446157 So we migrated this vfiler to a 8.1.2 head in another datacenter yesterday
na01> sis status Path State Status Progress /vol/vm64 Enabled Active Fingerprint is being recovered from volume (0% complete)
na01> df -sh Filesystem used saved %saved /vol/vm64net2/ 3829GB 3051GB 44%
Disk space is sawtoothing since the dedup job kicked off:
I've almost doubled the volume size. Experienced this with a smaller (700g) volume as well What is going on? Is there a way to avoid this ?
thanks
Look into netapp internal *bug 657692*. Wasnt available on public bug tracking as of 2 weeks ago.
we dealt with an internal bug relating to the fingerprint stuck at 0% on our dedupes with 8.1.2. I did not see the sawtoothing or volume size go nuts. We also didnt migrate vfilers.
Fix was to sis stop the dedupe job and restart it with '*sis start -s volume*'. basically rededupes from scratch your whole volume. Netapp actually had us run a perfstat while we fixed a few volumes with this to gather more data on the bug. All i have gotten out of them is that the 'math' gets off at some point under certain conditions. While i ran this i did not lose any previously deduped space, so my volume didnt instantly blow up. Just offering up in case its related somehow to your issue.
-dave
-- View this message in context: http://network-appliance-toasters.10978.n7.nabble.com/Fingerprint-is-being-r... Sent from the Network Appliance - Toasters mailing list archive at Nabble.com.
Dave, your workaround should help a lot on our next dozen migrations
The finger print finally finished and returned all the space - crazy bugs we are running into lately
thanks
On Mar 20, 2013, at 4:57 PM, "dave.withers" dave.withers@gnomefoo.com wrote:
Look into netapp internal *bug 657692*. Wasnt available on public bug tracking as of 2 weeks ago.
we dealt with an internal bug relating to the fingerprint stuck at 0% on our dedupes with 8.1.2. I did not see the sawtoothing or volume size go nuts. We also didnt migrate vfilers.
Fix was to sis stop the dedupe job and restart it with '*sis start -s volume*'. basically rededupes from scratch your whole volume. Netapp actually had us run a perfstat while we fixed a few volumes with this to gather more data on the bug. All i have gotten out of them is that the 'math' gets off at some point under certain conditions. While i ran this i did not lose any previously deduped space, so my volume didnt instantly blow up. Just offering up in case its related somehow to your issue.
-dave
-- View this message in context: http://network-appliance-toasters.10978.n7.nabble.com/Fingerprint-is-being-r... Sent from the Network Appliance - Toasters mailing list archive at Nabble.com. _______________________________________________ Toasters mailing list Toasters@teaparty.net http://www.teaparty.net/mailman/listinfo/toasters
-
dave.withers -
Fletcher Cocquyt