coredumps of OBIEE catalogcrawler

Question

We are encountering strange (and new for me) OBIEE phenomena at our customer deployment:

Disk space is taken by coredumps of OBIEE catalogcrawler process. New coredump appears every hour at :12 min in /opt/oracle/mw/instances/instance1.

catalogcrawler is responsible for indexing for full text search in OBIEE, and as far as I understand we don't use full text search in OBIEE.

So, it may be enough to just get rid of these cores.

We have have removed all the cores, but new cores created every hour, and diskspace will be eaten again eventually.

Do you know something about these coredumps?

I'm using OBIEE 10.1.1.1.5

Joel Acha · Answer

Your best bet is to raise an SR and find out if there is an 11.1.1.5 patch for this bug. If there isn't, then your only other option is to upgrade or a very crude alternative is to set some job which systematically deletes these core dumps.

783226 · Answer

Yes. 11.1.1.5.

Is there a way to overcome this bug without upgrading?

Joel Acha · Answer

Hi @783226

I presume that you meant version 11.1.1.5. If so, there is a bug listed in MOS - Bug 14643648 : CATALOGCRAWLER CRASH IN ODBC