On Thu, May 14, 2009 at 04:04:55PM -0400, [email protected] wrote: > So I've setup my test system. It's a dual quad core system with 2.33 ghz > procs and 8GB of RAM. I have 12 2.5" 146 SAS drives connected via a LSI > SAS expander and 1068 hba. For the boot drive, I'm using a 2.5" sata and > I also have a Intel SSD (32GB SLC) which I'm using for ZIL. Oh - i'm > running Express Community Edition snv_113. > > Can anyone share some insight on how best to benchmark fs performance? I > usually use iometer, but I can't get dynamo to compile on this platform. > > I see alot of people use iozone, but I don't really understand why some > of the performance numbers exceed the maximum throughput in theory for 12 > SAS disks.
Take a look at SWAT & vdbench: http://przemol.blogspot.com/2008/07/swat-vdbench-excellent-couple.html While filebench, iozone or bonnie create sort of "artificial" workload (but trying to be close to real), using SWAT & vdbench you can record your real (yes - REAL !) workload and then play it on benchmarked storage. I have been using it for a few months recording and playing our workload. Since then I have forgotten about iozone, filebench ... ;-) Regards Przemyslaw Bak (przemol) -- http://przemol.blogspot.com/ ---------------------------------------------------------------------- Dzwonki na komorkę! Sprawdz >> http://link.interia.pl/f2161 _______________________________________________ storage-discuss mailing list [email protected] http://mail.opensolaris.org/mailman/listinfo/storage-discuss
