Hi Bob,
Thanks for responding.
In answer to your question, it appears as though sessions simply tail off - no statistical summary or normal end-of-job detail, although now I think about it, tape devices appear to unload without difficulty since we've not seen any 'device in use' errors on subsequent jobs.
No outstanding processes of any kind on CM or clients, and no error message patterns to suggest a root cause. We do have an isolated case involving a single client whose job would hang on encountering an I/O error, but I've 'quarantined' that client so that it doesn't appear in any scripts other than its own.
I have the following parameters set in omnirc (since March 2009):
OB2SHMIPC=0
OB2IPCKEEPALIVE=1
OB2INETTIMEOUT=150
I'm not seeing any errors that would suggest a loss of connectivity or timeout, and the effect on backups is to show them as still running, while in the case of copy operations, virtual media are not recycled when a job fails to complete.
It's more of an annoyance than anything else - it's not happening every night, nor in the same job each time, but it's adding extra housekeeping I could do without :(
I'll have a look for your posts on Keepalive - perhaps there's something I've misssed...
Cheers,
Bob
My environment consists mostly of W2K8 systems, with small numbers of W2K3, W2012 and a couple of Linux and OpenVMS clients.
