lkml.org 
[lkml]   [2008]   [Jun]   [6]   [last100]   RSS Feed
Views: [more markup]  [less markup]  [headers]  [forward] 
 
Messages in this thread
/
SubjectRe: [patch] Re: PostgreSQL pgbench performance regression in 2.6.23+
FromMike Galbraith <>
DateFri, 06 Jun 2008 08:13:00 +0200
On Fri, 2008-06-06 at 01:03 -0400, Greg Smith wrote:

> I think I might not be testing exactly the same thing you did, though, 
> because the pattern doesn't match.  I think that my Q6600 system runs a 
> little bit faster than yours, which is the case for small numbers of 
> clients here.  But once we get above 8 clients your setup is way faster, 
> with the difference at 15 clients being the largest.  Were you perhaps 
> using batch mode when you generated these results?

No, those were with stock settings.

> Regardless, clearly your patch reduces the regression with the default 
> parameters to a mild one instead of the gigantic one we started with.

Unfortunately, after the recent reverts, we're right back to huge :-/

I'm trying to come up with a dirt simple solution that doesn't harm
other load types.  I've found no clear reason why we regressed so badly,
it seems to be a luck of the draw run order thing.  As soon as the load
starts jamming up a bit, it avalanches into a serialized mess again.  I
know the why, just need to find that dirt simple and pure win fix.

> Considering how generally incompatible this benchmark is with this 
> scheduler, and that there are clear workarounds (feature disabling) I can 
> document in PostgreSQL land to "fix" the problem defined for me now, I'd 
> be happy if all that came from this investigation was this change.  I'd 
> hope that being strengthened against this workload improves the 
> scheduler's robustness for other tasks of this type, which I'm sure there 
> are more of than just pgbench.

I consider pgbench to be a pretty excellent testcase.  Getting this
fixed properly will certainly benefit similar loads, Xorg being one
that's just not as extreme as pgbench.

> You get my vote for moving toward committing it+backport even if 
> the improvement is only what I saw in my tests.  If I can figure out how 
> to get closer to the results you got, all the better.

It's committed, but I don't think a back-port is justified.  It does
what it's supposed to do, but there's a part 2.  I suspect that your
results differ from mine due to that luck of the run order draw thing.

	-Mike



\
 
 \ /
  Last update: 2008-06-06 08:15    [from the cache]
©2003-2008