| To: | linux-xfs@xxxxxxxxxxx |
|---|---|
| Subject: | Re: optimizing raid performance with xfs |
| From: | Wessel Dankers <wsl@xxxxxxxxxxxx> |
| Date: | Wed, 19 Mar 2003 21:57:01 +0100 |
| In-reply-to: | <Pine.LNX.4.44.0303191521360.2065-100000@xxxxxxxxxxxxxxxxxx> |
| Mail-followup-to: | linux-xfs@xxxxxxxxxxx |
| References: | <200303191948.LAA21241@xxxxxxxxxxxxxxxxx> <Pine.LNX.4.44.0303191521360.2065-100000@xxxxxxxxxxxxxxxxxx> |
| Sender: | linux-xfs-bounce@xxxxxxxxxxx |
| User-agent: | Mutt/1.5.3i |
On 2003-03-19 15:30:58-0500, Joshua Baker-LePain wrote: > I've got another 3ware based system that's similar to yours (two cards > with a software stripe), and I found that increasing the chunk-size of the > stripe increased performance (at the cost of CPU load) -- I'm using 4096k > in production. In theory you should use a small stripe size for large files (parallellism for sequential access patterns) and a large stripe size for small files (low latency for random access). Of course YMMV so do test it out with your own specific workload (as opposed to a generic benchmark like Bonnie++). HTH, -- Wessel Dankers <wsl@xxxxxxxxxxxx> |
| <Prev in Thread] | Current Thread | [Next in Thread> |
|---|---|---|
| ||
| Previous by Date: | Re: optimizing raid performance with xfs, Steve Lord |
|---|---|
| Next by Date: | Re: kernel from CVS doesnt compile, Rusell Cattelan |
| Previous by Thread: | Re: optimizing raid performance with xfs, Christian Guggenberger |
| Next by Thread: | TAKE - Un-race-ify timer manipulation in pagebuf, Eric Sandeen |
| Indexes: | [Date] [Thread] [Top] [All Lists] |