2013-08-27 09:17 AM
OnTAP version 8.0.3.P2
I have two 500GB nfs shares, one with 450GB of data and another with zero. Both are mounted on a linux server and an rsync was performed copying data from one nfs share to another. Unfortunately, the destination nfs share is now at 100% disk space used. There should be close to 50GB free. Can someone shed some light on this phenomenon? Currently trying to comprehend dedup and snashots and if they play a role in the mysterious additional 50+GB on the destination share.
Thanks in Advance.
Solved! SEE THE SOLUTION
2013-08-27 10:33 AM
This could be because of ASIS. If the source has been deduped, the destination won't reflect that until after a dedupe cycle. Check df -s on the source to see how much data is deduped.
It could also be snapshots. If, for example, you ran a test rsync to the destination then deleted the data, it would be tied up in snaps on the destination until they roll off.
Also, if you don't exclude .snapshots from the rsync, then you risk copying all your source snaps to the destination, which is definitely not what you want.
2013-08-27 12:13 PM
Thanks Bill. As it turned out, there were multiple factors involved with my issue. First rsync converted hard links to actual files and .snapshot dirs were unintentionally included. Because hardlinks are essential for our application to run properly, I'll be performing an ndumpcopy instead. Hindsight, I should have investigated using ndumpcopy in the first place. Hopefully hardlink are preserved using ndumpcopy. Otherwise, I'll be increasing the quota a bit.