11 Posts
0
2076
Replacement SAS HDD Not Rebuilding (T310 Perc H700 RAID5)
I recently had a 600GB 15K SAS drive go into predictive fail in a 4-drive RAID 5 array. The server has the latest supported build of OMSA on the system, but the web interface is not working, so I am working with the command line, which I am much less familiar with. I pulled the predictive-fail drive from the system and replaced it with a refurbished drive. The drive is being recognized properly, but it was not added to the array. Instead it appears to be treated as a stand-alone drive. The pred-fail drive was 0:0:1, but for some reason 0:0:0 went into rebuild, not 0:0:1. The rebuild completed, but the array is still showing only the 3 drives and shows a status of "degraded".
My first though is to flag the new drive as a hot-spare (it is not seen as such) and hope it grabs it for the rebuild, or reboot the server and hope that corrects the issue, though that seems less likely to correct the issue, and more risky. Has anyone else seen this? Any suggestions would be appreciated!
DELL-Josh Cr
Moderator
Moderator
•
8.5K Posts
1
August 21st, 2018 10:00
Thanks, it looks like it is a 3 drive raid 5, currently on drives 1,2 and 3. Drive 0 is not doing anything. The controller has the warning for the firmware being out of date. https://www.dell.com/support/home/us/en/19/drivers/driversdetails?driverId=2948G
DELL-Josh Cr
Moderator
Moderator
•
8.5K Posts
1
August 20th, 2018 15:00
Hi,
Setting the drive as a hotspare is the next step and see if it will rebuild.
DELL-Josh Cr
Moderator
Moderator
•
8.5K Posts
0
August 21st, 2018 08:00
Can you post screenshots of the virtual disk and physical disk pages from OMSA?
Cybereal Tech Solutions
11 Posts
0
August 21st, 2018 08:00
Hello Josh,
Thank you for getting back to me. Oddly the issue has changed overnight.. As of yesterday, the replaced drive 0:0:1 was showing a capacity of 558.38GB with zero RAID Used Disk Space, and 558.38GB of Available RAID Disk Space, while the other three drives all showed the same capacity, but the metrics were flipped for used/available space. That gave me the impression that 0:0:1 had not been reintegrated into the array. As of this morning however, 0:0:1 is now showing the same metrics as the other drives, with fully used RAID disk space, and zero available.
To make matters more confusing, I just noticed that the size of the array is 1116.75GB, which makes no sense for a 4 x 600GB drive RAID 5 array, but makes perfect sense for a three drive RAID 5.. As you may have guessed, I am not the one who configured this server. What commands do we need to run to get to the bottom of this? Also, the array is still listed as "non-critical", so there is some sort of issue.
Thanks,
-Bob
Cybereal Tech Solutions
11 Posts
0
August 21st, 2018 10:00
PDF of requested output via OMSA CLI
Cybereal Tech Solutions
11 Posts
0
August 21st, 2018 10:00
Hi Josh,
I am stuck using the CLI (web interface is not responding). I am unable to paste the data into the forum as it's either too long (over 20k characters), produces some other random error, or just won't post. I have tried to work around this in several ways, with no success, so I have added a link (one post up) with the CLI output in PDF format.
-Bob
Cybereal Tech Solutions
11 Posts
0
August 21st, 2018 11:00
Thanks Josh- I overlooked the "Hot Spare = Global" on disk 0. When I pulled disk 1 it must have grabbed 0 for the array, which explains why that one was listed as rebuilding. At some point overnight it must have rebuilt 1 and set 0 back as the hot spare..
I'll likely leave the firmware as-is, as the server is slated for replacement sometime soon, and I don't want to risk anything further. Thanks again for your assistance- people like you are the reason Cybereal Technology Solutions is a proud Dell Partner!