Re: ORA-07445 core dump periodically happen

Hi, Jimmy:

Thank you for your suggestion.
The large pool and shared pool in our database is big enough. I have checked V$SGASTAT, there is a lot of free space for both pools. Even I turn off the MTS to switch to dedicated server configuration, the problem is still there. The resource limit is fine too.

But you say, you ended up increasing the number of threads that each user can have and increasing the number of processes that each user can have. Are you using the parallel server? Since the THREADS in init.ora is only for parallel
server? As I know, we can increse the PROCESSES in init.ora to increase the number of total processes, but how can we increase number of processes that each user can have?

Is it possible the core dump come from the application or oracle bug? Then how can we know this? And how can we find it.

Jingjing

jimmyn1_at_aol.com (JimmyN1) wrote in message news:<20010803214507.29004.00003998_at_ng-ch1.aol.com>...
> I had this with MTS on HP. I'm not sure what the sun equivalent kernel
> parameters are but we did the following:
>
> 1) if you have one defined, monitor your large pool at peak. we used mercury
> load runner to do this and we increased our shared pool.
>
> select * from v$sgastat;
>
> 2) Monitor your resources:
>
> select * from v$resource_limit
>
> Conclusions:
>
> We ended up increasing the number of threads that each user can have
>
> We increase the number of processes that each user can have
>
> We increased our large pool to 500MB
>
> After we did that, our ora-7445 disappeared.
Received on Mon Aug 06 2001 - 12:48:57 CDT