I have a RAID 1 (hardware) array, where the system crashed, and I removed the hard drives. It is my understanding that a RAID 1 drive should work as an individual drive, as the RAID data is stored at the end of the drive and will not interfere with use outside of a RAID array.

The drive is recognized as having one ext4 partition, which successfully mounts. However, when I mount to /mnt and try ls /mnt, I get:

[ 136.961505] EXT4-fs error (device sda1): htree_dirblock_to_tree:986: inode #2: block 9249: com is: bad entry in directory: rec_len is smaller than minimal - offset=0(0), inode=0, rec_len=0, name_len=0

I found a very similar issue: Server down and require manual fsck. I saw this error in dmesg

In this case, an ordinary file is erroneously marked as a directory, and the file need only be deleted. However, unlike in the example given for that question, my error does not include a directory number.

Any suggestions on how to get the data off these drives, or even some insight as to if my problem is likely similar to the one cited or not, would be greatly appreciated!


  1. The RAID array operated under Ubuntu 14.04, which was also the OS used when the given error was produced.

  2. The system running the drives crashed in a way that required a hard power down, as it was hanging in shut down. Reboot failed, and at run level 1 I first encountered this error.

  3. I copied that error over manually, so typos are possible, but I was careful and it does not appear that there are any.

  4. I can't remember for sure if the error was the same right at first, but after several rounds of fsck, this is what it was. After imaging the drives I kind of panicked about the data and didn't really keep track of what I was doing, and I am currently in the 10 hour waiting process to restore the one drive I messed with to the image of it that I took immediately after the crash.

Your Answer

By clicking “Post Your Answer”, you agree to our terms of service, privacy policy and cookie policy

Browse other questions tagged or ask your own question.