From: Seger, Mark <mark.se...@hp.com>
Subject: RE: [Ipmitool-devel] performance questions
To: "John Philips" <johnphilip...@yahoo.com>,
"Ipmitool-devel@lists.sourceforge.net" <Ipmitool-devel@lists.sourceforge.net>
Date: Friday, March 26, 2010, 11:01 AM
I was checking around internally and was told that on least our
systems there are 2 different interfaces and so look at this:
[r...@hpc01p001 ~]# time for i in `seq 1 10`; do ipmitool -d 0
-S /tmp/xxx sdr >/dev/null; done
real 0m18.793s
user 0m0.007s
sys 0m0.019s
[r...@hpc01p001 ~]# time for i in `seq 1 100`; do ipmitool -d 1
-S /tmp/xxx sdr >/dev/null; done
real 0m2.779s
user 0m0.043s
sys 0m0.142s
note in the second case I’m doing 10 times more intereractions
and it’s still completing in almost a 10th of the time. This is
with V1.8.10.
-markThat's interesting, but are you sure you're even getting valid results in
the second case? Docs say that '-d' refers to the device number, i.e.
/dev/ipmi0. Perhaps it runs so quickly because /dev/ipmi1 doesn't exist and
the command just issues an error.
------------------------------------------------------------------------------
Download Intel® Parallel Studio Eval
Try the new software tools for yourself. Speed compiling, find bugs
proactively, and fine-tune applications for parallel performance.
See why Intel Parallel Studio got high marks during beta.
http://p.sf.net/sfu/intel-sw-dev
_______________________________________________
Ipmitool-devel mailing list
Ipmitool-devel@lists.sourceforge.net
https://lists.sourceforge.net/lists/listinfo/ipmitool-devel
