
/* Push the audit even onto the stack */
stackItem = stack_push();
@@ -1384,7 +1385,7 @@ pgaudit_ExecutorCheckPerms_hook(List *rangeTabls, bool abort)
/* Log DML if the audit role is valid or session logging is enabled */
if ((auditOid != InvalidOid || auditLogBitmap != 0) &&
- !IsAbortedTransactionBlockState())
+ !IsAbortedTransactionBlockState() && !IsParallelWorker())
{
/* If auditLogRows is on, wait for rows processed to be set */
if (auditLogRows && auditEventStack != NULL)
@@ -1438,7 +1439,7 @@ pgaudit_ExecutorRun_hook(QueryDesc *queryDesc, ScanDirection direction, uint64 c
else
standard_ExecutorRun(queryDesc, direction, count, execute_once);
- if (auditLogRows && !internalStatement)
+ if (auditLogRows && !internalStatement && !IsParallelWorker())
{
/* Find an item from the stack by the query memory context */
stackItem = stack_find_context(queryDesc->estate->es_query_cxt);
@@ -1458,7 +1459,7 @@ pgaudit_ExecutorEnd_hook(QueryDesc *queryDesc)
AuditEventStackItem *stackItem = NULL;
AuditEventStackItem *auditEventStackFull = NULL;
- if (auditLogRows && !internalStatement)
+ if (auditLogRows && !internalStatement && !IsParallelWorker())
{
/* Find an item from the stack by the query memory context */
stackItem = stack_find_context(queryDesc->estate->es_query_cxt);
diff --git a/sql/pgaudit.sql b/sql/pgaudit.sql
index 8052426..7f0667b 100644
--- a/sql/pgaudit.sql
+++ b/sql/pgaudit.sql
@@ -1612,6 +1612,36 @@ COMMIT;
DROP TABLE part_test;
+--
+-- Test logging in parallel workers
+SET pgaudit.log = 'read';
+SET pgaudit.log_client = on;
+SET pgaudit.log_level = 'notice';
+
+-- Force parallel execution for testing
+SET max_parallel_workers_per_gather = 2;
+SET parallel_tuple_cost = 0;
+SET parallel_setup_cost = 0;
+SET min_parallel_table_scan_size = 0;
+SET min_parallel_index_scan_size = 0;
+
+-- Create table with enough data to trigger parallel execution
+CREATE TABLE parallel_test (id int, data text);
+INSERT INTO parallel_test SELECT generate_series(1, 1000), 'test data';
+
+SELECT count(*) FROM parallel_test;
+
+-- Cleanup parallel test
+DROP TABLE parallel_test;
+RESET max_parallel_workers_per_gather;
+RESET parallel_tuple_cost;
+RESET parallel_setup_cost;
+RESET min_parallel_table_scan_size;
+RESET min_parallel_index_scan_size;
+RESET pgaudit.log;
+RESET pgaudit.log_client;
+RESET pgaudit.log_level;
+
-- Cleanup
-- Set client_min_messages up to warning to avoid noise
SET client_min_messages = 'warning';

View File

@@ -0,0 +1,143 @@
commit cc708dde7ef2af2a8120d757102d2e34c0463a0f
Author: Tristan Partin <tristan.partin@databricks.com>
Date: 2025-06-23 02:09:31 +0000
Disable logging in parallel workers
When a query uses parallel workers, pgaudit will log the same query for
every parallel worker. This is undesireable since it can result in log
amplification for queries that use parallel workers.
