Drive Red Balling, Can't figure out why!

[guyonphone]

Hello Everyone,

 

One of my seagate 3 TB drives is red balling daily. I find myself having to constantly create a "new config" and re-add the drives in proper order because of this. The problem is that the drive appears to be perfectly healthy.

 

Attached are:

1. My smart test.

 

2. A Partial Syslog that shows that the drive is having issues writing to a sector.  ( i would have posted a full syslog but it exceeds the size limit) if you need to see more of it, i can post more.

 

3. Pic of Seagate Passed Tests

http://i.imgur.com/SqR4lDj.jpg

 

I have a norco 4224 case, so the drive is plugged into a norco backplane with a SAS 8087 to SAS 8087 cable, and is plugged into a AOC-SASLP-MV8. I am running a seasonic X-1250w power supply, so I know this is not a power issue.

 

My Unraid stats are:

Unraid 5.04

SimpleFeatures 1.011

Sickbeard

Sabnzbd

Air Video

Plex

 

Any help would be greatly appreciated as I cannot figure out what is going on here.

Partial_Syslog.txt

smart.txt

[redlaws]

surely you should replace that drive to prove that you have no more issues, check all leads & plugs.

 

I would not be mucking around if you replace it with a new drive & it rebuild ok then problem is solved.

if another reballs then you could lose lots of data.

[guyonphone]

I think you are partially correct, if another drive for lack of a better term "actually failed", i would just rebuild using a "new config", replacing the other drive, telling the array that parity was valid, worst case scenario is I might loose files that had been freshly written, but I wouldn't loose a whole lot. But You are correct in the sense, that still is not a risk i really want to be taking. It just basically means that I have to replace a drive which tests good, and that doesn't sit right with me. Maybe Ill try switching out the cable first and see if that has any affect.

 

Basically I would have to eat the cost for this drive, as how could I convince seagate to take it back if I can't prove it's bad :-(

 

One thing i forgot to mention, is that every time that I have been creating a "new config" i have been rebuilding the parity fresh. And the drive makes it through the parity rebuild every time, then ill check the next day and it will have red balled again. On top of all that, the drive has 45 gigs free, but it's actually not handling any of my new writes, i think it's just having problems reading a sector, or rather thats what the syslog says. If that was the case, shouldn't that come up in a smart attribute or the seagate long scan?

 

Since it isn't, that makes me believe that it might be a transitive issue. I would really like to pinpoint the actual problem.

[dheg]

Make sure the disk tray closes fine. I had some trouble with the 4224 because of loose trays. Make sure you push them tight!

 

Sent from my GT-I9305 using Tapatalk

 

[guyonphone]

I have changed the cable and attached the drive to a different Controller card. If the drive still fails, after that, i will just replace the drive. Otherwise i will assume that it was either a connection, cable, or controller card problem.

 

Thanks for you help redlaws and dheg. Ill report back here if I have problems again after a full parity check.

[dgaschk]

There is no requirement to prove a drive is bad. All vendors will RMA any covered drive.

[guyonphone]

Still having issues with the drive,  Im going to try and RMA it.

 

Dec 17 06:50:19 Unraid kernel: sd 1:0:1:0: [sdi] CDB:

Dec 17 06:50:19 Unraid kernel: cdb[0]=0x8a: 8a 00 00 00 00 00 fc 04 e0 b8 00 00 00 08 00 00

Dec 17 06:50:19 Unraid kernel: end_request: I/O error, dev sdi, sector 4228178104

Dec 17 06:50:19 Unraid kernel: end_request: I/O error, dev sdi, sector 4228178104

Dec 17 06:50:19 Unraid kernel: md: disk13 write error, sector=4228178040

Dec 17 06:50:19 Unraid kernel: md: recovery thread woken up ...

Dec 17 06:50:19 Unraid kernel: md: recovery thread has nothing to resync

Dec 17 09:13:21 Unraid shfs/user0: shfs_rmdir: rmdir: /mnt/disk15/Completed/Movies/_UNPACK_White House Down 2013 COMPLETE BLURAY-PCH cp(tt2334879) (39) Directory not empty

Dec 17 09:13:21 Unraid shfs/user0: shfs_rmdir: rmdir: /mnt/disk15/Completed/Movies/_UNPACK_White House Down 2013 COMPLETE BLURAY-PCH cp(tt2334879) (39) Directory not empty

Dec 17 09:13:40 Unraid shfs/user: shfs_readdir: fstatat: Alice in wonderland (3DBD) (2) No such file or directory

Dec 17 09:13:40 Unraid shfs/user: shfs_readdir: readdir_r: /mnt/disk13/Media/Movies (2) No such file or directory

Dec 17 09:14:22 Unraid shfs/user0: shfs_rmdir: rmdir: /mnt/disk15/Completed/Movies/White House Down 2013 COMPLETE BLURAY-PCH cp(tt2334879) (39) Directory not empty

Dec 17 09:14:31 Unraid shfs/user0: shfs_rmdir: rmdir: /mnt/disk15/Completed/Movies/White House Down 2013 COMPLETE BLURAY-PCH cp(tt2334879) (39) Directory not empty

Dec 17 09:15:04 Unraid shfs/user: shfs_readdir: fstatat: Alice in wonderland (3DBD) (2) No such file or directory

Dec 17 09:15:04 Unraid shfs/user: shfs_readdir: readdir_r: /mnt/disk13/Media/Movies (2) No such file or directory

Dec 17 09:15:34 Unraid shfs/user: shfs_readdir: fstatat: Alice in wonderland (3DBD) (2) No such file or directory

 

Do I also need to run a "reiserfsck --fix-fixable"?

[dgaschk]

See Check Disk Filesystems in my sig. Read the instructions fully and follow them carefully.