We paused backup operations and ran some numbers: We got about 105
MBytes/sec NFS throughput on a single 1 GigE connection from said
RHEL3 to the NAS, which seems quite good, dedicated linke and all
that. cpu consumption for for the rpcio daemon stayed, during raw
iozone testing, under 20%. I think I'm pretty happy with that?
overall system load was under 40%.
Something I noticed with NFS vs the fiber-attached Xserves, there is
no longer a 70%-89% "iowait" bottleneck eating CPU power and making
the system otherwise unresponsive to inputs/outputs.
Since iSCSI is treated as block device, is does it incur the same
iowait penalty as other block devices?
Thanks for your input. So far the NAS tested to be 3x faster at
sustained writes and random reads, which seems perfect to our use. I
can finally keep...7+ days on disk instead of .5 days, a marked
improvement.
--Eugene
--
Eugene Vilensky
evilensky AT gmail DOT com
To sign off this list, send email to listserv AT listserv.temple DOT edu and
type "signoff networker" in the body of the email. Please write to
networker-request AT listserv.temple DOT edu if you have any problems with this
list. You can access the archives at
http://listserv.temple.edu/archives/networker.html or
via RSS at http://listserv.temple.edu/cgi-bin/wa?RSS&L=NETWORKER
|