Running 4.7-beta1

 

Rebooted the server and array wont start now.  Looked back in the log file from before it shutdown and saw these two lines

 

Jan 23 14:01:45 Tower kernel: REISERFS error (device md20): reiserfs-2025 reiserfs_cache_bitmap_metadata: bitmap block 32768 is corrupted: first bit must be 1

Jan 23 14:01:45 Tower kernel: REISERFS (device md20): Remounting filesystem read-only

 

Drive is a Samsung F4 with firmware update and precleared with -A, if i'm assuming that md20 is the same as disk20.  The drive is empty so not worried about data on the drive

 

Also note that when I remove disk20 unraid simulates the disk and the array starts up

 

Attached is the old syslog, when I shutdown the server, and new is the log that just booted up

 

Thanks

 

Josh

syslog-old.zip

syslog-new.txt

The file-system on that disk is corrupted.  You just need to use the procedure described here to fix it:

http://lime-technology.com/wiki/index.php?title=Check_Disk_Filesystems

 

Yes, md20 is disk20.

You would use reiserfsck on /dev/md20

 

Joe L.

 

 

Josh -

 

Wondering what happened to cause this.  I know you ran a successful test with the F4 for 4.7b1 - does this indicate that maybe there is a problem?

Not sure, making me a little nervous.  Time stamp on the error at the time I was running a script that lists the files on all disks, combination of the find and ls command

Not sure, making me a little nervous.  Time stamp on the error at the time I was running a script that lists the files on all disks, combination of the find and ls command

that would not have caused the corruption.  It would have found the corruption when the file system was detected by the OS as being corrupted.

 

Are you sure the firmware patch was successfully applied to the drive?  Did you have any abrupt power downs?  Or do you have any un-readable sectors on that disk?  (That would do it too)  Or bad/improperly configured memory strips. (Wrong voltage/timing/clock speed)

okay per wiki

 

root@Tower:~# cd

root@Tower:~# samba stop

root@Tower:~# umount /dev/md20

umount: /dev/md20: not found

 

there are no md anything in my /dev directory

 

however unraid is giving me the option to rebuild the disk now.  Should I rebuild?

syslog.txt

okay per wiki

 

root@Tower:~# cd

root@Tower:~# samba stop

root@Tower:~# umount /dev/md20

umount: /dev/md20: not found

 

there are no md anything in my /dev directory

 

however unraid is giving me the option to rebuild the disk now.  Should I rebuild?

Because you un-assigned the drive  (ore removed it) and booted without it unRAID is now going to re-construct it.

 

You can let it, but when it is done and your array is back online, you'll probably still need to fix the file-system as described earlier before you can write to the disk20 drive.

(The re-constructed version will likely have the same corruption)

 

Joe L.