Dana,

The backup is done by nodes.  Each process will select a node and copy all
the files for that node.  What probably happened is that one of the
processes started on a node with a very large amount of data.  Meanwhile
the other process selected all the other nodes and they all had much less
data.  When there were no unselected nodes left, process 1 quit; process 2
will continue until it finishes the big node it is working on.  What I
don't know is if process 2 will quit when it finishes
the big node or if it will go look for more to do, that is back up nodes
that have arrived in the diskpool after process 1 finished.  I suspect it
will look for more to do.

I can't think of any way to balance this when you have a node with much
more data than average.  You can cancel the long running process and let
the data get backed up the next time.

--
--------------------------
--------------------------
Bill Colwell
Bill Colwell
C. S. Draper Lab
Cambridge, Ma.
bcolwell AT draper DOT com
--------------------------
In <OFF94D2A34.09154904-ON862568BA.005389E4 AT equitable-of-iowa DOT com>, on 
04/07/00
In <OFF94D2A34.09154904-ON862568BA.005389E4 AT equitable-of-iowa DOT com>, on 
04/07/00
   at 11:33 AM, "Dana Mitchell;" <Dana.Mitchell AT ING-DM DOT COM> said:

>Hello All,

>I am running TSM 3.7 under OS/390 2.4 with a primary disk pool of 173gb.
>At 6AM I issue "BACKUP STG PRODDISK1  PRODCOPY1   MAXPROCESS=2" command
>to begin backing up this pool.   Two processes are then started and all
>is well.   Less than an hour later, one of the processes ends
>'successfully' after only backing up 12GB.  The other process is still
>going at 10AM and its up to 70GB and counting.

>Does anyone know how the first process decides it's time to end?  I would
>like to start these backups earlier in the morning and keep two processes
>going but I am having trouble understanding what causes these to end,
>seemingly prematurely.

>Any help would be appreciated.

>Dana Mitchell
>Equitable of Iowa