pg.ddx.io pgsql-performance@postgresql.org mailing list archive
help / color / mirror / Atom feedWhich benchmark to use for testing FS?
3+ messages / 2 participants
[nested] [flat]
* Which benchmark to use for testing FS?
@ 2006-08-23 22:23 Jeff Davis <pgsql@j-davis.com>
2006-08-24 01:50 ` Re: Which benchmark to use for testing FS? Michael Stone <mstone+postgres@mathom.us>
0 siblings, 1 reply; 3+ messages in thread
From: Jeff Davis @ 2006-08-23 22:23 UTC (permalink / raw)
To: pgsql-performance
I am planning to test various filesystems on some new hardware I'm
getting. Is pgbench a good way to try out the filesystem?
I'm currently planning to test some or all of:
Linux: ext2, ext3, XFS, JFS, reiser3, reiser4
FreeBSD: UFS, UFS+SU
So, I'm looking for a good way to test just the filesystem performance
through PostgreSQL (since database access is different than normal FS
activity). Would pgbench give me a good approximation?
Also, do ext2 or UFS without soft updates run the risk of losing or
corrupting my data?
I saw Chris Browne did some benchmarks back in 2003 and determined that
JFS was a good choice. However, I assume things have changed somewhat
since then. Does anyone have a pointer to some newer results?
Regards,
Jeff Davis
^ permalink raw reply [nested|flat] 3+ messages in thread
* Re: Which benchmark to use for testing FS?
2006-08-23 22:23 Which benchmark to use for testing FS? Jeff Davis <pgsql@j-davis.com>
@ 2006-08-24 01:50 ` Michael Stone <mstone+postgres@mathom.us>
2006-08-24 16:05 ` Re: Which benchmark to use for testing FS? Jeff Davis <pgsql@j-davis.com>
0 siblings, 1 reply; 3+ messages in thread
From: Michael Stone @ 2006-08-24 01:50 UTC (permalink / raw)
To: pgsql-performance
On Wed, Aug 23, 2006 at 03:23:03PM -0700, Jeff Davis wrote:
>Also, do ext2 or UFS without soft updates run the risk of losing or
>corrupting my data?
I suggest you check the list archives; there's a lot of stuff about
filesystems and disk configuration in there.
Mike Stone
^ permalink raw reply [nested|flat] 3+ messages in thread
* Re: Which benchmark to use for testing FS?
2006-08-23 22:23 Which benchmark to use for testing FS? Jeff Davis <pgsql@j-davis.com>
2006-08-24 01:50 ` Re: Which benchmark to use for testing FS? Michael Stone <mstone+postgres@mathom.us>
@ 2006-08-24 16:05 ` Jeff Davis <pgsql@j-davis.com>
0 siblings, 0 replies; 3+ messages in thread
From: Jeff Davis @ 2006-08-24 16:05 UTC (permalink / raw)
To: Michael Stone <mstone+postgres@mathom.us>; +Cc: pgsql-performance
On Wed, 2006-08-23 at 21:50 -0400, Michael Stone wrote:
> On Wed, Aug 23, 2006 at 03:23:03PM -0700, Jeff Davis wrote:
> >Also, do ext2 or UFS without soft updates run the risk of losing or
> >corrupting my data?
>
> I suggest you check the list archives; there's a lot of stuff about
> filesystems and disk configuration in there.
>
I spent a while looking in the list archives, but the list archives have
been misbehaving lately (you click on a search result and the message
that appears doesn't have the same subject as the one you clicked on).
They may have fixed that (they are aware of the problem, according to
pgsql-www). Also, the messages I was able to find were mostly from a
long time ago.
If you have a pointer to a particularly useful thread please let me
know.
Regards,
Jeff Davis
^ permalink raw reply [nested|flat] 3+ messages in thread
end of thread, other threads:[~2006-08-24 16:05 UTC | newest]
Thread overview: 3+ messages (download: mbox mbox.gz follow: Atom feed)
-- links below jump to the message on this page --
2006-08-23 22:23 Which benchmark to use for testing FS? Jeff Davis <pgsql@j-davis.com>
2006-08-24 01:50 ` Michael Stone <mstone+postgres@mathom.us>
2006-08-24 16:05 ` Jeff Davis <pgsql@j-davis.com>
This inbox is served by DDX for PostgreSQL; see mirroring instructions
for how to clone and mirror all data and code used for this inbox