[grundig@teleline.es: Re: Reiserfs corruption on 2.4.17-rc1!]

From: Diego Calleja (grundig@teleline.es)
Date: Tue Dec 18 2001 - 16:11:22 EST


On Tue, 18 Dec 2001 22:08:28 Diego Calleja wrote:

On Tue, 18 Dec 2001 12:58:46 Nikita Danilov wrote:
> Diego Calleja writes:
> > Well, I've run badblocks in 2.4.16
> > results:
> >
> >

> > Dec 18 00:22:41 diego kernel: is_tree_node: node level 18771 does not
> match
> > to the expected one 1
> > Dec 18 00:22:41 diego kernel: vs-5150: search_by_key: invalid format
> found
> > in block 10667. Fsck?
> > Dec 18 00:22:41 diego kernel: vs-13070: reiserfs_read_inode2: i/o
> failure
> > occurred trying to find stat data of [4160 68377 0x0 SD]
> >
> > This is my opinion:
> > -Something (reiserfs, anything) has caused fs corruption
> > -It should be repaired by reiserfsck, but it's broken :-((
>
> You mean you cannot access /sbin/reiserfsck, because root file system is
> unstable? Sorry, probably I missed start of this thread on lkml.

I mean this:
root@diego:~# reiserfsck /dev/hdc5

<-------------reiserfsck, 2001------------->
reiserfsprogs 3.x.0j

Will read-only check consistency of the partition
Will put log info to 'stderr'
Do you want to run this program?[N/Yes] (note need to type Yes):Yes
Analyzing journal..nothing to replay (no transactions older than last
flushed one found)
Fetching on-disk bitmap..done
Checking S+tree../ 1 (of 2)/ 3 (of 110)/ 15 (of 149)node (10667) with
wrong level (18771) found in the tree (should be 1)
Segmentation fault
root@diego:~#

>
> > -This corruption should NOT have happened, reiserfsck
> shouldn't
> > have to be used.
>
> Unfortunately this looks like "standard" reiserfs tree corruption.
> Take
>
> ftp://ftp.namesys.com/pub/reiserfsprogs/pre/reiserfsprogs-3.x.0k_pre11.tar.gz,
>
> build it and run reiserfsck from there. If your root file system is
> dead, you will have to do this from some other media (like floppy).
Well, I've a copy of all data in another partition...I'm running from it.
I've downloaded that version....well, it don't crash, here's output:
root@diego:~/reiserfsprogs-3.x.0k_pre11/fsck# ./reiserfsck -i /dev/hdc5

<-------------reiserfsck, 2001------------->
reiserfsprogs 3.x.0k_pre11

Will read-only check consistency of the filesystem on /dev/hdc5
Will put log info to 'stdout'

Do you want to run this program?[N/Yes] (note need to type Yes):Yes
###########
reiserfsck --check started at Tue Dec 18 21:59:12 2001
###########
Replaying journal..
0 transactions replayed
reserved=8193
Checking S+tree../ 1 (of 2)/ 3 (of 110)/ 15 (of 149)node (10667) with
wrong level (18771) found in the tree (should be 1)
whole subtree skipped
ok
Comparing bitmaps..free block count 1970606 mismatches with a correct one
1971004.
on-disk bitmap does not match to the correct one. 91 bytes differ
Bad nodes were found, Semantic pass skipped
There were found 3 corruptions which can be fixed only during
--rebuild-tree
###########
reiserfsck finished at Tue Dec 18 22:06:23 2001
###########

If I do --rebuild tree, will I lose my data?

>
> > -I'm not a kernel hacker, so I can't try anything...what I
> know is
> > that
> > /etc in hc5 doesn't work. /usr, /var....works
> correctly.
> >
> > Well, I'd like to know what's happened in my drive. Can somebody try
> to
> > give an explanation?
> >
> > Diego Calleja
>
> Nikita.
Diego Calleja

-
To unsubscribe from this list: send the line "unsubscribe linux-kernel" in
the body of a message to majordomo@vger.kernel.org
More majordomo info at http://vger.kernel.org/majordomo-info.html
Please read the FAQ at http://www.tux.org/lkml/



This archive was generated by hypermail 2b29 : Sun Dec 23 2001 - 21:00:18 EST