HP MSA2012i, virtual disk is offline

[Rezart_gr]

Hi,

I had a couple of days ago 2 HDs on the same virtual disk failing. The virtual disk is on raid 5 and had a spare. After replacing the 2 hard drives, it started to reconstruct the Raid and looks like failed yesterday, by giving me another error and by presenting 2 HDs as "Leftover". After searching on google, i cleaned the metadata of these hard drives and presented them again to the virtual disk and the reconstruction started again.
However it looks like during the night the reconstruction failed and for some reason the whole Virtual Disk went Offline!
Questions:
1. How can I bring it online?
2. It has now also 2 HDs as "leftovers", can I clear the metadata again and represent them?

This is what I can see at the notifications:

     
Warning  2013-07-31 04:19:19 8 A220121 Vdisk D_0 drive down (Channel:0 ID:8 SN:9QJ2AS5D Encl:0 Slot:8)  
Critical 2013-07-31 04:19:19 314 A220122 FRU type: drive, problem: encl 0 deviceID 8. Vendor: ATA Product ID: ST31000340NS , S/N: 9QJ2AS5D rev: SN06. Related event ID: 220121, type: 8  
Warning  2013-07-31 04:19:19 18 A220123 Vdisk reconstruct failed. Command failed (error code 1). (Vdisk: D_0, SN: 00c0ffd5ab39000093695e4d00

000000)  
Warning  2013-07-31 04:19:19 44 A220124 Unwritable cache data exists for volume (volume: , SN: 00c0ffd5ab3900002a35654d01000000) comprising 1% of cache space  
Critical 2013-07-31 04:19:19 3 A220125 Vdisk offline: D_0, SN: 00c0ffd5ab39000093695e4d00000000

Thanks
Rezart

[StorageMike]

Hi

I would advise against clearing the meta data  and reusing the disks marked as leftover.  These disks are faulty in some way and need to be replaced.  I would arrange to get these disks swapped out and once replaced present them as spares.

Thanks

[Mark Matthews]

I've recently gone through the same, horrible thing with HP support, and it didnt end well for the data...

The Vdisk was RAID5 with a dedicated spare.

One disk would fail and the spare would kick in (sometimes it wouldnt and had to clear metadata in order to make it Available to add it back as a spare first) and reconstruction would begin.

However there was another disk in the same Vdisk with problems that would only show in the logs (not actually fail the disk) which caused the reconstruct to fail every time.

The only way around it was to get 2 replacement disks sent from HP (luckily was still in warrany) and delete the recreate the Vdisk and all its volumes, hence losing all your data!

There is a 'hidden' Trust option you can do in the CLI that lets you mount the Vdisk long enough for you backup any data, but its a one way ride, once you Trust it, the only thing left to do is delete and recreate...

This is a good incentive to always keep your controller and disk firmware upto date!

Thanks,

Mark...

[danletkeman]

I would highly suggest getting a replacement plan together.  We had the same thing multiple times with both of our 2012i's and after 3 years of corrupt vdisks and calling support we finally just got rid of them.