Hello Andres,
>> In conclusion, and very egoistically, I would prefer if this patch could
>> wait for the checkpoint scheduling patch to be considered, as it would
>> basically invalidate the X00 hours of performance tests I ran:-)
>
> These two patches target pretty independent mechanics. If you patch were
> significantly influenced by this something would be wrong. It might
> decrease the benefit of your patch a mite, but that's not really a
> problem.
That is not the issue I see. On the principle of performance testing it
really means that I should rerun the tests, even if I expect that the
overall influence would be pretty small in this case. This is my egoistic
argument. Well, probably I would just rerun a few cases to check that the
impact is "mite", as you said, not all cases.
Another point is that I'm not sure that this patch is ripe, in particular
I'm skeptical about the hardcoded 1.5 without further testing. Maybe it is
good, maybe 1.3 or 1.6 is better, maybe it depends and it should just be a
guc with some advises about how to set it. So I really think that it needs
more performance figures than "it has a positive effect on one load".
Well, this is just my opinion, no need to care too much about it:-)
--
Fabien.