Re: Identifying source of iowait - is it the DB or the raid array

Dave

This is a less than expert opinion. My guess is that you won't have an easy time finding the answer. My guess is that certain information can only be found from the I/O subsystem, like your question whether it is the raid array cache. So you'll want to find the available information resources for the I/O system, and track them. But then some information in Oracle can provide more context to the specific situation, like your question like whether there are a lot of table scans being perfomed.

   Bottom line, I think you're going to have to act like a detective, gather clues everywhere, and assemble them to draw a conclusion. And since you won't know beforehand when an event will occur, you've got to gather information continually, then try to decipher the events that led to the event.

Dennis Williams

On 6/22/05, David Turner <dnt9000_at_yahoo.com> wrote:
> Periodically we have a problem with our raid arrays
> and suddenly iowait shoots up. Does anyone have a
> method of identifying the source of iowait. I find it
> hard to tell whether say the cache on a raid array was
> disabled versus more processes performing table scans.
> I understand historical tracking but often don't have
> this info on the databases. Currently working with 9i.
>
> Thx, Dave
>
>
>
> ____________________________________________________
> Yahoo! Sports
> Rekindle the Rivalries. Sign up for Fantasy Football
> http://football.fantasysports.yahoo.com
> --
> http://www.freelists.org/webpage/oracle-l
>

--
http://www.freelists.org/webpage/oracle-l