unlinked inode, aborted journal and read only file system

[zigzam]

I am running centos 5 on adaptec 2405 raid 10

 I get this error:
EXT3-fs error (device sda8): ext3_lookup:
unlinked inode 55804731 in dir #55804693
Aborting journal on device sda8.
ext3_abort called.
EXT3-fs error (device sda8):
ext3_journal_start_sb: Detected aborted journal
Remounting filesystem read-only 
  

How do I fix this?


Tried

 tune2fs -O ^has_journal /dev/sda8
 
fsck -a /dev/sda8 (this immediately returned that the filesystem was
clean and without error)
 
tune2fs -j /dev/sda8 
  

Thank you!

[Rubberman]

Run fsck -f /dev/sda8 to force it to check the file system, even if it appears to be "clean".

[Lakshmipathi]

yes - I would agree with Rubberman , try forcing fsck.

[zigzam]

I will give that a try. Also was told to try:

 fsck -cr /dev/sda8 
  

Is that recommended?

[Rubberman]

The -c option does a bad-block check of the disc. Not a bad idea. You can do that with the -f option also.

[zigzam]

So use:

  fsck -crf /dev/sda8 
  

With the journal on or off?

[Lakshmipathi]

With the journal on or off?

I'm not sure about option to turn on/off journal. How would you do that?

But I strongly recommend you to continue fsck with Journaling on - Since the only purpose of using journal is to recover file system from crash.So run fsck with journal on.

[Rubberman]

Actually, for fsck, turn it off. You can't fix a mounted file system anyway. An ext3 file system is just an ext2 file system w/ journaling, which can be enabled/disabled by mounting or remounting with the appropriate options. When you run fsck against either an ext2 or ext3 file system, it actually uses the e2fsck application. Here is the relevant paragraph from the man page:

       e2fsck  is  used  to  check a Linux second extended file system (ext2fs).  E2fsck
       also supports ext2 filesystems containing a journal,  which  are  also  sometimes
       known as ext3 filesystems, by first applying the journal to the filesystem before
       continuing with normal e2fsck processing.  After the journal has been applied,  a
       filesystem will normally be marked as clean.  Hence, for ext3 filesystems, e2fsck
       will normally run the journal and exit, unless its superblock indicates that fur-
       ther checking is required.

[Lakshmipathi]

Actually, for fsck, turn it off. You can't fix a mounted file system anyway. An ext3 file system is just an ext2 file system w/ journaling, which can be enabled/disabled by mounting or remounting with the appropriate options. When you run fsck against either an ext2 or ext3 file system, it actually uses the e2fsck application. Here is the relevant paragraph from the man page:

       e2fsck  is  used  to  check a Linux second extended file system (ext2fs).  E2fsck
       also supports ext2 filesystems containing a journal,  which  are  also  sometimes
       known as ext3 filesystems, by first applying the journal to the filesystem before
       continuing with normal e2fsck processing.  After the journal has been applied,  a
       filesystem will normally be marked as clean.  Hence, for ext3 filesystems, e2fsck
       will normally run the journal and exit, unless its superblock indicates that fur-
       ther checking is required.

Thanks for clarification Rubberman. But the man page says , " by first applying the journal to the filesystem before continuing with normal e2fsck processing " .

What I'm worried is , If this current corrupted file system /dev/sda8 - has previously unfinished task -- will be replayed if journal turned off ?

[zigzam]

The issue was caused due to a crash then power cycle. Maybe that will help determine the commands needed.