Re: How fast should an IO be on a SYMMETRIX?

"Sal L" <member46676_at_dbforums.com> wrote in message news:3550820.1067801502_at_dbforums.com
> And could you comment the statement regarding tkprof disk
> statistics taken
>
> from Knowledge Xpert for Oracle Administration:
>
> "disk Number of data blocks read from disk for each parse, fetch,
> or execute. If a single multiblock read returns eight blocks,
> this figure is incremented only once for the physical read."
> According this one you shouldn't divide 375/8 to get read request count.

Hi, Sal,

The unit of reading from disk is block, not number of times of reading. I can prove it with a simple test
(http://www.stormloader.com/yonghuang/computer/diskreadunit.txt) The article at Knowledge Xpert is correct in saying it's "Number of data blocks read" but I don't know why the author is not consistent and goes on to say the counter is incremented only once for a multiblock read.

As an aside, Cary Millsap's book "Optimizing Oracle Performance" not only says "disk" in tkprof is number of blocks read or written, but "buffer" is also number of buffers, not number of buffer reads. This still differs from Sybrand's interpretation (see the thread "Help with interpreting TKPROF output" around Aug 31).

Yong Huang

-- 
Posted via Mailgate.ORG Server - http://www.Mailgate.ORG