We have a problem with a celerra ns20 nas. Is there a way of copy raw data of corrupted filesystem to another storage system? We have two LUNs offline because a raid group faulted and is affecting a filesystem with valuable data on it. We looking for ways to recover the maximun amount of data.
Data is written over a cifs share, not encrypted.
I'm thinking about copy all LUNs data with someting like a dd command.
How drives died ?
Did Celerra send any error message on failed drives beforehead or just simple drive taken offline ?
If 2 then it's bad luck as not much can be done in this case.
Best option is to address this issue to EMC support, i suppose that Celerra is not covered by service contract so one time paid basis is your option.
We can try to recover some info from Celerra, but no guaranties.
The event was registered on navisphere event log. Drive x failed, no hot spares avaliable.
The filesystem has aprox. 6,5TB size, uxfs format, and 9 LUNs. Two of them are the faulted LUNs.
I tried to mount or fsck but get a input/output error.
All correct, it would be I/O error, as double drive failure under 1 RG.
I am not aware of any simple way of restoring data in this case, either physically read survived data from hdds (including dead ones), or try to restore backup.
Please elaborate what array model you have: is it Celerra NS20 or NS-120 ?
NS20 went EOSL so EMC won't be able to take a look at it, you can try to take hdds to some data recovery organisation, maybe they will be able to help.
If NS-120 Service Request under EMC can be raised, some portion of data can be recovered that way.
If only ONE disk failed, then the raid group would be de-graded (showing as faulted) as the raid should be protecting the data (Raid 5 can handle one faulted disk). With a Raid 5 raid group, with one disk faulted you should still be able to access the data on the LUNs as the remaining drives should be able to calculate the data from parity.
Do you know what the raid type of the raid group with the failed disk is?
If you trespass the LUN to the peer SP you "might" be able to get to the data on the LUNs.