Backup Central

Tsm server v6.2.2.
Redhat 6.
Ts1120 tape drives in 3494 library.

Have a process requesting a volume mount. Library thinks the volume is there, however, the mount won't time out. Mount wait on the devclass devininition is 60 minutes, already has been waiting for 90 minutes.
Cancel proc still pending, can't update the volume to read-only or unavailable, tsm says in use.

How to kill this process then try and figure out what is up?

Unfortunately library is offsite, so trying to avoid a trip down.

Thanks for any assistance.


Gary Lee
Senior System Programmer
Ball State University
phone: 765-285-1310

=

This is the gnarly situation where a pending I/O operation is blocking.
The best you can do remotely is use TSM and 3494 queries to ascertain where the tape seems to be within the process of mounting it. In some cases, you can jog the tape drive at the OS level to cause I/O status to be returned to the host, and possibly unblock the TSM operation. The tape/drive problem will remain, however. In my experience, the most common problem is a tape drive defect (vacuum failure, mechanical problem, leader block issue) such that the load doesn't complete, and the tape is stuck in the drive. Performing an unload operation (which can be initiated remotely) may get it unstuck: sometimes, a tape drive power cycle may be a more severe way of getting it out. An IBM CE is the final recourse in that scenario.

Richard Sims