file corruption, advice needed

file corruption, advice needed

[Christian]

Hi,

recently i noticed some corrupt (only 1 really, about 9 MB big) or even 
disappeared (say a dozen or perhaps more).
i run reiserfsck on the very partition while it was mounted RO, no 
corruptions were found. should i trust it (and blame some application 
for the messed up files) or should i try with --rebuild-sb (needs a 
--rebuild-tree afterwards, right?) ?

i have to admit, that i had "not so good" experience with --rebuild-* 
operations, but this was 1 or 2 years ago.

so, can you give me advice on this? should i still backup this parition?
(it's my root ( / ) partition, so it's not that important)

this is with Debian/stable i386 (AMD K7 SMP), reiserfsprogs 3.6.11.
debugreiserfs on this rw mounted parition shows:

--------
root@atlant:~# debugreiserfs /dev/sda1
debugreiserfs 3.6.11 (2003 www.namesys.com)


Filesystem state: consistency is not checked after last mounting

Reiserfs super block in block 16 on 0x801 of format 3.6 with standard 
journal
Count of blocks on the device: 2239484
Number of bitmaps: 69
Blocksize: 4096
Free blocks (count of blocks - used [journal, bitmaps, data, reserved] 
blocks): 680307
Root block: 36301
Filesystem is NOT cleanly umounted
Tree height: 5
Hash function used to sort names: "r5"
Objectid map size 26, max 972
Journal parameters:
         Device [0x0]
         Magic [0x6961a7e1]
         Size 8193 blocks (including 1 for journal header) (first block 18)
         Max transaction length 1024 blocks
         Max batch size 900 blocks
         Max commit age 30
Blocks reserved by journal: 0
Fs state field: 0x0:
sb_version: 2
inode generation number: 5996064
UUID: 00000000-0000-0000-0000-000000000000
LABEL:
Set flags in SB:
root@atlant:~#
-----------

i can provide more information / fsdumps if there is interest. but i 
think it's rather hard to do anything now, because i don't even know if 
it is a filesystem thing anyway.

Thank you for your time,
Christian.

Re: file corruption, advice needed

[Vitaly Fertman]

On Thursday 02 October 2003 14:11, Christian wrote:
> Hi,

Hi

> recently i noticed some corrupt (only 1 really, about 9 MB big) or even
> disappeared (say a dozen or perhaps more).
> i run reiserfsck on the very partition while it was mounted RO, no
> corruptions were found. should i trust it (and blame some application
> for the messed up files) or should i try with --rebuild-sb (needs a
> --rebuild-tree afterwards, right?) ?

if reiserfsck does not find any corruption that means that reiserfs on-disk
structures are consistent. no rebuild operation is needed.


> i have to admit, that i had "not so good" experience with --rebuild-*
> operations, but this was 1 or 2 years ago.
>
> so, can you give me advice on this? should i still backup this parition?
> (it's my root ( / ) partition, so it's not that important)

reiserfsck is considered as stable now, but backups are always useful, 
thus you would have this dozen of files backuped now.


-- 
Thanks,
Vitaly Fertman


> this is with Debian/stable i386 (AMD K7 SMP), reiserfsprogs 3.6.11.
> debugreiserfs on this rw mounted parition shows:
>
> --------
> root@atlant:~# debugreiserfs /dev/sda1
> debugreiserfs 3.6.11 (2003 www.namesys.com)
>
>
> Filesystem state: consistency is not checked after last mounting
>
> Reiserfs super block in block 16 on 0x801 of format 3.6 with standard
> journal
> Count of blocks on the device: 2239484
> Number of bitmaps: 69
> Blocksize: 4096
> Free blocks (count of blocks - used [journal, bitmaps, data, reserved]
> blocks): 680307
> Root block: 36301
> Filesystem is NOT cleanly umounted
> Tree height: 5
> Hash function used to sort names: "r5"
> Objectid map size 26, max 972
> Journal parameters:
>          Device [0x0]
>          Magic [0x6961a7e1]
>          Size 8193 blocks (including 1 for journal header) (first block 18)
>          Max transaction length 1024 blocks
>          Max batch size 900 blocks
>          Max commit age 30
> Blocks reserved by journal: 0
> Fs state field: 0x0:
> sb_version: 2
> inode generation number: 5996064
> UUID: 00000000-0000-0000-0000-000000000000
> LABEL:
> Set flags in SB:
> root@atlant:~#
> -----------
>
> i can provide more information / fsdumps if there is interest. but i
> think it's rather hard to do anything now, because i don't even know if
> it is a filesystem thing anyway.
>
> Thank you for your time,
> Christian.

Re: file corruption, advice needed

[Christian]

Vitaly Fertman wrote:

> if reiserfsck does not find any corruption that means that reiserfs on-disk
> structures are consistent. no rebuild operation is needed.
> 

[....]

>>so, can you give me advice on this? should i still backup this parition?
>>(it's my root ( / ) partition, so it's not that important)
> 
> 
> reiserfsck is considered as stable now, but backups are always useful, 
> thus you would have this dozen of files backuped now.

hm, yes, that makes sense. i am already suspecting some application, but 
i'll watch out for further corruptions.

Thank you for the quick (!!) response,
Christian.

file corruption, advice needed

[Christian]

 > this is with Debian/stable i386 (AMD K7 SMP), reiserfsprogs 3.6.11.

oh, i forgot: kernel 2.4.21-5-k7-smp (Debian Kernel) is running. this 
seems to be vanilla with some patches applied, details available here:

http://nerdbynature.de/bits/README.Debian.1st.gz

Thanks,
Christian.