ALM Process From Within a Batch Hangs Randomly With No Error or Crash Dump Being Generated (Doc ID 2428982.1)

Last updated on AUGUST 02, 2023

Applies to:

Symptoms

While running batch, one of the ALM processes sometimes hangs during execution.

Its normal execution time is 2 to 2.5 hrs. Once hung, no error/crash dump was generated in as long as 6 hours of wait time.

After it hangs, process is killed, rmipc.sh is executed to remove all hanging semaphore/shared memory (after shutting down all OFSAA services) and process is restarted (when the batch is restarted, it continues by restarting this task alone).

Then it executes successfully but the issue appears again at the 2nd or 3rd subsequent run.

Other observations:

1). Process itself also hangs while being run separately. Running process from outside the batch, on separate instrument tables, also doesn't hang.

2). After checking log for the final products being processed, a filter was created on those and process was executed again, having a successful result.

3). Other jobs (ALM processes running in parallel) were put on hold, so that this can run in isolation, but the same happens.

4). Despite Severity "0" and "1" errors being resulted from Cash Flow Edits, this does not seem to be a data issue, as rerunning usually resolves the issue.

5). Other ALM processes are running fine.

Steps in KM Note: <Document 1307768.1> How to Improve ALM Processing Performance are already applied.

Cause

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