I had a hardware problem with one of my sata controllers that appears to have corrupted one drive.  Fortunately the only one on that controller.

 

I've fixed the hardware problem and restarted unraid but get the following reiserfs errors . . .

 

Nov 11 11:04:16 Tower emhttp: disk6 mount error: 32

Nov 11 11:04:16 Tower kernel: REISERFS warning: reiserfs-5090 is_tree_node: node level 0 does not match to the expected one 3

Nov 11 11:04:16 Tower kernel: REISERFS error (device md6): vs-5150 search_by_key: invalid format found in block 32941. Fsck?

Nov 11 11:04:16 Tower emhttp: shcmd (124): rmdir /mnt/disk6

Nov 11 11:04:16 Tower kernel: REISERFS (device md6): Remounting filesystem read-only

Nov 11 11:04:16 Tower kernel: REISERFS error (device md6): vs-13070 reiserfs_read_locked_inode: i/o failure occurred trying to find stat data of [1 2 0x0 SD]

 

I stopped the array, uninstalled that drive in Devices,  then closed down and physically removed the drive from the computer, rebooted the system and still get the error, even though the drive is not physically present.  On the main unraid screen the drive shows as not installed and unformatted.

 

So what do I do now?  I assume the parity is corrupt and my only option is to format the drive, rebuild parity and start again.  Am I right?  Is my data toast?

 

Full syslog attached

syslog-2010-11-11.txt

I had a hardware problem with one of my sata controllers that appears to have corrupted one drive.  Fortunately the only one on that controller.

 

I've fixed the hardware problem and restarted unraid but get the following reiserfs errors . . .

 

Nov 11 11:04:16 Tower emhttp: disk6 mount error: 32

Nov 11 11:04:16 Tower kernel: REISERFS warning: reiserfs-5090 is_tree_node: node level 0 does not match to the expected one 3

Nov 11 11:04:16 Tower kernel: REISERFS error (device md6): vs-5150 search_by_key: invalid format found in block 32941. Fsck?

Nov 11 11:04:16 Tower emhttp: shcmd (124): rmdir /mnt/disk6

Nov 11 11:04:16 Tower kernel: REISERFS (device md6): Remounting filesystem read-only

Nov 11 11:04:16 Tower kernel: REISERFS error (device md6): vs-13070 reiserfs_read_locked_inode: i/o failure occurred trying to find stat data of [1 2 0x0 SD]

 

I stopped the array, uninstalled that drive in Devices,  then closed down and physically removed the drive from the computer, rebooted the system and still get the error, even though the drive is not physically present.  On the main unraid screen the drive shows as not installed and unformatted.

 

So what do I do now?  I assume the parity is corrupt and my only option is to format the drive, rebuild parity and start again.  Am I right?  Is my data toast?

 

Full syslog attached

 

See this post.  He had the exact same problem and was able to fix it.

I had a hardware problem with one of my sata controllers that appears to have corrupted one drive.  Fortunately the only one on that controller.

 

I've fixed the hardware problem and restarted unraid but get the following reiserfs errors . . .

 

Nov 11 11:04:16 Tower emhttp: disk6 mount error: 32

Nov 11 11:04:16 Tower kernel: REISERFS warning: reiserfs-5090 is_tree_node: node level 0 does not match to the expected one 3

Nov 11 11:04:16 Tower kernel: REISERFS error (device md6): vs-5150 search_by_key: invalid format found in block 32941. Fsck?

Nov 11 11:04:16 Tower emhttp: shcmd (124): rmdir /mnt/disk6

Nov 11 11:04:16 Tower kernel: REISERFS (device md6): Remounting filesystem read-only

Nov 11 11:04:16 Tower kernel: REISERFS error (device md6): vs-13070 reiserfs_read_locked_inode: i/o failure occurred trying to find stat data of [1 2 0x0 SD]

 

I stopped the array, uninstalled that drive in Devices,  then closed down and physically removed the drive from the computer, rebooted the system and still get the error, even though the drive is not physically present.  On the main unraid screen the drive shows as not installed and unformatted.

 

So what do I do now?  I assume the parity is corrupt and my only option is to format the drive, rebuild parity and start again.  Am I right?  Is my data toast?

 

Full syslog attached

 

See this post.  He had the exact same problem and was able to fix it.

An un-assigned drive is treated EXACTLY the same as a failed drive or a drive that is disconnected.  It is STILL part of the disk configuration and will be simulated by parity and the remaining data drives.

 

It is almost guaranteed that if the actual drive has file system corruption that the simulated drive will have the same corruption.

 

The solution is to run reiserfsck on the damaged file system and fix the corruption.  If you do it with the physical disk disconnected, or logically un-assigned, the fix will be applied to the physical drive when you eventually re-assign it and re-construct the fixed contents onto it.

 

Joe L.