Wanda,

BINGO!

I found the count of MaxNumMP on clients to be either 0, 2, or 999.  I have
*never* set that value so I guess it was introduced randomly by an upgrade
(probably from ADSM 3.1 to TSM 4.1).

I used the command you recommended to set them all to 1, and the failing
nodes worked OK.  Both versions.

Tonight I'm trying direct-to-tape backups of the large clients that write
into the LTO.  The smaller clients talk to an 800 MB disk pool of 20 x 40
MB volumes.  Should be interesting...

Thanks again.

Tab







"Prather, Wanda" <Wanda.Prather AT JHUAPL DOT EDU>@VM.MARIST.EDU> on 03/11/2002
11:42:07 AM

Please respond to "ADSM: Dist Stor Manager" <ADSM-L AT VM.MARIST DOT EDU>

Sent by:  "ADSM: Dist Stor Manager" <ADSM-L AT VM.MARIST DOT EDU>


To:   ADSM-L AT VM.MARIST DOT EDU
cc:
Subject:  Re: AIX client fails direct to tape backup


I think it was introduced in the 3.7 server.

Depends on server level, not client level though; I've still got some WinNT
clients running 3.1.0.6, and they have MAXNUMMP set (although as I recall
there was an issue when I first moved to that server level, that clients
which had NEVER had their MAXNUMMP set, would not behave as they should by
default, until I set it explicitly; or clients that had MAXNUMMP that
defaulted to 0 didn't behave the same as ones that were explicitly set to
0,
or something like that.  Anyway, I can't check now, because I remember I
did
a global update node * maxnummp=1 to clean it all up.

But that was a long time and several sever version ago.  Just thought I
would mention it...