| To: | Rajagopal Ananthanarayanan <ananth@xxxxxxx> |
|---|---|
| Subject: | Re: Parallel read performance crawls |
| From: | KELEMEN Peter <fuji@xxxxxxx> |
| Date: | Thu, 24 May 2001 20:07:10 +0200 |
| Cc: | linux-xfs@xxxxxxxxxxx |
| In-reply-to: | <3B0D2EB0.CFF8F78B@sgi.com>; from ananth@sgi.com on Thu, May 24, 2001 at 08:54:24AM -0700 |
| Organization: | ELTE Eotvos Lorand University of Sciences, Budapest, Hungary |
| References: | <20010524162005.A26407@chiara.elte.hu> <3B0D2EB0.CFF8F78B@sgi.com> |
| Reply-to: | KELEMEN Peter <fuji@xxxxxxx> |
| Sender: | owner-linux-xfs@xxxxxxxxxxx |
| User-agent: | Mutt/1.3.17i |
* Rajagopal Ananthanarayanan (ananth@xxxxxxx) [20010524 08:54]:
> FWIW, in the following, I see not only read performance drop
> with 2 bonnies, but also re-write. Are you using the same disk
> as the target for the 2 bonnies? If so, I would strongly suspect
> disk head contention. Delayed allocation/writes in the block
> output case seem to have a positive effect: clusters of writes
> are fairly large; there is no such clustering at the higher
> level for reads (re-write is read+write iirc).
I'm using a 80G LVM volume striped across two disks. ReiserFS
shows 18M(write)/8M(read) performance when run in parallel. I
will try to restructure my LVM setup.
Peter
--
.+'''+. .+'''+. .+'''+. .+'''+. .+''
Kelemen Péter / \ / \ / fuji@xxxxxxx
.+' `+...+' `+...+' `+...+' `+...+'
|
| <Prev in Thread] | Current Thread | [Next in Thread> |
|---|---|---|
| ||
| Previous by Date: | Re: Errr... what's all this about then?, Eric Sandeen |
|---|---|
| Next by Date: | Re: Crashes in various ext2 functions while running xfstest/check, Seth Mos |
| Previous by Thread: | Re: Parallel read performance crawls, Rajagopal Ananthanarayanan |
| Next by Thread: | Re: Parallel read performance crawls, KELEMEN Peter |
| Indexes: | [Date] [Thread] [Top] [All Lists] |