Signed-off-by: Tristan Partin <tristan.partin@databricks.com>
diff --git a/expected/pgaudit.out b/expected/pgaudit.out
index 8772054..9b66ac6 100644
--- a/expected/pgaudit.out
+++ b/expected/pgaudit.out
@@ -2556,6 +2556,37 @@ DROP SERVER fdw_server;
NOTICE: AUDIT: SESSION,11,1,DDL,DROP SERVER,,,DROP SERVER fdw_server;,<not logged>
DROP EXTENSION postgres_fdw;
NOTICE: AUDIT: SESSION,12,1,DDL,DROP EXTENSION,,,DROP EXTENSION postgres_fdw;,<not logged>
+--
+-- Test logging in parallel workers
+SET pgaudit.log = 'read';
+SET pgaudit.log_client = on;
+SET pgaudit.log_level = 'notice';
+-- Force parallel execution for testing
+SET max_parallel_workers_per_gather = 2;
+SET parallel_tuple_cost = 0;
+SET parallel_setup_cost = 0;
+SET min_parallel_table_scan_size = 0;
+SET min_parallel_index_scan_size = 0;
+-- Create table with enough data to trigger parallel execution
+CREATE TABLE parallel_test (id int, data text);
+INSERT INTO parallel_test SELECT generate_series(1, 1000), 'test data';
+SELECT count(*) FROM parallel_test;
+NOTICE: AUDIT: SESSION,13,1,READ,SELECT,,,SELECT count(*) FROM parallel_test;,<not logged>
+ count
+-------
+ 1000
+(1 row)
+
+-- Cleanup parallel test
+DROP TABLE parallel_test;
+RESET max_parallel_workers_per_gather;
+RESET parallel_tuple_cost;
+RESET parallel_setup_cost;
+RESET min_parallel_table_scan_size;
+RESET min_parallel_index_scan_size;
+RESET pgaudit.log;
+RESET pgaudit.log_client;
+RESET pgaudit.log_level;
-- Cleanup
-- Set client_min_messages up to warning to avoid noise
SET client_min_messages = 'warning';
diff --git a/pgaudit.c b/pgaudit.c
index 004d1f9..f061164 100644
--- a/pgaudit.c
+++ b/pgaudit.c
@@ -11,6 +11,7 @@
#include "postgres.h"
#include "access/htup_details.h"
+#include "access/parallel.h"
#include "access/sysattr.h"
#include "access/xact.h"
#include "access/relation.h"
@@ -1339,7 +1340,7 @@ pgaudit_ExecutorStart_hook(QueryDesc *queryDesc, int eflags)
{
AuditEventStackItem *stackItem = NULL;
- if (!internalStatement)
+ if (!internalStatement && !IsParallelWorker())
{
/* Push the audit even onto the stack */
stackItem = stack_push();
@@ -1420,7 +1421,7 @@ pgaudit_ExecutorCheckPerms_hook(List *rangeTabls, List *permInfos, bool abort)
/* Log DML if the audit role is valid or session logging is enabled */
if ((auditOid != InvalidOid || auditLogBitmap != 0) &&
- !IsAbortedTransactionBlockState())
+ !IsAbortedTransactionBlockState() && !IsParallelWorker())
{
/* If auditLogRows is on, wait for rows processed to be set */
if (auditLogRows && auditEventStack != NULL)
@@ -1475,7 +1476,7 @@ pgaudit_ExecutorRun_hook(QueryDesc *queryDesc, ScanDirection direction, uint64 c
else
standard_ExecutorRun(queryDesc, direction, count, execute_once);
- if (auditLogRows && !internalStatement)
+ if (auditLogRows && !internalStatement && !IsParallelWorker())
{
/* Find an item from the stack by the query memory context */
stackItem = stack_find_context(queryDesc->estate->es_query_cxt);
@@ -1495,7 +1496,7 @@ pgaudit_ExecutorEnd_hook(QueryDesc *queryDesc)
AuditEventStackItem *stackItem = NULL;
AuditEventStackItem *auditEventStackFull = NULL;
- if (auditLogRows && !internalStatement)
+ if (auditLogRows && !internalStatement && !IsParallelWorker())
