From: Bruce Evans (bde@kralizec.zeta.org.au)
Date: 08/24/93


From: bde@kralizec.zeta.org.au (Bruce Evans)
Subject: Re: SCSI Performance (Yet Again)
Date: 25 Aug 1993 11:46:26 +1000

In <1993Aug22.150338.1810@jonh.wimsey.bc.ca> jhenders@jonh.wimsey.bc.ca (John Henders) writes:

>root@hip-hop.suvl.ca.us (Remco Treffkorn) writes:

>>... Iozone is certainly
>>not perfect, but gives you a relative measure if compiled correctly.

> Exactly. In my comparisons, I even noted that the numbers I see for
>performance under DOS for my Adaptec 1452b are similar, though slightly
>higher. ...

The values reported by iozone are almost guaranteed not to be exceeded
for physical i/o to normal files, so low values are good for proving
that your i/o system is poor. It's high values that are usually
meaningless (they usually mean that part of the i/o was non-physical).

Read speeds much lower than write speeds are probably caused by
insufficent read-ahead by the file system. This problem is often masked
by read-ahead by the controller.

-- 
Bruce Evans  bde@kralizec.zeta.org.au