* Performance testing results
@ 2007-06-27 16:15 Valerie Clement
2007-06-27 21:19 ` Mingming Cao
2007-06-28 13:47 ` Jean noel Cordenner
0 siblings, 2 replies; 5+ messages in thread
From: Valerie Clement @ 2007-06-27 16:15 UTC (permalink / raw)
To: ext4 development; +Cc: Dave Kleikamp, Mingming Cao, Avantika Mathur
Hi,
I re-ran some FFSB tests on ext3, ext4 and xfs filesystems, on a
2.6.22-rc5 kernel and with the latest ext4-git-tree (update of June 26).
The results are available here:
http://www.bullopensource.org/ext4/20070627/ffsb-write.html
Regards,
Valérie
^ permalink raw reply [flat|nested] 5+ messages in thread
* Re: Performance testing results
2007-06-27 16:15 Performance testing results Valerie Clement
@ 2007-06-27 21:19 ` Mingming Cao
2007-06-28 13:47 ` Jean noel Cordenner
1 sibling, 0 replies; 5+ messages in thread
From: Mingming Cao @ 2007-06-27 21:19 UTC (permalink / raw)
To: Valerie Clement; +Cc: ext4 development, Dave Kleikamp, Avantika Mathur
On Wed, 2007-06-27 at 18:15 +0200, Valerie Clement wrote:
> Hi,
> I re-ran some FFSB tests on ext3, ext4 and xfs filesystems, on a
> 2.6.22-rc5 kernel and with the latest ext4-git-tree (update of June 26).
>
> The results are available here:
> http://www.bullopensource.org/ext4/20070627/ffsb-write.html
>
> Regards,
> Valérie
>
>
Thanks a lot Valerie!!
^ permalink raw reply [flat|nested] 5+ messages in thread
* Re: Performance testing results
2007-06-27 16:15 Performance testing results Valerie Clement
2007-06-27 21:19 ` Mingming Cao
@ 2007-06-28 13:47 ` Jean noel Cordenner
1 sibling, 0 replies; 5+ messages in thread
From: Jean noel Cordenner @ 2007-06-28 13:47 UTC (permalink / raw)
To: Valerie Clement
Cc: ext4 development, Dave Kleikamp, Mingming Cao, Avantika Mathur
Valerie Clement a écrit :
> Hi,
> I re-ran some FFSB tests on ext3, ext4 and xfs filesystems, on a
> 2.6.22-rc5 kernel and with the latest ext4-git-tree (update of June 26).
>
> The results are available here:
> http://www.bullopensource.org/ext4/20070627/ffsb-write.html
>
> Regards,
> Valérie
An update of the Iozone test is also avialable:
http://www.bullopensource.org/ext4/20070627/iozone.html
cheers,
Jean noel
^ permalink raw reply [flat|nested] 5+ messages in thread
* Re: Performance testing results
@ 2007-06-28 4:46 sftf
2007-06-28 3:52 ` Eric Sandeen
0 siblings, 1 reply; 5+ messages in thread
From: sftf @ 2007-06-28 4:46 UTC (permalink / raw)
To: valerie.clement, linux-ext4
Hi!
IMHO this test is not 100% correct.
ext3 | data=writeback
ext4 | data=writeback,extents,delalloc
xfs | defaults is ordered !
So you have compared ext's in writeback (which fastest mode) vs xfs in ordered.
^ permalink raw reply [flat|nested] 5+ messages in thread
* Re: Performance testing results
2007-06-28 4:46 sftf
@ 2007-06-28 3:52 ` Eric Sandeen
0 siblings, 0 replies; 5+ messages in thread
From: Eric Sandeen @ 2007-06-28 3:52 UTC (permalink / raw)
To: sftf; +Cc: valerie.clement, linux-ext4
sftf wrote:
> Hi!
> IMHO this test is not 100% correct.
> ext3 | data=writeback
> ext4 | data=writeback,extents,delalloc
> xfs | defaults is ordered !
>
> So you have compared ext's in writeback (which fastest mode) vs xfs in ordered.
Actually xfs's default (only) mode is more like writeback than ordered.
So that's a fair comparison. I bet it is xfs's barriers that hurt
it, though - while they are pretty much required on a single disk with a
volatile write cache, I think xfs's barrier implementation hurts it more
than barriers for ext*
-Eric
^ permalink raw reply [flat|nested] 5+ messages in thread
end of thread, other threads:[~2007-06-28 13:54 UTC | newest]
Thread overview: 5+ messages (download: mbox.gz follow: Atom feed
-- links below jump to the message on this page --
2007-06-27 16:15 Performance testing results Valerie Clement
2007-06-27 21:19 ` Mingming Cao
2007-06-28 13:47 ` Jean noel Cordenner
-- strict thread matches above, loose matches on Subject: below --
2007-06-28 4:46 sftf
2007-06-28 3:52 ` Eric Sandeen
This is a public inbox, see mirroring instructions
for how to clone and mirror all data and code used for this inbox;
as well as URLs for NNTP newsgroup(s).