{
/* Find an item from the stack by the query memory context */
stackItem = stack_find_context(queryDesc->estate->es_query_cxt);
diff --git a/sql/pgaudit.sql b/sql/pgaudit.sql
index 6aae88b..de6d7fd 100644
--- a/sql/pgaudit.sql
+++ b/sql/pgaudit.sql
@@ -1631,6 +1631,36 @@ DROP USER MAPPING FOR regress_user1 SERVER fdw_server;
DROP SERVER fdw_server;
DROP EXTENSION postgres_fdw;
+--
+-- Test logging in parallel workers
+SET pgaudit.log = 'read';
+SET pgaudit.log_client = on;
+SET pgaudit.log_level = 'notice';
+
+-- Force parallel execution for testing
+SET max_parallel_workers_per_gather = 2;
+SET parallel_tuple_cost = 0;
+SET parallel_setup_cost = 0;
+SET min_parallel_table_scan_size = 0;
+SET min_parallel_index_scan_size = 0;
+
+-- Create table with enough data to trigger parallel execution
+CREATE TABLE parallel_test (id int, data text);
+INSERT INTO parallel_test SELECT generate_series(1, 1000), 'test data';
+
+SELECT count(*) FROM parallel_test;
+
+-- Cleanup parallel test
+DROP TABLE parallel_test;
+RESET max_parallel_workers_per_gather;
+RESET parallel_tuple_cost;
+RESET parallel_setup_cost;
+RESET min_parallel_table_scan_size;
+RESET min_parallel_index_scan_size;
+RESET pgaudit.log;
+RESET pgaudit.log_client;
+RESET pgaudit.log_level;
+
-- Cleanup
-- Set client_min_messages up to warning to avoid noise
SET client_min_messages = 'warning';

View File

@@ -0,0 +1,143 @@
commit 8d02e4c6c5e1e8676251b0717a46054267091cb4
Author: Tristan Partin <tristan.partin@databricks.com>
Date: 2025-06-23 02:09:31 +0000
Disable logging in parallel workers
When a query uses parallel workers, pgaudit will log the same query for
every parallel worker. This is undesireable since it can result in log
amplification for queries that use parallel workers.
Signed-off-by: Tristan Partin <tristan.partin@databricks.com>
diff --git a/expected/pgaudit.out b/expected/pgaudit.out
index d696287..4b1059a 100644
--- a/expected/pgaudit.out
+++ b/expected/pgaudit.out
@@ -2568,6 +2568,37 @@ DROP SERVER fdw_server;
NOTICE: AUDIT: SESSION,11,1,DDL,DROP SERVER,,,DROP SERVER fdw_server,<not logged>
DROP EXTENSION postgres_fdw;
NOTICE: AUDIT: SESSION,12,1,DDL,DROP EXTENSION,,,DROP EXTENSION postgres_fdw,<not logged>
+--
+-- Test logging in parallel workers
+SET pgaudit.log = 'read';
+SET pgaudit.log_client = on;
+SET pgaudit.log_level = 'notice';
+-- Force parallel execution for testing
+SET max_parallel_workers_per_gather = 2;
+SET parallel_tuple_cost = 0;
+SET parallel_setup_cost = 0;
+SET min_parallel_table_scan_size = 0;
+SET min_parallel_index_scan_size = 0;
+-- Create table with enough data to trigger parallel execution
+CREATE TABLE parallel_test (id int, data text);
+INSERT INTO parallel_test SELECT generate_series(1, 1000), 'test data';
+SELECT count(*) FROM parallel_test;
+NOTICE: AUDIT: SESSION,13,1,READ,SELECT,,,SELECT count(*) FROM parallel_test,<not logged>
+ count
+-------
+ 1000
+(1 row)
+
+-- Cleanup parallel test
+DROP TABLE parallel_test;
+RESET max_parallel_workers_per_gather;
+RESET parallel_tuple_cost;
+RESET parallel_setup_cost;
+RESET min_parallel_table_scan_size;
+RESET min_parallel_index_scan_size;
+RESET pgaudit.log;
