On Mon, Mar 12, 2012 at 4:45 PM, Jeff Janes <jeff.janes@gmail.com> wrote:
>>> Do you think the difference is in the CPU architecture, or the
>>> IO subsystem?
>>
>> That is an excellent question. I tried looking at vmstat output, but
>> a funny thing kept happening: periodically, the iowait column would
>> show a gigantic negative number instead of a number between 0 and 100.
>
> On which machine was that happening?
The IBM server.
>> This makes me a little chary of believing any of it. Even if I did,
>> I'm not sure that would fully answer the question. So I guess the
>> short answer is that I don't know, and I'm not even sure how I might
>> go about figuring it out. Any ideas?
>
> Rerunning all 4 benchmarks (both 16MB and 32MB wal_buffers on both
> machines) with fsync=off (as well as synchronous_commit=off still)
> might help clarify things.
> If it increases the TPS of Nate@16MB, but doesn't change the other 3
> situations much, then that suggests the IO system is driving it.
> Basically moving up to 32MB is partially innoculating against slow
> fsyncs upon log switch on that machine.
Mmm, yeah. Although, I think it might have been 64MB rather than 32MB
that I tested on that machine.
> Does the POWER7 have a nonvolatile cache? What happened with
> synchronous_commit=on?
Haven't tried that yet.
> Also, since all data fits in shared_buffers, making
> checkpoint_segments and checkpoint_timeout be larger than the
> benchmark period should remove the only other source of writing from
> the system. With no checkpoints, no evictions, and no fysncs, it is
> unlikely for the remaining IO to be the bottleneck.
Another thing to test.
Meanwhile, here are some TPS graphs at 16MB and 32MB on the IBM POWER7
machine. 32 clients, 1800 seconds, scale factor 300, synchronous
commit off.
--
Robert Haas
EnterpriseDB: http://www.enterprisedb.com
The Enterprise PostgreSQL Company
