Superblock marked dirty

Good morning!
I met a problem on a FS with AIX 5.3
It's not possible to mount the FS because of a dirty superblock.
I tried few things without success. I need your help to solve my problem guys. Do you have any idea please?

Thanks a lot

drp01,/home/root # mount /GSPRES/data
Replaying log for /dev/lvdt1.
Fatal: I/O error
mount: 0506-324 Cannot mount /dev/lvdt1 on /GSPRES/data: The media is not formatted or the format is not correct.
0506-342 The superblock on /dev/lvdt1 is dirty. Run a full fsck to fix.
drp01,/home/root # fsck -y -p /GSPRES/data
The current volume is: /dev/lvdt1
Primary superblock is valid.
J2_LOGREDO:log redo processing for /dev/lvdt1
Fatal: I/O error
exec module "/sbin/helpers/jfs2/logredo64" failed.
logredo failed (rc=5). fsck continuing.
Primary superblock is valid.
*** Phase 1 - Initial inode scan
*** Phase 2 - Process remaining directories
*** Phase 3 - Process remaining files
*** Phase 4 - Check and repair inode allocation map
*** Phase 5 - Check and repair block allocation map
Block allocation map is corrupt (FIXED)
Superblock marked dirty because repairs are about to be written.
Fatal: I/O error
drp01,/home/root # dd count=1 bs=4k skip=31 seek=1 if=/dev/lvdt1 of=/dev/lvdt1
1+0 records in.
1+0 records out.
drp01,/home/root # mount /GSPRES/data
Replaying log for /dev/lvdt1.
Fatal: I/O error
mount: 0506-324 Cannot mount /dev/lvdt1 on /GSPRES/data: The media is not formatted or the format is not correct.
0506-342 The superblock on /dev/lvdt1 is dirty. Run a full fsck to fix.
drp01,/home/root # lsfs /GSPRES/data
Name Nodename Mount Pt VFS Size Options Auto Accounting
/dev/lvdt1 -- /GSPRES/data jfs2 874512384 rw,cio yes no
 
drp01,/home/root # oslevel
5.3.0.0

I found this other website kristijan.org Manually fixing dirty JFS2 filesystems but that doesn't work for me

drp01,/home/root # fsdb /GSPRES/data
 
File System: /GSPRES/data
 
File System Size: 874485288 (512 byte blocks)
Aggregate Block Size: 4096
Allocation Group Size: 1048576 (aggregate blocks)
 
> su
...
[12] s_state: 0x00000002 [31] s_fpack: 'lvdt1'
FM_DIRTY [32] s_fname: ''
...
display_super: [m]odify, napshot info or e[x]it: m
Please enter: field-number value > 12 0x0
 
[12] s_state: 0x00000000 [31] s_fpack: 'lvdt1'
FM_CLEAN [32] s_fname: ''

fsck doesn't work after that

This is your clue................

To run an extensive filesystem check there'll be another switch to add to your fsck command which will then cause it to check everything (which a normal fsck doesn't do). Look at your fsck man pages to see what that switch is. It will probably take a while to execute.

(It's been a while since I was involved with AIX so can't tell you straight out but it's the same with most flagship O/S's)

This also may help........

http://archive.rootvg.net/cgi-bin/anyboard.cgi/aix?cG=63130393&cmd=get&gV=0&p=&v=2&zu=36313039

Thanks for your answer.
I didn't find an other way to execute a full fsck... even with the man page.

And for the link, I have already done the instruction without success...

An other clue?

Hi,

You say "fsck doesn't work after that" - what output are you now getting from the fsck command, or does the fsck command not work at all?

Cheers,
ZB

Read somewhere that fsck switch -o full works on AIX too.
Dunno but worth a try.

Here the result of fsck after the modification with fsdb

drp01,b0nneMINE # fsck -y -p /GSPRES/data



The current volume is: /dev/lvdt1
Primary superblock is valid.
J2_LOGREDO:log redo processing for /dev/lvdt1
Fatal: I/O error
exec module "/sbin/helpers/jfs2/logredo64" failed.
logredo failed (rc=5).  fsck continuing.
Primary superblock is valid.
*** Phase 1 - Initial inode scan
*** Phase 2 - Process remaining directories
*** Phase 3 - Process remaining files
*** Phase 4 - Check and repair inode allocation map
*** Phase 5 - Check and repair block allocation map
Block allocation map is corrupt (FIXED)
Superblock marked dirty because repairs are about to be written.
Fatal: I/O error

---------- Post updated at 11:28 AM ---------- Previous update was at 11:27 AM ----------

Not better

drp01,b0nneMINE # fsck -y -p -o full /GSPRES/data



The current volume is: /dev/lvdt1
Primary superblock is valid.
J2_LOGREDO:log redo processing for /dev/lvdt1
Fatal: I/O error
exec module "/sbin/helpers/jfs2/logredo64" failed.
logredo failed (rc=5).  fsck continuing.
Primary superblock is valid.
*** Phase 1 - Initial inode scan
*** Phase 2 - Process remaining directories
*** Phase 3 - Process remaining files
*** Phase 4 - Check and repair inode allocation map
*** Phase 5 - Check and repair block allocation map
Block allocation map is corrupt (FIXED)
Superblock marked dirty because repairs are about to be written.
Fatal: I/O error

I'm no AIX expert but are you running fsck with the file system mounted?
If so, it won't fix the super block.

Nop, not possible to mount it, that's why I try fsck to fix it

Are there other filesystems of this VG mounted?
Please post the output of

lsvg <vgname>
lslv /dev/lvdt1

Do you have a backup of the data?
Does the errpt show more information? Maybe problems with underlying PV/hdisk involved?

I had a PVid missing on the VG.
It's solve now. Sorry I made a wrong analyse at the beginning!

Thanks all!