Unsolved
This post is more than 5 years old
3 Posts
0
13374
November 19th, 2015 13:00
DiskGroup Failed on MD36xxf
Hi,
Early this week we lost a physical disk, the rebuilt to the hot spare worked well and was finished. Then last night another drive failed, that make our 41TB Disk pool to fail. So this morning I revive the last drive failed with the following script: revive physicaldisk [6,2];
That kick a DiskGroup reconstruction, but it still at 0% since this morning (8hrs ago). The DiskGroup is still down and volume not accessible.
Do anybody have an idea to speed things up?
Thank you in advance,
David Dumais
0 events found
No Events found!
DELL-Daniel Ca
243 Posts
0
November 19th, 2015 15:00
Hey David.
Sounds like we have a little mess brewing. This really has me curious - a dynamic disk pool is a bunch of RAID 6 parity stripes randomly spread across 10 disks each stripe. So, a 2 disk loss would at the most degrade some of the data.
What I'm wondering here is if something was still trying to rebuild when you revived the 2nd disk, and threw off the reconstruct, and may have damaged some of the parity. Either way, we really need to see a support bundle to see what's missing, and see about putting it back together.
I'll send you an email, and you can reply to that with the diagnostics.
I'll keep you posted.
ddumais
3 Posts
0
November 19th, 2015 17:00
A mistake was made in my post, It's a disk Group, so basically a RAID5 with a dedicated hot spare. The rebuild to the hot spare was finish, so from my thinking I could lose another drive... that what happen but the result was not expected.
I will send you the bundle tomorrow.
Thank you.
David