4.0.6 upgrade, hit Oracle max processes limit - Is there a suggested value for this parameter?
![]()
I've recently upgraded from RTC 3.0.1.2 to 4.0.6. We had been using a max processes limit of 300, but this does not appear to be high enough for 4.0.6. The system hit the num processes limit once during the actual upgrade process. It then hit it again during the nightly ETL jobs. (no users using the system at all yet)
I am wondering if there is a suggested number for the max processes, or if we just need to bump it up and guess if it is enough.
|
3 answers
![]() Hi Jason, Usually decreasing the number of JDBC and Mediator pool connections would be the alternative to raising the Oracle limits. Are you using WAS or Tomcat? You do need to set the Oracle limit higher than the JDBC/Mediator connections due to ETLs and other system connections, but I am afraid there is no rule of thumb that I am aware of. If you can provide the additional info, it might be able to comment further. If you can get your DBA to give you the list of users accessing the databases, then it should also be able to guesstimate how many you (should) require. Paul Comments I'm running with Tomcat.
Unfortunately the DBA's were locked out of the DB when the max limit was hit and they had to bounce the DB. So we couldn't tell what was going on. I now have access to see the current sessions and have been keeping an eye on that as well as the number of processes in v$process. I've yet to see the number of processes get over 60, even when running all the ETL processes. So I have no idea what would have caused it to jump up to over 300 processes. The only idea I have is perhaps the first-time bulk migrations have some large, parallel sql that pushed it over the limit...
Hi again Jason,
|