Hello,
              This is actually my first correspondence in this facility. I
hope this message is successful.

Regarding database space, what is important is not so much the size of the
drives on the nodes being backed up, but rather the number of files or
objects being backed up. I do a query actlog begind=-n s='backed up' to
display the adsm message that indicates total objects backed up. Once I am
able to accumulate the statistics for each node over a period, I multiply
the total number of objects by 1k to determine the database size required.
The 1k takes into consideration the tapepool, copypool, and a 25% overhead.
This is not a totally scientific process, but it is a reasonable estimate.
If you are not able to determine the number of objects, then the size can
be estimated as 1-5% of the total space on the nodes or servers. I suppose
that 3% would be a good start.