Closed artemgavrilov closed 4 months ago
Attention: Patch coverage is 80.00000%
with 1 line
in your changes missing coverage. Please review.
Project coverage is 84.99%. Comparing base (
dacb41f
) to head (63f274a
).
Files | Patch % | Lines |
---|---|---|
pg_stat_monitor.c | 75.00% | 0 Missing and 1 partial :warning: |
:umbrella: View full report in Codecov by Sentry.
:loudspeaker: Have feedback on the report? Share it here.
PG-644
This PR introduces two changes:
pgsm_track_application_names
that allows to disable application name trackingIt turned out that taking
application name
is expensive operation, it requires locking inside PG. This causes performance dramatically on big amount of connections. For example on TPC-C benchmark with 512 threads PGSM with application name tracking (default) gives -15% QPS, while without tracking only -3% QPS.Docs PR: https://github.com/percona/pgsm-docs/pull/63