[OmniOS-discuss] background snapshot destruction impact

[Tobias Oetiker]

I am running omnios r151008 on a recent box with 3gig SAS disks
arranged in a raidz2 pool.

I have this 7TB filesystem where I have my zimbra server store its
backups.  The backup-sets are highly redundant, so for a time I had
deduplication turned on to great effect, but as the size of the
backup-sets grew, I ran out of ram and performance on the system
became realy bad. So I turned deduplication of. Performance for
newly written data is fine again.

The other day, I destroyed a bunch of old snapshots on this
filesysem. Thanks to background destruction this did not realy take
a long time. But now as background destruction is progressing, the
system remains very slugish when doing io on the pool where
the destruction is taking place.

I put up some graphs and raw data on

  http://tobi.oetiker.ch/background-destruction/

to show the state of things.

My Questions:

a) would it be faster to send/receive the content of the
   deduplicated filesystem to a new non deduplicated
   and then destroy the entire filesystem (not the pool).

b) is there a way to monitor progress on the background
   destruction?

c) shouldn't the smarter write throttle change
   https://github.com/illumos/illumos-gate/commit/69962b5647e4a8b9b14998733b765925381b727e
   have helped with this by makeing zfs do its internal things
   with a lower priority.

cheers
tobi


-- 
Tobi Oetiker, OETIKER+PARTNER AG, Aarweg 15 CH-4600 Olten, Switzerland
http://it.oetiker.ch tobi at oetiker.ch ++41 62 775 9902 / sb: -9900