Unsolved
1 Rookie
•
3 Posts
0
44
Drive replacement is being recognized as "spare"
Hello. We have an old Dell EqualLogic PS4100 that is still in use and we're stuck on an issue where one of the replacement disks is being recognized as a spare instead of a replacement and as a result is not rebuilding automatically.
The current setup is 6x 2TB disks in RAID50. Initially we had 2 failures and those were replaced. We did them one by one, simply removing and replacing the disks. They would start rebuilding automatically. It was generally a smooth process.
Shortly after, a 3rd disk failed. We followed the same process, removing the failed disk and inserting the replacement. During the first attempt, there was an issue with the replacement disk not being approved. It was removed and the original failed disk was put back in the slot again. During the second attempt, a different replacement disk was inserted and it was detected as a "spare".
After reading some suggestions here, we tried to failover the controller. Unfortunately, the disk still remained a spare and wouldn't begin the rebuild process. Understand we need to somehow put this spare back into the raid set. Upon running raidtool, it shows the set as RAID 5 instead of RAID 50. Everywhere else (e.g., General Member Information) it's shown as RAID 50.
We haven't yet tried rebooting the whole device. We're saving that for the last resort. Is there anything else we can try? Thank you for any assistance.
Some additional info:
RAID policy is RAID 50 (per General Member Information)
Controller firmware V7.1.1 (R400572)
Driver Status: Ok
RAID LUN 0 Ok.
05 Drives (0,2,5,3,4)
RAID 5 (64KB sectPerSU)
Capacity 7,829,783,838,720 bytes
Available Drives List: 1
DELL-Marco B
Moderator
Moderator
•
3.4K Posts
0
March 21st, 2024 10:58
Hello,
please can you give me more details about the new hard disk? Which is the firmware version, where did you buy it, it is a Dell part?
It seems that this is a firmware issue, like the system doesn't recognized to rebuild.
Thanks
dwilliam62
1 Rookie
1 Rookie
•
1.5K Posts
0
March 21st, 2024 22:01
Hello,
I would make sure you are on the correct EQL arrat member. There are not enough drives showing for it to be RAID50. RAID50 is two RAID5s put together. That RAIDset shows complete, and not degraded so a rebuild is not needed on that member. Did you maybe go to the Group IP instead of the IP address of the member with the failed drives?
Regards,
Don
#iworkfordell
daz3dandconfuz3d
1 Rookie
1 Rookie
•
3 Posts
0
March 22nd, 2024 01:55
Thanks for the suggestions.
The replacement hard drive details and screenshots are below. These drives were purchased from a Dell distributor. Actually, for the first 2 failed drives, the distributor sent us several drives that were "not approved". Eventually, they got the correct firmware and sent us a working drive.
For the current failed drive, we went through the same process of trial an error with different replacement drives and eventually landed on this current drive that seems to have passed the approval stage but is only being recognize as a spare. All drives the distributor sent us were the same spec. Variations in DOM and Config Code (1311 vs 1341) made no difference.
Dell Enterprise Plus Constellation ES.3
"Enterprise Plus Certified by Dell"
P/N 9ZM275-157
MODEL ST2000NM0023
F/W GE0B
Below are also some screenshots of the drives. The initial 2 drives that were successfully replaced are in slot 5 and 2. The current one we're trying to replace is in slot 1.
Regarding the IP, there's just 1 member configured here but I tried connecting through different IPs of the EQL and the raidtool reports the same results.
Meanwhile, we've requested a short downtime window from management to try a reboot. Once we try it, will report back with results. If that doesn't work, we're kind of out of ideas. Maybe we can ask the distributor to try a new drive.
Thanks again for the assistance.
(edited)
DELL-Joey C
Moderator
Moderator
•
3.1K Posts
0
March 22nd, 2024 08:36
Hi,
Yeah, probably a reboot of the storage. I'm pretty sure you have done a backup before the downtime.
The RAID status seems to be OK, has no alert. I'm sure you have already followed the process of replacing the failed drive, since you already done 2 previously.
dwilliam62
1 Rookie
1 Rookie
•
1.5K Posts
0
March 22nd, 2024 16:38
@daz3dandconfuz3d Hello, there is no need for a reboot. As you can see the RAID status is OK. I does not require the use of that spare. If you reboot nothing will change. You do not have enough drives to support RAID50. The policy selected maybe RAID50 but it is in fact only a RAID5. If it needed a spare the RAID status would say DEGRADED.
Re: Drives. It is not the model or firmware of the drives that makes them compatible. The drives are internally modified to be recognized by the firmware. You could buy idenical OEM dives and they would not work. The vendor could have gotten them from old EQL arrays. Make sure you back up your data regularly
Regards,
Don
#iworkfordell
daz3dandconfuz3d
1 Rookie
1 Rookie
•
3 Posts
0
March 26th, 2024 07:13
Sorry for the slow response. The vendor requested we try a shutdown of the SAN. We did as requested and nothing changed. The spare was still recognized as a spare. We decided to keep the spare and leave things as they are based on what the SAN reported (OK state and not degraded).
It's still a little confusing how it went from RAID50 to RAID5. Based on my understanding of RAID50 (this particular setup), it's 2x RAID5 in a stripe, and each of those RAID5 can tolerate a single failure. The initial failure of 2 disks that happened simultaneously confirms it must've really been a RAID50 because if it was a RAID5, those 2 failures would've ended it.
Thanks for all the responses and help!
dwilliam62
1 Rookie
1 Rookie
•
1.5K Posts
0
March 26th, 2024 17:01
@daz3dandconfuz3d Hello,
You are very welcome. I have never seen a R50 with a 1/2 chassis configuration. I've been support EQL since 2004.
So I can't explain what you saw. If it had gone from R50 to R5 some how you would also have gained space since R50 costs you another drive to build the 2nd RAIDset.
If your vendor has more compatible drives you could add them in and fill the chassis. EQL only supports 1/2 and Full drive configuraions. It would take a significant time to rebuild but you would have more space, better performance and very likely R50.
Yes, R50 is two R5 RAIDsets striped together
Regards,
Don
#iworkfordell