Re: Long wait times for db file scattered read

>
> unless my maths is wrong you are waiting twice as long on average for each
> read in the second run, however the average wait appears to be 0.001 or
> 0.002 of a second in the two cases, which rather suggests to me you are
> getting the data from a cache on the disk subsystem. I'd be looking at
> whether I really need to request 1million blocks from this table when I'm
> going to end up with 69 rows returned. What does the sql look like?

Thank you for your response. You and another responder have both suggessted the wait times are not excessive. Is there a rule of thumb for this evaluation? Please note the final number of rows returned are 1000. I'm in agreement that it appears the AK7 index is not very selective.

Aha, I just found out 3 indexes in this table were not being analyzed as part of the batch run -- even though I was assured everything under the Europe schema was being analyzed. A case study for not overlooking the obvious in spite of what you're told.

>
> btw is this 10g or a third party tool formatting the report like this?
>
> The tool is Trace Analyzer you can get off Metalink. Created by Ruben Sierra I believe. I really like it's comprehensiveness and easy readability.

Thanks again for your interest and help.

John

> --
> Niall Litchfield
> Oracle DBA
> Audit Commission UK
> *****************************************
> Please include version and platform
> and SQL where applicable
> It makes life easier and increases the
> likelihood of a good answer
> ******************************************
Received on Thu Mar 18 2004 - 19:25:20 CST