Re: BUG #8470: 9.3 locking/subtransaction performance regression

10.04.2015, 19:57, Alvaro Herrera kirjoitti:
>> I didn't try this patch, but I built today's REL9_3_STABLE with the patch
>> from your mail on the 31st of Dec
>> (http://github.com/saaros/postgres/tree/alvaro-multixact-optimize-9.3) and
>> ran the older version of my appliaction's test suite which has a case that
>> times out after 3 minutes with unpatched 9.3.  With the current patch the
>> test case successfully completes in ~20 seconds which is a major improvement
>> although it's still slower than 9.2  It also passes all other test cases in
>> my test suite.
>>
>> Do you think it'll make it to 9.3.3?
>
> I just pushed this to 9.5.  I'm not sure how much value I place on a
> backpatch at this point.  There were a number of complaints about the
> performance regression, but other than yours I got no report on whether
> the fix was good, or on whether it was stable.  If you have other
> thoughts, I'm willing to listen.  Anyway if you could test the patch I
> pushed, it'd be great.

The component where this was an issue was totally rewritten already a
while ago to allow us to take 9.3 into use and as I can't run the old
version anymore I can't verify the fix in a real application.  The code
we had there was almost identical to the test case I posted.

I don't think the fix needs to be backpatched.

Thanks,
Oskari