Have you all noticed how certain errors come and go throughout the ages
and versions? The memory leak, which can be fixed with an event, the
4031, which has been permanently fixed in every single release of Oracle
since 7.0, and others.

The root causes stay the same, though. Go back to the docs from when the
problems were thoroughly detected, diagnosed and repaired and you will
find ever-green advise for 4031-handling...

I'm not saying there couldn't be some new and exciting reason for 4031,
but mostly it's the same old, same old.

Mogens


Joe Armstrong-Champ wrote:

> Tracy,
>
> We upgraded a database from 9i to 10gr2 on aix 5.2 last weeek and got
> a 4031 which killed an etl process that runs daily. I'm not the dba
> working on the problem but I know that 1 of the fixes is to increase
> the data limit (ulimit -d). That seemed to help somewhat but not
> entirely. The dba who is working on it is also playing with some sga
> parms. I'll pass your info to her to see if it helps.
>
> Joe


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