Question Details

Browse

linux file system corruption

By Terry Biscane - Aug. 01, 2008

"How can I recover from the following situation?
After rebooting a Linux 4 system, we are now seeing the following error message on the screen:

ext2fs partition type 0x83. ....
ext3-fs: journal inode is deleted....
mount: error 22 mounting ext3
error 2 mounting name
switchroot: mountfailed:22
umount/initrd/dev failed:2
kernal panic - not syncing.

Any idea how this happened and or how to fix it?"


Answers

Add Answer
  1. By Stan Dupont on Aug. 01, 2008

    You will need to boot from either your disaster recovery boot media or from installation media and try to run e2fsk on the filesystem(s) that are damaged.


    It appears (though there is not enough information) that the beginning of the disk has crashed. It looks like bad blocks have been encountered in the inode table, one of them being the ext3 journal file's inode. If this is the case, then your root filesystem is probably non-recoverable. You can try hitting the other copies of the inode tables in the partition, but if you don;t have the original map, it could be tough to locate them. -- If your data is stored on other filesystems, then you may be able to recover the info on those filesystems by mounting them and transferring them to either another filesystem or a tape while booted from the media mentioned above.


    This is an area I have a lot of expertise in, but it is very time consuming and one has to be in front of the actual system to perform all the actions needed to get everything that may be possible to get. You can expect to pay $100-$200 per hour for someone to provide data recovery services and, even then, it is without guarantees since the disk(s) may have truly crashed across all boundaries.


    So, if the data is valuable and you have no other copy, you may want to consider recovery services. Otherwise, it is best to try to run tests on and then re-format the drive(s) and then re-install.


    Once you have a system up and configured as you need it, the most important thing to do is to create a boot CD, backup your entire system on media of your choosing, and then boot from the CD and restore the entire system. Then you know you can get back to ground zero when needed. -- Then you need to back up the system on a regular basis so you can always get back to the last good backup in a reasonable amount of time and not lose very much data... -- And if your data is really valuable, have incremental backups occurring to a second source throughout the day and create a master backup each night. -- And if it is critical data, archive off-site in another geographically remote location on a continous basis.


    Sorry you have such an issue to face, but it would occur on any OS.


    I use BackupEdge from Microlite for my master backup and disaster recovery needs. It is a commercial package. It is worth every penny for what it does in simplifying disaster recovery.


    I use rsync for a standby server that can be converted to a live server in about 30 minutes.


    For incremental backups during the day, many other factors come into play. Oracle, MySQL, etc. need snapshot services. And then you need, in addition, an outside service like Sentinel or Iron Mountain...


    0 Votes
  2. By leehj hj on Sep. 01, 2008

    grub


    0 Votes
  3. By leehj hj on Sep. 01, 2008

    i cant't help you.but i want points .sorry.


    0 Votes
  4. By Mike Farley on Sep. 03, 2008

    try system recovery cdrom. it is very usefull and will allow you to boot into a live linux distribution with tools. or you can try to get a live cdrom from the linux distrubution you used to install with.


    0 Votes
Share your knowledge