On Wed, 2005-09-07 at 20:34 -0500, Matthew S. Hallacy wrote:
> Hello, 
> 
>  After an unclean shutdown I have a jfs partition that will not properly fsck:
> 
> processing started: 9/6/2005 21.18.36
> 
> FSCK  Device /dev/md0 is currently mounted READ ONLY.
> The current device is:  /dev/md0
> Open(...READ/WRITE EXCLUSIVE...) returned rc = 0
> Primary superblock is valid.
> The type of file system for the device is JFS.
> Block size in bytes:  4096
> Filesystem size in blocks:  2148376576
> **Phase 0 - Replay Journal Log
> LOGREDO:  Log superblock contains invalid magic number.
> logredo failed (rc=-268).  fsck continuing.

This is not good.  The journal is located at the end of the volume.
Could there be a hw or raid problem?  Any syslog errors?

A failure to replay the journal could account for some of the errors
that were found.

> **Phase 1 - Check Blocks, Files/Directories, and  Directory Entries
> Extended attributes for file set object DF22494908 have an invalid format.
> Extended attributes for file set object DF22495041 have an invalid format.
> Extended attributes for file set object DF22495051 have an invalid format.
> Extended attributes for file set object DF22495053 have an invalid format.
> Extended attributes for file set object DF22495055 have an invalid format.
> Extended attributes for file set object DF22495056 have an invalid format.
> ......|..
> Duplicate reference to 1 block(s) beginning at offset 275215147 found in file 
> system object FF35544170.
> Inode F35544170 has references to cross linked blocks.
> File system object FF35544170 has corrupt data (39).
> Duplicate reference to 1 block(s) beginning at offset 1677788153 found in 
> file system object DF55240801.
> Inode F55240801 has references to cross linked blocks.
> Duplicate reference to 1 block(s) beginning at offset 1677788152 found in 
> file system object DF55240801.
> Inode F55240801 has references to cross linked blocks.
> File system object DF55240801 has a node with an invalid self field.
> Duplicate reference to 1 block(s) beginning at offset 1677788154 found in 
> file system object DF55240810.
> Inode F55240810 has references to cross linked blocks.
> File system object DF55240810 has a node with an invalid self field.
> Duplicate reference to 1 block(s) beginning at offset 1677788159 found in 
> file system object DF55240833.
> Inode F55240833 has references to cross linked blocks.
> Duplicate reference to 1 block(s) beginning at offset 1677788158 found in 
> file system object DF55240833.
> Inode F55240833 has references to cross linked blocks.
> File system object DF55240833 has a node with an invalid self field.
> File system object DF67214016 has a node with an invalid self field.
> Invalid data (43) detected in file system object DF67746400.
> File system object DF67746400 has a node with an invalid self field.
> Invalid data (43) detected in file system object DF68029056.
> File system object DF68029056 has a node with an invalid self field.
> Insufficient dynamic storage available for required workspace (1,6).  CANNOT 
> CONTINUE
> processing terminated:  9/7/2005 0:20:59  with return code: -10097  exit 
> code: 8.
> 
> 
> The hardware in question is two 12 disk 3ware arrays RAID-0'd together via 
> software raid, I am able to mount it read-only and nothing appears to be
> corrupted
> 
> /dev/md0              8.1T  2.9T  5.2T  36% /home
> 
> As you can see, there is plenty of disk space available
> 
> free -m
>              total       used       free     shared    buffers     cached
> Mem:          3040       1106       1933          0        820         11
> -/+ buffers/cache:        275       2764
> Swap:         6999          2       6997
> 
> There's also plenty of RAM, and a lot more swap. I'm not quite sure what 
> fsck.jfs
> is referring to when it says 'Insufficient dynamic storage' =/

32-bit hardware?  I wonder if fsck's address space is exhausted.  I
think I need to look at improving jfs_fsck's memory usage.  The memory
usage goes up with the number of inodes, and I think you have a lot of
inodes.

> I would appreciate any ideas about how to fix this, shuffling 3TB of data 
> around to reformat the array would be quite a pain (and before anyone asks,
> this _is_ the backup system =)

Okay.  I found a place to save some memory by removing some OS/2 cruft
from fsck.jfs.  A source tarball should be showing up soon at:
http://kernel.org/pub/linux/kernel/people/shaggy/jfsutils-1.1.8.1.tar.gz

Let's hope that it helps enough to let fsck complete for you.

> -- 
> Matthew S. Hallacy 
> http://www.poptix.net                           GPG public key 0x01938203

Shaggy
-- 
David Kleikamp
IBM Linux Technology Center



-------------------------------------------------------
SF.Net email is Sponsored by the Better Software Conference & EXPO
September 19-22, 2005 * San Francisco, CA * Development Lifecycle Practices
Agile & Plan-Driven Development * Managing Projects & Teams * Testing & QA
Security * Process Improvement & Measurement * http://www.sqe.com/bsce5sf
_______________________________________________
Jfs-discussion mailing list
[email protected]
https://lists.sourceforge.net/lists/listinfo/jfs-discussion

Reply via email to