On Wed, 29 Dec 1999, E Hunter wrote:
>
> Its my feeling that testing this against the Auspex wouldn't be fair
> if NetApps is at RAID 4. We'll see if I'm wrong.
No, it's not fair, but I think you'll find an F760 will still beat
a similarly configured Auspex. ;-) If you look at the results of the
SPECsfs97 benchmarks, you'll notice that a lot of vendors "cheat" by
loading up their testbeds with hundreds of drives or dozens of small
filesystems or only using RAID 0+1 or only striping the outer tracks
of each drive, etc. One has to wonder how much performance you lose
with a real-world configuration. Netapp's benchmark results are with
RAID 4, full disk usage, single filesystems. No cheating.
Check the Auspex results and take a large grain of salt with you
(http://www.spec.org/osg/sfs97/results/res99q3/sfs97-19990716-00109.html).
More drives, more storage adapters, more filesystems, more NVRAM,
comparable CPU power == less performance, both in absolute NFS ops/sec
and response times (7462 ops/sec @ 9.4 ms for a single F760 vs. 6030 @
16.7 for a single-node Auspex).
-- Brian Tao (BT300, taob@risc.org) "Though this be madness, yet there is method in't"
This archive was generated by hypermail 2b29 : Tue Jan 04 2000 - 10:20:56 EST