Open
Conversation
NikolayS
force-pushed
the
claude/pg-stat-statements-rows-filtered-ddwRA
branch
from
1938136 to
3e66693
Compare
Add a new rows_filtered column to pg_stat_statements that tracks the total number of rows removed by filter conditions across all plan nodes. This helps identify queries that would benefit from better indexing. The feature is controlled by a new GUC pg_stat_statements.track_rows_filtered (default: off). When enabled, the module uses INSTRUMENT_ROWS instrumentation to collect nfiltered1 (scan/join qual) and nfiltered2 (recheck conditions) from all plan nodes via planstate_tree_walker. Key implementation details: - Uses INSTRUMENT_ROWS (not INSTRUMENT_ALL) to avoid timing overhead - Only increments tuple counters, which has negligible performance impact - GUC is superuser-only (PGC_SUSET), can be changed at runtime - Stats file header bumped to invalidate old stats files This addresses a common DBA need to identify inefficient sequential scans where adding an index could significantly improve performance.
NikolayS
force-pushed
the
claude/pg-stat-statements-rows-filtered-ddwRA
branch
from
3e66693 to
b577803
Compare
This file contains hidden or bidirectional Unicode text that may be interpreted or compiled differently than what appears below. To review, open the file in an editor that reveals hidden Unicode characters.
Learn more about bidirectional Unicode characters
Sign up for free
to join this conversation on GitHub.
Already have an account?
Sign in to comment
2 participants
Add this suggestion to a batch that can be applied as a single commit.This suggestion is invalid because no changes were made to the code.Suggestions cannot be applied while the pull request is closed.Suggestions cannot be applied while viewing a subset of changes.Only one suggestion per line can be applied in a batch.Add this suggestion to a batch that can be applied as a single commit.Applying suggestions on deleted lines is not supported.You must change the existing code in this line in order to create a valid suggestion.Outdated suggestions cannot be applied.This suggestion has been applied or marked resolved.Suggestions cannot be applied from pending reviews.Suggestions cannot be applied on multi-line comments.Suggestions cannot be applied while the pull request is queued to merge.Suggestion cannot be applied right now. Please check back later.
Add a
rows_filteredcolumn topg_stat_statementsto track the total number of rows removed by filter conditions during query execution. This metric helps identify queries that may benefit from better indexing, as highrows_filteredvalues indicate that many rows are being scanned but then discarded.The implementation:
INSTRUMENT_ROWSbeforeExecutorStart(usingINSTRUMENT_ROWSinstead ofINSTRUMENT_ALLto avoid the overhead of timing instrumentation)ExecutorEndto sumnfiltered1(scanqual/joinqual) andnfiltered2(other quals) from all nodesExample output:
This helps identify inefficient queries where many rows are scanned but few are returned, suggesting an index might help.