+RESET pgaudit.log_client;
+RESET pgaudit.log_level;
-- Cleanup
-- Set client_min_messages up to warning to avoid noise
SET client_min_messages = 'warning';
diff --git a/pgaudit.c b/pgaudit.c
index 1764af1..0e48875 100644
--- a/pgaudit.c
+++ b/pgaudit.c
@@ -11,6 +11,7 @@
#include "postgres.h"
#include "access/htup_details.h"
+#include "access/parallel.h"
#include "access/sysattr.h"
#include "access/xact.h"
#include "access/relation.h"
@@ -1406,7 +1407,7 @@ pgaudit_ExecutorStart_hook(QueryDesc *queryDesc, int eflags)
{
AuditEventStackItem *stackItem = NULL;
- if (!internalStatement)
+ if (!internalStatement && !IsParallelWorker())
{
/* Push the audit event onto the stack */
stackItem = stack_push();
@@ -1489,7 +1490,7 @@ pgaudit_ExecutorCheckPerms_hook(List *rangeTabls, List *permInfos, bool abort)
/* Log DML if the audit role is valid or session logging is enabled */
if ((auditOid != InvalidOid || auditLogBitmap != 0) &&
- !IsAbortedTransactionBlockState())
+ !IsAbortedTransactionBlockState() && !IsParallelWorker())
{
/* If auditLogRows is on, wait for rows processed to be set */
if (auditLogRows && auditEventStack != NULL)
@@ -1544,7 +1545,7 @@ pgaudit_ExecutorRun_hook(QueryDesc *queryDesc, ScanDirection direction, uint64 c
else
standard_ExecutorRun(queryDesc, direction, count, execute_once);
- if (auditLogRows && !internalStatement)
+ if (auditLogRows && !internalStatement && !IsParallelWorker())
{
/* Find an item from the stack by the query memory context */
stackItem = stack_find_context(queryDesc->estate->es_query_cxt);
@@ -1564,7 +1565,7 @@ pgaudit_ExecutorEnd_hook(QueryDesc *queryDesc)
AuditEventStackItem *stackItem = NULL;
AuditEventStackItem *auditEventStackFull = NULL;
- if (auditLogRows && !internalStatement)
+ if (auditLogRows && !internalStatement && !IsParallelWorker())
{
/* Find an item from the stack by the query memory context */
stackItem = stack_find_context(queryDesc->estate->es_query_cxt);
diff --git a/sql/pgaudit.sql b/sql/pgaudit.sql
index e161f01..c873098 100644
--- a/sql/pgaudit.sql
+++ b/sql/pgaudit.sql
@@ -1637,6 +1637,36 @@ DROP USER MAPPING FOR regress_user1 SERVER fdw_server;
DROP SERVER fdw_server;
DROP EXTENSION postgres_fdw;
+--
+-- Test logging in parallel workers
+SET pgaudit.log = 'read';
+SET pgaudit.log_client = on;
+SET pgaudit.log_level = 'notice';
+
+-- Force parallel execution for testing
+SET max_parallel_workers_per_gather = 2;
+SET parallel_tuple_cost = 0;
+SET parallel_setup_cost = 0;
+SET min_parallel_table_scan_size = 0;
+SET min_parallel_index_scan_size = 0;
+
+-- Create table with enough data to trigger parallel execution
+CREATE TABLE parallel_test (id int, data text);
+INSERT INTO parallel_test SELECT generate_series(1, 1000), 'test data';
+
+SELECT count(*) FROM parallel_test;
+
+-- Cleanup parallel test
+DROP TABLE parallel_test;
+RESET max_parallel_workers_per_gather;
+RESET parallel_tuple_cost;
+RESET parallel_setup_cost;
+RESET min_parallel_table_scan_size;
+RESET min_parallel_index_scan_size;
+RESET pgaudit.log;
+RESET pgaudit.log_client;
+RESET pgaudit.log_level;
+
-- Cleanup
-- Set client_min_messages up to warning to avoid noise
SET client_min_messages = 'warning';