>Occasionally we see the scheduler freeze up on Unix systems
>(Solaris).  There are no errors in the scheduler log - the last entry is
>just a file which was being backed up.  The scheduler process is running,
>but not doing anything. ...

Paul - Problems like this can be due to irregularities in the client
       file system, or even a disk defect.  In your Solaris system, I'd
look in /var/log/messages for any indications of a hardware problem, and
take a good look at the directory that the backup got stuck in, for
oddities like perhaps circular symbolic links or crud in general.
Certainly, see if the last file reported in the backup can be read.
And see if there are any dsmerror.log entries.
Check the server Activity Log at the time to see if any corrolary problems.
Consider using a 'find' or 'ls' command to run through the file system as
TSM would and see if those native operations have any trouble.
If no problems found in these basic checks, consider running the client
in Trace mode to see what it's doing when it gets stuck.

   Richard Sims, BU