LeRoy on 27 Aug 2011 17:10:39 -0700

[Date Prev] [Date Next] [Thread Prev] [Thread Next] [Date Index] [Thread Index]

Re: [PLUG] Debian unstable locking up and corrupting filesystem

Hash: SHA1

On 08/27/2011 01:14 PM, David B. van Balen wrote:
> Thanks for this, the system rescue cd was very useful. At the very
> least, I've been able to backup the latest data in my home directory,
> which was my biggest worry.
> I'm not sure if the computer will successfully restart, but I'm starting
> to think I need a new hard drive... or maybe a new computer. It's about
> 5-6 years old, which is 2-3 years less than I kept my last computer, but
> then the last one wasn't failing.
> I'm a little weary of purchasing again from Compaq/HP, for obvious
> reasons, along with the fact that I had to wait a year after buying this
> one before a SATA driver for this disk (the one that's now failing)
> showed up in the Linux kernel. I'm also not necessarily sold on another
> Dell, like my previous one, although I'm also not against it. Does
> anyone have recommendations for a good, Linux friendly company I can
> purchase a desktop computer from? Also, any thoughts on AMD vs. Intel? I
> currently have a 64bit AMD, but am thinking of going multi-core 32 bit
> for my next computer.
> Anyway, back to the original topic, it looks like it's my boot partition
> with the issues. I was able to successfully run e2fsck on my data
> partition (where my home directory is) and backup my data. However,
> running it on my boot partition resulted in:
> root@sysresccd /root % e2fsck -f -c -k -C 0 -p -v /dev/sda1
> /dev/sda1: recovering journal
> Error reading block 24432 (Attempt to read block from filesystem
> resulted in short read).
>      (i.e., without -a or -p options)
> Re-running without -p had me first ignore the error and then force a
> rewrite. After that there were various things:
> - orphaned inodes (normal after a non-clean shutdown)
> - another short read, this time for block 512046
> - a couple multiply claimed blocks shared with 1 file, so I told it to
> clone the multiply-claimed blocks (I assume that could have good or bad
> consequences, depending on the details of the situation).
> - The free blocks count was wrong for group #1 (93, counted=91) and
> group #15 (8894, counted=8896), both of which I told it to fix.

Your journal is fouled up so I would you should delete it.  See

http://linux.empulsegroup.com/removing and re-enabling journal for ext3.html

(I realize that it is not good style to use spaces in file names but the
author has made the link difficult)

To remove the journal use tune2fs
tune2fs -O ^has_journal /dev/sda1

I did not see notice how the bad blocks check portion of your fsck came
out.  If there are numerous bad blocks on a drive it is time to get a
new one.

e2fsck -f -c -k -C 0 -v /dev/sda1

Since you have already run the bad blocks check I do not think you need
to do it again.  Thus you can eliminate the -c and -k options.

After you run e2fsck use tune2fs to reinstall the journal.
tune2fs -j /dev/sda1

Before doing any of this I would suggest reading the man pages along
with the above web site.
- -- 
 Rev. LeRoy D. Cressy  mailto:leroy@lrcressy.com   /\_/\
                       http://lrcressy.com        ( o.o )
                       Phone:  215-535-4037        > ^ <
		       Cell:   267-307-3527

gpg fingerprint:  62DE 6CAB CEE1 B1B3 359A  81D8 3FEF E6DA 8501 AFEA

For info on enigmail:    http://lrcressy.com/linux/mozilla.pdf
For info on gpg:         http://www.gnupg.org/

Jesus saith unto him, I am the way, the truth, and the life:
no man cometh unto the Father, but by me. (John 14:6)
Version: GnuPG v1.4.10 (GNU/Linux)
Comment: Using GnuPG with Mozilla - http://enigmail.mozdev.org/

Philadelphia Linux Users Group         --        http://www.phillylinux.org
Announcements - http://lists.phillylinux.org/mailman/listinfo/plug-announce
General Discussion  --   http://lists.phillylinux.org/mailman/listinfo/plug