1 Rookie
•
35 Posts
0
876
MD3200 - Filesystem RAW after power failure
Following a power failure, Windows Disk management is showing an MD3200 Disk as "RAW" and wants to format it. However, everything looks good on MDSM, and aChkDsk identifies it as NTFS (but throws a lot of "Arribute Record Corrupt" errors). How do I go about getting Disk management to recognize it as NTFS?
Suggestions on the net include running Convert x: /fs:ntfs, and chkdsk x: /f (but I am leery of issuing any command that writes to the disk without some expert advice).
DiegoLopez
4 Operator
4 Operator
•
2.7K Posts
1
May 3rd, 2022 07:00
Hello @CU.Dell.User,
I think there is more to see in MDSM.. it should be advised to have the logs checked by Dell (if the MD3200 is still under warranty). Maybe some disks were affected.
If you are having "Arribute Record Corrupt" this could mean there are several data corrupt in the hard drives. You can try chkdsk c: /r to see if this fix the errors.. Otherwise you will need to maybe change the hard drive with errors and restore the data from backup.
Regards.
CU.Dell.User
1 Rookie
1 Rookie
•
35 Posts
0
May 3rd, 2022 08:00
Thank you Diego. You are right. Even though immediately after reboot, MDSM looked fine, overnight it identified a disk as failed, and is rebuilding. I will wait for the rebuild to complete and then try chkdsk c: /r and provide an update. (the unit is not under warranty, -unfortunately :-).
CU.Dell.User
1 Rookie
1 Rookie
•
35 Posts
0
May 7th, 2022 09:00
Closure: Once the Rebuild was finished, MDSM looked fine again (except for the "Failed Disk Pool Physical Disk" message). But, windows could not access the drive & the "Attribute Record Corrupt" errors were still present.
The problem was resolved with CHKDSK x: /f. It was fairly quick & done in a few minutes. Afterward, I ran a CHKDSK x: /r (which implies /f so it would have worked too. But /r is obviously very slow -took about 15 hours for this ~12TB drive). It did not find any bad sectors but reported another message ("CHKDSK discovered free space marked as allocated in the volume bitmap") and corrected that.
I am still puzzled why the disk failed, and whether it was coincidental, or whether the power failure caused it. There were never any warnings of impending failure before this episode & both batteries (in the controllers) are Optimal. And, even after the rebuild was complete windows could not access the drive until a CHKDSK /F was run. Maybe that is normal? Well, the good news is everything seems to be working fine now.
DELL-Joey C
Moderator
Moderator
•
3.4K Posts
0
May 8th, 2022 18:00
Hi @CU.Dell.User,
We may not certainly know for sure what's the main root cause, unless the engineers have a good look at the storage logs. To my opinion, it's the power failure that might be the root cause. A sudden surge of power can cause other components to fail not only the disks.
CU.Dell.User
1 Rookie
1 Rookie
•
35 Posts
0
May 10th, 2022 13:00
Just so the future readers have the full picture, the enclosure was powered via an APC UPS (which shuts down the server via a USB signal but it has no way of telling the MD3200 to shut down -AFAIK?)
So, I guess/hope there would not have been a power surge (am I trusting these $200 UPS too much?) On top of that these controllers have battery backup, so if the power is lost should the battery not do an orderly shutdown.