Dm_oracle Cores Dump When Restarted (Doc ID 2953983.1)

Last updated on JUNE 13, 2023

Applies to:

Symptoms

On environment upgraded from Oracle Communications Billing and Revenue Management (BRM) 7.5 to BRM 12 PatchSet (PS) 4, customer is running a multischema system with 10 schemas, with each server running 10 dm_oracle and with dm_n_be=122.
Two chains are used, one for running batch apps (such as pin_bill_accts) and one for online calls (billing fulfillment and billing enquiries).
dm_trans_timeout is set to 60 for batch and set to 5 for online.

When shutting down processes for a chain by using pin_ctl stop, followed by "pin_ctl start" for the same processes (ie pin_ctl stop cm dm_oracle followed by pin_ctl start cm dm_oracle), most of dm_oracle backend processes generate core dumps. The frontend processes do not generate core dump files.

See the stack trace (all cores are the same):

The problem is that file system on which all these core files are generated is getting filled up, as the number of core files is high and they have a big size.

Cause

My Oracle Support provides customers with access to over a million knowledge articles and a vibrant support community of peers and Oracle experts.