what does  this mean, and how might i resolve?

 

Feb 8 12:18:12 JDGJrMedia kernel: attempt to access beyond end of device

Feb 8 12:18:12 JDGJrMedia kernel: md5: rw=1, want=1953525016, limit=1953522992

Feb 8 12:18:12 JDGJrMedia kernel: REISERFS: abort (device md5): Journal write error in flush_commit_list

Feb 8 12:18:12 JDGJrMedia kernel: REISERFS: Aborting journal for filesystem on md5

 

info for that drive:

OK /dev/md5 /mnt/disk5 /dev/sdh Hitachi_HDS721010KLA330_GTF002PAJSEGPF 31°C 86806 77612 0 1.00T 658.62G 66% 341.55G

I stopped the array and restarted. Now the drive is reported as 'unformatted'.

 

do i need to take the array down until i can get a new physical drive and rebuild the data?

When you stopped the array, something was still open on that drive, still being used in the background, and therefore the unmount did not complete.  That is the usual explanation when a restart unexpectedly turns up an "Unformatted" drive, and the answer always is to shutdown and reboot, which forces everything to close.  On reboot, you should be back up and running.  If that drive or the parity drive are still not green, you may need the Trust My Array procedure to return to all green, and it does it rather quickly.

 

Here is a relevant FAQ entry:  http://lime-technology.com/wiki/index.php?title=FAQ#Why_is_a_disk_showing_as_Unformatted.3F

 

As to what happened, the size difference is about 2MB (2024KB), *almost* equal to a typical HPA, carved out at the end of the drive by some motherboards (particularly Gigabyte) when they first see the board, probably for a restorable copy of the BIOS.  However, I have never seen this happen *after* a drive has already been in use, fully partitioned.  I would check the file system on this drive with reiserfsck, see Check Disk Filesystems.

This drive has been running for about 20 days, and contains approx 600G of data.

 

the check command returned this:

 

bread: Cannot read the block (244190637): (Invalid argument).

 

reiserfs_open: Your partition is not big enough to contain the

filesystem of (244190637) blocks as was specified in the found super block.

Failed to open the filesystem.

 

If the partition table has not been changed, and the partition is

valid  and  it really  contains  a reiserfs  partition,  then the

superblock  is corrupted and you need to run this utility with

--rebuild-sb.

 

do you think it is safe to rewrite the superblock?

Can you provide a syslog for me first?  Also, is this a Gigabyte motherboard?  The partition messages still look like a forced HPA.  Are you technically comfortable with tools like HDAT2?  Haven't had a chance to use it myself, but it should allow you to remove the HPA, if that proves to be what has happened.  Can you think of anything unusual you may have done recently, that involved this drive?  Have you made any changes in the BIOS settings?

It is a gigabyte motherboard. No hardware/bios changes in 3+ weeks. I was renaming a lot of files from my Vista box today, when I started to get 'you need permissions' messages.

 

I went ahead and rebuilt the superblock, and it then asked to rebuild the tree. Both completed successfully and the drive is back on-line.

 

thanks for getting me started.