On Fri, Jul 12, 2024 at 11:19:05AM -0500, Nathan Bossart wrote:
> I suppose it would be silly to allow even lower values for
> autovacuum_naptime (e.g., by moving it to ConfigureNamesReal and setting
> the minimum to 0.1).
I've thought about that as well, and did not mention it as this would
encourage insanely low naptime values resulting in fork() bursts.
> That's a neat trick. I was confused why this test generates an autovacuum
> worker at all, but I now see that you are pausing it before we even gather
> the list of tables that need to be vacuumed.
Yep. More aggressive signals aren't going to help. One thing I also
considered here is to manipulate the db list timestamps inside a
USE_INJECTION_POINTS block in the launcher to make the spawn more
aggressive. Anyway, with 600ms in detection where I've tested it, I
can live with the responsiveness of the patch as proposed.
> Looks reasonable to me.
Thanks. I'll see about stressing the buildfarm tomorrow or so, after
looking at how the CI reacts.
--
Michael