Arev32 locking up (AREV32)
At 12 DEC 2012 05:45:55PM John Bowen wrote:
We are running Arev32 in OI 9.3.1. We have a large table that has 140K in the Lk and 7 Gig in the Ov. There are about 200,000 records in the table. This table is an audit table that is used to track changes in multiple tables.
I ran a process last Friday that removed records that were over 6 months old. I deleted about 22,000 records. The delete process completed, but users could not log into Arev32. The login process hung on the 'Initializing..' window. Users that were already in the system were able to continue to work. We also have an external process that connects to OI tables via Oengine. This process was not able to connect to Oengine after the delete process completed and received an error (Error: Engine creation FAILED, code *-3102*). To resolve the issue we had to reboot the server that OI is installed on.
I ran a select on the table yesterday and today and received the same results (users were not able to log into Arev32). We had to restart the server to resolve the issue.
Is there a know issue with tables that have grown that large? Is there a better work around than rebooting the server?
At 13 DEC 2012 10:40AM Bob Orsini wrote:
John, first I would try doing a select in OI and see if that creates the same problem. Also check the table for any MFS and see if it is creating an issue. There are no Known issues deleting records in large tables.
At 13 DEC 2012 10:46AM John Bowen wrote:
The only mfs on the table is si.mfs. I will try to run a select on the table in OI today.
At 17 DEC 2012 03:29PM Jared Bratu wrote:
When you receive the 3102 error have the other sessions of Arev32 crashed or do they continue to function correctly?
Is the Linearhash service still operating on the server after the error code displays?
At 18 DEC 2012 10:38AM John Bowen wrote:
Any Arev32 sessions that are active can continue to work. No new sessions will start up. The LH service continues to work for the active Arev32 sessions.
At 18 DEC 2012 10:42AM John Bowen wrote:
We have not had any further lock-ups since I backed up the file and cleared the table. I am working on setting up a sandbox environment where I can run tests against the large table with out affecting the production environment.