Re: Add log_autovacuum_{vacuum|analyze}_min_duration

2025-08-14 17:56 に kasaharatt wrote:
> Hi,
> 
> 2025-08-14 13:26 に Shinya Kato wrote:
>> Hi,
>> 
>> On Wed, Aug 13, 2025 at 5:44 PM kasaharatt 
>> <kasaharatt@oss.nttdata.com> wrote:
>> 
>>> >>>> > Approach 2:
>>> >>>> > - log_autovacuum_min_duration: Changed behavior, which controls only
>>> >>>> > autovacuum logging.
>>> >>>> > - log_autoanalyze_min_duration: New parameter, which controls
>>> >>>> > autoanalyze logging.
>>> >>>>
>>> >>>> My vote is for this approach. It is probably OK to change the
>>> >>>> behavior of
>>> >>>> log_autovacuum_min_duration, as the new GUC will have the same
>>> >>>> default
>>> >>>> value.
>>> >>>
>>> >>>
>>> >>> Thank you for voting. I also think this approach is reasonable to
>>> >>> implement.
>>> >>
>>> >>
>>> >> A new patch is attached.
>>> >> Thoughts?
>>> I reviewed this patch.
>>> I also have no particular objections to the Approach 2.
>> 
>> Thank you for the review!
>> 
>>> > +    <term><literal>log_autoanalyze_min_duration</literal>,
>>> > <literal>toast.log_autoanalyze_min_duration</literal>(<type>integer</type>)
>>> (snip)
>>> > +       "toast.log_autoanalyze_min_duration",
>>> This patch adds the log_autoanalyze_min_duration parameter fot TOAST
>>> tables.
>>> However since PostgreSQL currently does not support ANALYZE on TOAST
>>> tables,
>>> isn't this parameter unnecessary?
>> 
>> You're right; that was a mistake. I've fixed it in the v4 patch.
> Thanks for the patch updating.
> 
>> Additionally, I removed the above setting from the test files in
>> src/test/modules/xid_wraparound/t/ (001_emergency_vacuum.pl,
>> 002_limits.pl, 003_wraparounds.pl). The reason is that these tests
>> check for autovacuum logs, not autoanalyze logs. You can run the test
>> with the following command:
>> make check -C src/test/modules/xid_wraparound 
>> PG_TEST_EXTRA='xid_wraparound'
> Yeah, I agree that.
> 
> I confirmed this patch could apply to HEAD and all make check tests 
> were passed.
> I've manually tested it in a few different ways, and it's working as
> expected so far.
> 
> Unless there are any objections from others, I think I can mark this
> patch as "Ready for Committer".

Since there were no particular comments, I've marked it as Ready for 
Committer.

Best regards,
-- 
Kasahara Tatsuhito
NTT DATA Japan Corporation