Maybe someone on the list has experienced a similar thing
and can jump in ...

 TSM server v5.1.5.4 on AIX 5.1 ML4.
 8 AIT3 (SCSI) tape drives connected to an [ADIC] "SNC" converter ->
 -> Brocade FC switch -> dedicated HBA on the AIX system.

 Server works fine most of the time. However, about once a week,
 a TSM process writing to tape locks up (usually a migration).
 The tape drive and tape volumes involved vary.

 CANCEL PROCESS has no effect (waited > 12 hours), nor do
 tricks like updating the tape involved to acc=UNAVAILABLE.

 HALTing the TSM server leaves the TSM server process "<exiting>"
 (once I waited >4 hours in the hope for something to time out,
 but nothing changed). An AIX re-boot brings everything back to normal.

 Unfortunately, there are no pertinent messages in the TSM log,
 nor in the AIX error log.

Sure we know from experience that (SCSI connected) AIT drives
have a non-zero I/O error rate ... but on SCSI, in the worst
case there always would be an I/O time-out.

Experiences, anyone?

Wolfgang J. Moeller, Tel. +49 551 201-1516/-1510, moeller AT gwdvms.dnet.gwdg 
DOT de
GWDG, D-37077 Goettingen, F.R.Germany     |    Disclaimer: No claim intended!
http://www.gwdg.de/~moeller/ ---- <moeller AT gwdg DOT de> ---- <w.moeller AT 
ieee DOT org>