Guys,

I've never seen this before in our environments, and then I see it twice in
one day on two separate systems...

TSM Client for AIX (4.3.3 - SP2 CWS) 3.7.1.0 - Server AIX 4.1.3.0 (4.3.3
SP2 WH2 Wide).

We have the client scheduler running on these boxes, but suddenly backups
stopped running for no apparent reason, when the client appeared to hang...

Invoking 'dsmc' from the command line, followed by 'q se' or 'q files' etc.
is fine, but as soon as I try an 'inc' it seems to hang, with no data being
sent to the server. The session on the server sits in idle until it
eventually times out. Obviously, the server is contactable from the client
as the 'q se' and 'q files' commands work fine, but it's only when a backup
attempt is made do I get problems. I tried (out of desparation) a 'sel' of
a filespace, a 'show locks' on the server (none), a stop / start of the TSM
server process, checked the dsm.sys and opt, even uninstalled and
re-installed the clients (at the same level), but all with the same
response / lack of response.

Checking the dsmerror.log reveals:

Destroy Mutex failed: 16

on each occasion. I'm not entirely sure if this is an error generated by my
control c-ing from the client and severing the server connection (if there
is one at this point), or as a result of the error. I guess I could
check...

Anyone come across this before, what on earth is this 'mutex destroy' which
is failing (I've had a quick check through adsm.org and seen it mentioned
in confusing despatches), and how can I fix it?

Rgds,

David McClelland
---------------------------
Tivoli Storage Management Team
Tivoli Storage Management Team
IBM EMEA Technical Centre,
Mail Point SGJ3, IBM, North Harbour, Portsmouth PO6 3AU, England
Tel: 02392-56 0218
Mob: 07711 120 931
Internet: davidmcclelland AT uk.ibm DOT com