Horton
Menu

Re: Add pg_stat_autovacuum_priority

Andres Freund <andres@anarazel.de>

View thread
From: Andres Freund <andres@anarazel.de>
To: Tom Lane <tgl@sss.pgh.pa.us>
Cc: Nathan Bossart <nathandbossart@gmail.com>, Alexander Lakhin <exclusion@gmail.com>, Sami Imseih <samimseih@gmail.com>, Bharath Rupireddy <bharath.rupireddyforpostgres@gmail.com>, Robert Treat <rob@xzilla.net>, satyanarlapuram@gmail.com, pgsql-hackers <pgsql-hackers@postgresql.org>, tndrwang@gmail.com
Date: 2026-04-08T20:40:03Z
Lists: pgsql-hackers 
Hi,

On 2026-04-08 15:21:38 -0400, Tom Lane wrote:
> Hmm, yeah I suppose a caller that doesn't care about leakage could
> skip the pfree.  But are there really any of those?  The complaint
> that prompted 02502c1bc concerned databases with many many thousands
> of relations.

Most of the calls come from SQL callable functions that just return a value to
the user. For those I don't think we gain anything by doing retail pfrees, if
anything it'd hurt performance.


> I now realize that what you said upthread about caching the results
> might be a bigger problem, ie if the pgstats code does retain all
> these values then we'd have a memory bloat problem there.  Maybe
> we need a more aggressive API change that includes a way to specify
> "don't cache this result".

I can't stop myself from pointing out that the whole reason this caching stuff
exists is because you insistet that we provide an equivalent for snapshotting
semantics of the pre-shmem stats system...

Note that the whole cached state does automatically get reset at the end of
the transaction (AtEOXact_PgStat()->pgstat_clear_snapshot()), just like it did
before the shmem stats stuff.

Greetings,

Andres Freund

Commits

Same data as JSON: GET /api/v1/messages/:b64id/commits the thread's linked commits as JSON, with link sources. API reference →

  1. Fix double-free in pg_stat_autovacuum_scores.

  2. Add LOG_NEVER error level code.

  3. Add pg_stat_autovacuum_scores system view.

  4. Remove recheck_relation_needs_vacanalyze().

  5. Add elevel parameter to relation_needs_vacanalyze().

  6. Teach relation_needs_vacanalyze() to always compute scores.

  7. Refactor relation_needs_vacanalyze().

  8. Add rudimentary table prioritization to autovacuum.

  9. Fix per-relation memory leakage in autovacuum.

  10. Fix recently-understood problems with handling of XID freezing, particularly