2010-07-20 11:08 PM
I just added a used disk to an existing filer and zeroed it (this was fine). Then added it to an existing aggregate and immediately got a string of:
Tue Jul 20 21:56:08 PDT [netapp0a: raid.tetris.cksum.embed:CRITICAL]:
Invalid checksum entry on Disk /aggr0/plex0/rg0/0c.00.7 Shelf 0 Bay 7
[NETAPP X286_S15K5146A15 NA01] S/N [3LN0GZPC00009736X1W9], block #911993, during write operation.
So I immediately ran "disk fail 0c.00.7". But now "disk maint" won't work. Is there any way to run diagnostics on a failed disk?
2010-07-21 01:52 AM
Sounds like the fail command does what it says on the tin. The disk is 'failed' and can not be used as it is fail, ie dead.
Can you try it in a 2nd filer?
Hope it helps
2010-07-21 08:37 AM
We don't have a second filer. We could physically visit the location (it is remote), and remove the drive. But then how would we get the NetApp to "forget" it has ever seen that serial number?
2010-07-28 12:24 PM
I found the answer:
> priv set advanced
> disk unfail xxxxxx
This is a used disk, that when added to an aggregate immediately gave:
But later passed all "disk maint" tests. Go figure.