Re: Cache buffer chains latch bug (8.1.7.2)

I put the 2 statspack reports on my web page (work has this thing with sending attachments).

http://www.reddragon.org/~zifnab/oracle/sp_6_11.lst

     this is a 16 minute slice from our production database the day 
     after we originally upgraded (just FYI).  This isn't really 
     useful in looking at the below 2 reports, but I think it 
     does point out the problem we had quite well.



-----------------------------------------------------------

The index root block that caused all the contention from above is under child latch number 1612 in the below 2 traces. These tests below were generated with 30 concurrent sessions.

http://www.reddragon.org/~zifnab/oracle/sp_7_8.lst

     this was prior to applying the patch.

http://www.reddragon.org/~zifnab/oracle/sp_9_10.lst

     this was after the patch.

I think the most telling portion of it is the child latch breakdowns.

The other interesting thing to notice is that the waits for cache buffer handles has all but disappeared as well.

scott

h.de.boer_at_itcg.nl (Herman de Boer) wrote in <bde5777e.0201151041.38191533_at_posting.google.com>:

>Hello Scott,
>
>I would like to see tha stats before and after.
>
>Kind Regards,
>
>Herman de Boer
>sr. consultant
>IT Consultancy Group bv
Received on Thu Jan 17 2002 - 13:54:49 CST