This question should be more specific I know - but I have a few OMS servers that
use a repository database and sometimes it just stops accepting requests once it gets
overloaded. The last time it happened right after a backup and we had
to shutdown the database. I'm not looking for a solution just a few suggestions
on where to look to pin-point the cause. I looked in the alert logs
but it's hard to determine where the problems are originating. I think this whole
install needs some major tuning.
A suggestion on where to start would help - the docs are not helping much.
To observe the things, you can check the following:
1. Verify the db alert log if there are any dead locks or any ORA errors.
2. Verify the high Time /High Waits/executions queries from AWR reports, to determine which tables exactly in the repository are effected.
3. Also would suggest to run the EMDIAG kit and check for the module Repository, to get some hits, if some issues exists.