Capacity reporting dilemma

mikam

Newcomer
Joined
Apr 22, 2013
Messages
3
Reaction score
0
Points
0
Location
Estonia
PREDATAR Control23

Hi

I encountered a situation where we have a client node (a small CentOS 6.6 box) which has about 80GB of data and a 120GB total drive capacity. On the TSM server, it's filespace size is displayed ok (118GB) and the util % is also displayed correctly. However, a logical_mb query reports 580GB (which is suspiciously much because the retention policy is only verexists=2 verdeleted=1 and both the retextra and retonly values are 30 days.

But to top it off, reporting_mb (the value which means how much would the occupancy be without deduplication) reported 10TB occupancy for this particular node. I ran "repair occupancy" command for the node and the value dropped to 2TB.

There has been no issues that could logically explain why the values are off. Not a "hard link" issue, not a "followsymlinks yes" issue, not a "forgotten mount" issue (there has been no external disk mounts).

Has anyone here stumbled upon something similar ?
 
Top