Re: Berserk Autovacuum (let's save next Mandrill)

On Sat, 28 Mar 2020 at 19:21, David Rowley <dgrowleyml@gmail.com> wrote:
> Thank you.  Pushed.

I'm unsure yet if this has caused an instability on lousyjack's run in
[1]. I see that table does have 30,000 rows inserted, so it does seem
probable that it may receive an autovacuum now when didn't before. I
did a quick local test to see if swapping the "ANALYZE pagg_tab_ml;"
to "VACUUM ANALYZE pagg_tab_ml;" would do the same on my local
machine, but it didn't.

I'll keep an eye on lousyjack's next run.  If it passes next run, I
may add some SQL to determine if pg_stat_all_tables.autovacuum_count
for those tables are varying between passing and failing runs.

David

[1] https://buildfarm.postgresql.org/cgi-bin/show_log.pl?nm=lousyjack&dt=2020-03-28%2006%3A33%3A02