Fedora 16 died! UNEXPECTED INCONSISTENCY. Duplicate or bad block in use!

[wajda]

Hi Guys,

New year 2012 for me began with crashing my Fedora 16. Can anybody please help me with the this?

I've been updating Fedora 16 (x86-64) on my Dell laptop using yum when the kernel suddenly crashed badly with the black screen similar to the blue one in windows. The system became totally unresponsive so I had to hard reboot it. But since then when it started booting it fails with the message: _Fedora-16-x86_6: UNEXPECTED INCONSISTENCY; RUN fsck MANUALLY (see the attached photo for entire message)

The issue I am facing is very similar to this one, but the fsck message is a bit different. In my case it says:

dracut Warning: _Fedora-16-x86_6: Duplicate or bad block in use!
dracut Warning: _Fedora-16-x86_6: Multiply-claimed block(s) in inode 1487: 1180126 1180127
dracut Warning: _Fedora-16-x86_6: Multiply-claimed block(s) in inode 3281: 1180126 1180127
dracut Warning: _Fedora-16-x86_6: (There are 2 inodes containing multiply-claimed blocks.)
dracut Warning: _Fedora-16-x86_6: File /var/log/gdm/:0-greeter.log (inode #1487, mod time Sun Jan 1 13:40:13 2012)
dracut Warning: has 2 multiply-claimed block(s), shared with 1 file(a):
dracut Warning: _Fedora-16-x86_6: /var/lib/rpm/__db.001 (inode #3281, mod time Sun Jan 1 13:52:55 2012)

I tried to fsck /dev/sda* with no result. Some partitions are Ok, the others are reported to be in use, so fsck cannot check them. I tried touch /forsefsck, but I get the same message and the shell screen when I reboot.

My questions:
1. How could that happen that ext4 being journaling filesystem became corrupted so easily because of a stupid kernel panic followed by hard reboot?
2. [SOLVED] How to cure this (if possible) completely without any undesired consequences?
3. If I happen to reinstall the OS, can I choose better filesystem, so that I never fall into such problem again? What about BTRFS or NILFS? I heard they support snapshots allowing easily recovering in cases like this, it that true?

[GallComp]

Read the forum rules, specially the cross-posting section:

Do Not Cross-Post (Double-Post)
Posting your thread in several parts of the forum is immature. Just because it's in more than one location doesn't mean that more people are going to answer it. In fact quite the opposite normally as experienced members have little time for this sort of behaviour, staff members will close all offending threads. Posts or threads advertising other posts are also seen as cross-posting.

Your post on that thread has been deleted.

[wajda]

Ok, I just managed ti fix the issue - I realized that I've been running fsck on wrong partitions. I'm using LVM, so running it on /dev/mapper/vg-something_lv-root/ helped. It found couple of dozens problems on my partition (ext4) showing me some totally meaningless numbers of nodes that need to be fixed asking me if I agree or not. Well, I doubt I could even have a choice in this case It does some magic and the system booted normally. Yum and rpm seem to be working fine, so I assume no data was left corrupted (hopefully).

So I'm closing the question #2, but leaving the other two questions opened - about how could such corruption happen on a journaling FS and which FS would have better resistance than ext4 to such problems?

Thank you!

---------- Post added at 06:01 PM ---------- Previous post was at 05:58 PM ----------

Originally Posted by GallComp

Read the forum rules, specially the cross-posting section...

Ouch! No probs, sorry for that!