I was running some stats on node Occucapancy and filespace utilized, when
noticed that in a few cases, the amount of data backed up exceeded the sum of
all the filespace.

I was using the following commands:

select node_name, backup_mb from auditocc

select node_name,capacity,pct_util from filespaces

as well as q occ and q filespace and adding up the totals by hand just to
compare.

In the one case I've investigated the most, the client is AIX and the filespace
type is JFS. Is this caused by AIX's LVM and/or the JFS? Assuming that the info
on the amound backed up is accurate, is there any way to get more accurate data
on the total filespace on the client?

Steve Cochran
Dartmouth College