public inbox for [email protected]
help / color / mirror / Atom feedperformance impact on read replicas during heavy vacuums on primary
2+ messages / 2 participants
[nested] [flat]
* performance impact on read replicas during heavy vacuums on primary
@ 2026-02-03 01:02 Dirk Krautschick <[email protected]>
2026-02-04 16:29 ` Re: performance impact on read replicas during heavy vacuums on primary Greg Sabino Mullane <[email protected]>
0 siblings, 1 reply; 2+ messages in thread
From: Dirk Krautschick @ 2026-02-03 01:02 UTC (permalink / raw)
To: pgsql-performance
Is there a scenario, where a very heavy vacuum on the primary side is
affecting significantly
the read replica(s) in matter of performance (not the overall apply lag,
which is expected) of
selects? I do not get this reproduced ...only the lag is clear but even
hammering down the
primary with vaccum keeps my standby answering fine.
Cheers
^ permalink raw reply [nested|flat] 2+ messages in thread
* Re: performance impact on read replicas during heavy vacuums on primary
2026-02-03 01:02 performance impact on read replicas during heavy vacuums on primary Dirk Krautschick <[email protected]>
@ 2026-02-04 16:29 ` Greg Sabino Mullane <[email protected]>
0 siblings, 0 replies; 2+ messages in thread
From: Greg Sabino Mullane @ 2026-02-04 16:29 UTC (permalink / raw)
To: Dirk Krautschick <[email protected]>; +Cc: pgsql-performance
Is this a theoretical question, or have you seen performance issues? Yes,
it is certainly possible for vacuum to affect the replica performance. More
vacuum activity means more WAL means more WAL replay on the replica, using
up CPU and I/O. In most real-world cases, this will not be a problem, as
the replica should be able to keep up quite well even for near-constant
vacuum activity. You would need a very high WAL volume and a very
underprovisioned replica before seeing any ill effects. There are plenty of
other performance bottlenecks that will likely hit you before this one.
Cheers,
Greg
^ permalink raw reply [nested|flat] 2+ messages in thread
end of thread, other threads:[~2026-02-04 16:29 UTC | newest]
Thread overview: 2+ messages (download: mbox mbox.gz follow: Atom feed)
-- links below jump to the message on this page --
2026-02-03 01:02 performance impact on read replicas during heavy vacuums on primary Dirk Krautschick <[email protected]>
2026-02-04 16:29 ` Greg Sabino Mullane <[email protected]>
This inbox is served by agora; see mirroring instructions
for how to clone and mirror all data and code used for this inbox