Marco,

Tivoli could do either of two things for us to avoid the problem you describe 
(below, and
which used to drive me nuts):

1. Provide a parameterless 'cancel reclamation' command (as was done at 
3.1.2.40 for
expirations)

or possibly

2. Cause 'upd stg XXX recl=100' to cancel any running reclamations for that pool

In the absence of this, you have to write some sort of script (outside TSM 
because TSM
scripting doesn't let you do it) which issues a 'q pr' and picks out the 
process id's which
you then pass back in individual 'can pr NNN' commands.

How you do this depends on your server platform and its scheduling/scripting 
facilities.
I have OS/390 and my friendly CONTROL-O administrator provided a very neat 
solution.

Good luck,
Geoff

>Date: Fri, 20 Apr 2001 10:45:18 +0300
>From: Marco Supino <marco AT SD.CO DOT IL>
>Subject: Space reclamation.
>MIME-Version: 1.0
>Content-Type: multipart/mixed ; boundary="------------DB54185992B9557EA4653642"
>
>Hi.
>
>Can i make the TSM Server stop any space reclamation between tapes when
>the disk storage pools reaches its high limit, so it can migrate it to
>tapes ? what happens now is that nodes backups are delayed because of
>tape reclamation,
>
>Any help appreciated.
>
>Thanks.
>
>Marco.