Re: Cache buffer chains latch bug (8.1.7.2)

I have a Test environment running Oracle Apps R11 in 8.1.7.1 and do see Cache Buffer Chains Latch contention -- this is on a server with 2 CPUs but 3 test environments (the other two test environments are 8.0.6 while Production is still 8.0.5). I will be upgrading the Test to 8.1.7.2.1 + 1967363 on Thursday. I'll post my findings a few days after that.

What did you say about the number of CPUs being relevant ? That the problem manifests on servers with few (<4) CPUs ?

Hemant K Chitale
http://hkchital.tripod.com
"Scott Gamble" <zifnab_at_nospam.reddragon.org> wrote in message news:Xns918FC0854CE3Fzifnabnospamreddrago_at_64.152.100.163...
> Oracle Version = 8.1.7.2.1
> OS = Tru64 5.1 (might be 5.0 escapes me at the moment)
>
> Just wanted to keep those intersted in the cache buffer chains latch
> bug that was present in 8.1.7.1 and 8.1.7.2 that I had posted about a
> couple of times.
>
> We upgraded one of our databases today that is being replaced, and
> prior to applying the 'one off' patch for the fix of bug #1967363
> we were able to reproduce the latching problem that we were having
> finally with our test case. Appears that the number of
> CPU's is critical in causing the bug to occur. With our normal
> 4 cpu dev boxes we could not cause anything near the problem we had
> originally, however today on one of our old production machines we
> did reproduce the problem.
>
> After applying the one off patch for the above bug everything looked
> completely normal. (normal for our application which still has a few
> problems of its own)
>
> As of right now the machine is running with a normal production
> work load on it, and we are showing no signs of the massive
> latch contention that ground the DB to a halt.
>
> If anyone is intersted, I can post some of the relevant pieces from
> our statspacks prior to the one off patch application,
> and after it if anyone is interested.
>
> Scott Gamble
>
Received on Tue Jan 08 2002 - 07:53:37 CST