Storage Discuss,

I've just finished building a large storage array.  The array is comprised
of seven (7) shelves each with 24 disks.  The disks (1TB SATA II) are
connected to a 3ware 9650SE in each shelf.  None of the controllers have
BBU's.

I create 4 ZFS volumes (4.4TB usable) of six (6) disks per shelf.  I export
them as iSCSI LU's with COMSTAR.

On the head, I mount all 28 LUN's making 4 RAIDZ's in a single ZFS.  Each
RAIDZ is comprised of one LU of each shelf.

I then share the large ZFS with NFS.  I have a Linux NFS client (NFSv3)
mount it and perform a bonnie run.  The linux server has 4GB of RAM.

Here is the output:
[r...@qa-www-s106:~/bonnie-64-read-only]# ./Bonnie -d /limefs/ -s 8192
File '/limefs//Bonnie.3111', size: 8589934592
Writing with putc()...done
Rewriting...done
Writing intelligently...done
Reading with getc()...done
Reading intelligently...done
Seeker 1...Seeker 3...Seeker 2...start 'em...done...done...done...
              -------Sequential Output-------- ---Sequential Input--
--Random--
              -Per Char- --Block--- -Rewrite-- -Per Char- --Block---
--Seeks---
Machine    GB M/sec %CPU M/sec %CPU M/sec %CPU M/sec %CPU M/sec %CPU  /sec
%CPU
            8   0.8  1.1   1.9  0.4   1.9  0.4  73.5 99.2 106.2  7.3     8
0.1
[r...@qa-www-s106:~/bonnie-64-read-only]#

As you see, the read performance is pretty good, but the write performance
is horrible.

Also, I keep seeing this in the Linux server's /var/log/messages:
Jun 30 22:09:57 qa-www-s106 kernel: nfs: server 172.17.96.200 not
responding, still trying
Jun 30 22:09:57 qa-www-s106 kernel: nfs: server 172.17.96.200 not
responding, still trying
Jun 30 22:09:57 qa-www-s106 kernel: nfs: server 172.17.96.200 not
responding, still trying
Jun 30 22:09:58 qa-www-s106 kernel: nfs: server 172.17.96.200 OK
Jun 30 22:09:58 qa-www-s106 kernel: nfs: server 172.17.96.200 OK
Jun 30 22:09:58 qa-www-s106 kernel: nfs: server 172.17.96.200 OK

>From some googling, I believe this has to do with cache flushes, but I'm not
certain.  Can anyone shed light on this?

Best regards,
Michael Schenck


-- 
Michael Schenck - www.tek-ops.com
_______________________________________________
storage-discuss mailing list
[email protected]
http://mail.opensolaris.org/mailman/listinfo/storage-discuss

Reply via email to