Hi

TSM Server v6.3.1

Some odd behaviour when migrating a disk pool to tape

The disk pool (devicetype=disk) is 6tb in size and has approximately
2.5tb of data in it from last nights backups.


I backup the stgpool, it copies 2.5tb to tape. Fine with this

I mig stgpool lo=0 maxpr=3. Start fine.

(I do not use a duration parameter)

First migration finishes after a few minutes,  migrating 500mb. Lower
than i expected, i guess
Second migration finishes after 30 minutes, migrating 138gb.
Third migration finishes after 1.5 hours, migrating 819gb

But the pool now shows about 24% utilised, so still about 1.5tb of
data remaining


The tape pool the migrations are writing to has colloocation=group specified.
I have three collocgroups, containg approximately 250 nodes. All of
the nodes on the server are within these 3 groups
I noticed that one of the clients was still backing up. It's a slow
backup, always has been. That node is in one of the collocgroups.
When that client backup completed, i ran migration again with lo=0 and
it is now beyond 1tb and still running. Pct utilisation of the disk
pool is now down to 10%


So, because the backup of that node is still running, will that
prevent migration from migrating data from that specific collocgroup
while a backup of a client within that group is in process?

Any comments welcome.

Regards