xfs
[Top] [All Lists]

Re: U320 Large Array Performance

To: lord@xxxxxxx
Subject: Re: U320 Large Array Performance
From: "Rick Smith" <rgsmith72@xxxxxxxxxxx>
Date: Wed, 05 Feb 2003 12:48:06 -0800
Cc: linux-xfs@xxxxxxxxxxx
Sender: linux-xfs-bounce@xxxxxxxxxxx
Steve,
As a test, I shortened the size of one of the 15 disks in the array (using fdisk) in order to create a slightly smaller filesystem and thus requiring a smaller number of allocation groups. The resultant filesystem only required 256 allocation groups and I was able to improve performance over 14 disks for the first time. The results from xfs_bmap for this 15 disk array was very similar to the output that I sent you for the 14 disk array. There were no extent gaps between files written in succession and each group of files seem to have the same AG.
I feel the problem is the result of >256 allocation groups. Any thoughts on this?


Rick

From: Steve Lord <lord@xxxxxxx>
To: Rick Smith <rgsmith72@xxxxxxxxxxx>
CC: linux-xfs@xxxxxxxxxxx
Subject: Re: U320 Large Array Performance
Date: 05 Feb 2003 12:02:08 -0600

On Tue, 2003-02-04 at 16:47, Rick Smith wrote:
> Steve,
> Here is the output from xfs_info, please excuse the formatting. I am using
> a chunk size of 4K in the raid. Larger sizes seem to degrade performance.
> Thanks for your help.
>


Can you do one more thing for me, with these configurations send me the
xfs_bmap -v output for a file on each filesystem.

Thanks,

Steve

--

Steve Lord                                      voice: +1-651-683-3511
Principal Engineer, Filesystem Software         email: lord@xxxxxxx


_________________________________________________________________
The new MSN 8: smart spam protection and 2 months FREE* http://join.msn.com/?page=features/junkmail



<Prev in Thread] Current Thread [Next in Thread>