This post is more than 5 years old
2 Posts
0
2930
March 15th, 2017 10:00
PowerEdge 2900 one of 4 disk failed in RAID 5 array
Hi,
this is my situation regarding RAID5 array.
Problem is on Dell Power Edge 2900 server with PERC 6/i controller.
Last week one of 4 disks in RAID5 array failed due the power cut off and VD is still online but degraded. Server OS MS 2003 works and data is OK.
I've replaced failed HDD with brand new (same model but not certified by dell), and PERC 6/i controller did not start rebuilding process automatically.
I've read on this pages that is necessary to assign new HDD as Global Hot Spare to begin rebuild process, and I did this but VD is still showing only 3 online drives and status degraded.
I didn't tray to start rebuilding process from BIOS because currently can't afford downtime.
Is there a way to start rebuilding process from within OMSA - latest version installed or I have to enter BIOS RAID controller management?
Also I've try to assign this new HDD as a Global Hot Spare and as Dedicated Hot Spare with no result.
Please advise what to do because I can't confirm that this Hot Spare will step in correctly if another HDD fails.
Physical Disk 0:0:0 Online
Physical Disk 0:0:1 Online
Physical Disk 0:0:2 Online
Physical Disk 0:0:3 Ready Global Hot Spare
theflash1932
11 Legend
•
16.3K Posts
0
March 15th, 2017 14:00
Yes. Assign the disk as a hot-spare (Global or Dedicated).
I know you said you already did, but that is all you can do, and the process is the same from the controller's BIOS utility. If it is not rebuilding, it is probably because it is not a valid hot-spare for any configured VD:
Is the drive actually showing that it is still configured a hot-spare? It actually succeeded in assigning it as a hot-spare?
mkoprivnj
2 Posts
0
March 29th, 2017 00:00
Hi,
just wanted to update you with the latest information/findings regarding my RAID5 setup.
Apparently, this RAID5 setup was with 3 HDD plus one assigned as dedicated hot spare. I've confirmed that last week when virtual disk encountered second disk failure and dedicated hot spare did its job.
Rebuilding process started automatically and the server is up and running.
Next, I took this failed disk out from bay2 and put it back in after approximately two minutes (remember this disk was marked as failed) but after putting it back in the same bay the PERC 6/i controller recognizes this previously failed disk as OK and start rebuilding the array. Surprisingly everything went well and the VD is healthy again.
My question is how is this situation possible, and how PERC controller marks HDD as failed and after two minutes of inactivity marks the same disk as OK and rebuild it back to existing VD array?
Olso I'm considering replacing the RAID5 array with RAID10 and this are steps I'm planning to perform:
1. make complete system image
2. destroy existing RAID5
3. create RAID10 array with four disks
5. restore system from image to newly created VD
All suggestions are well appreciated.
Thanks.