Partition problems

[blackbird]

This is my first post here, so I'll say Hello to all of you. I'm glad to find a great community here, and I hope to get some help here (and point others here as well).

I'd like to resurrect this thread if I may, as I have a similar dilemma. I've been using Mandrake 9.2 steadily for awhile on my 2nd ide drive, while I tinker away at my redhat 9 install on my 1st drive (planet CCRMA, installing xfce4, etc.). Mandrake is my sanctuary while I continually experiment and break my redhat. I don't really do anything risky so I never expected to find problems, and thanks to my ignorance, I never backed up my data.

 

So, my Mandy partitions: hdb1 - /boot hdb2 - / hdb3 - swap hdb5 - /usr hdb6 - /var hdb7 - /home

all except for swap (obviously) are ext3

The problem is with hdb7, no! my precious files!!

 

I wait to see if anything happens but nothing. I try no and I get error! and I either log in my superuser password or type ctrl-D. So I try to e2fsck -b 8193 /dev/hdb7 and I just get the error message repeated, by some process of elimination I get to e2fsck -b 32768 /dev/hdb7 and then I get

 

ext3 recovery flag clear, but journal has data. Recovery flag not set in backup superblock, so running journal anyway. e2fsck unable to set superblock flags on /home1

 

So... what do I do next? Is there any hope of recovering my /home partition with data intact?

[Ixthusdan]

Rather than resurrect an old thread, it is perfectly acceptable to start a new one, even if it might be a similar problem. It gets better attention this way! B)

[blackbird]

Thanks, I'll take your word for it, new thread it is.... :D

 

[edit] ............ oh it is on a new thread, I wish I didn't start another thread..................oops

Edited April 1, 2004 by blackbird

[blackbird]

one annoyance to this is that I don't seem to have mke2fs. I checked my fstab file and deleted a couple of lines for samba mounts, but I doubt that 's the trouble maker.

[bvc]

have you ran any other partition apps to see if they detect anything weird (try parted)////can you mount it from rh?

 

If it can be mounted from another partition what do you see?.... have you tar'd it up and uncompressed it in another partition to see what you can salvage? I'm just shooting in the dark here, but hey, it's better than no reply at all :D

[blackbird]

Hi, thanks for replying

When trying to mount in rh I get

[root@localhost ben]# mount -t ext3 /dev/hdb7 /mnt/mdk_home

mount: wrong fs type, bad option, bad superblock on /dev/hdb7,

    or too many mounted file systems

. What are the tools to use for partitions besides parted? My first option is to recover my data; won't parted wipe my paritition clean?

I haven't mentioned how this came to pass, but even mentioning won't really clarify things. I usually leave my computer on all day, so when I came home and moved the mouse to wake up the computer kde was running very slowly. I had gimp, quanta, xine and konqueror running so I figured that my memory was gobbled up. I saw that my hard drive led would flicker once and a while so I figured the system was using swap. I tried to close down my applications but kde was just too slow so I gave up and rebooted. Smartbootmanager came up and I looked for hd1 but it wasn't there, I rebooted and checked bios to see that that too couldn't find my poor hard drive. I powered down, opened the case and checked the connections, but everything seemed fine. Powered up again and bios saw my 2nd hard drive, but couldn't boot up Mdk due to my ext3 filesystem problems.

[Ixthusdan]

Your description can't rule out a hardware failure. In the course of recovery, the partition table is either rough or semi-messed up. I would be tempted to do a data recovery with some good tools and then reformat the drive. Bottom line, my hunch is the drive actually did this, rather than software.

[bvc]

My point about running partition apps is that some of them run certain checks when they start and could fix a problem if found.

[Ixthusdan]

Yes, if a repair is possible that is always preferred. :P

[blackbird]

my hdb drive looks to parted like this

Disk geometry for /dev/hdb: 0.000-38166.679 megabytes

Disk label type: msdos

Minor    Start    End  Type      Filesystem  Flags

1          0.031    101.975  primary ext3        boot

2        101.975 4102.536  primary ext3       

3    4102.537 5098.754  primary linux-swap 

4    5098.755  38162.219  extended              lba

5    5098.786  13099.877  logical ext3       

6      13099.909  16096.376  logical ext3       

7      16096.408  38162.219  logical           

Information: Don't forget to update /etc/fstab, if necessary.

Trying to do a simple check on hdb7 with parted results with:

Error: Unable to open /dev/hdb7 - unrecognised disk label.

The fstab file looks like this:

/dev/hdb2 / ext3 defaults 1 1

/dev/hdb1 /boot ext3 defaults 1 2

none /dev/pts devpts mode=0620 0 0

/dev/hdb7 /home ext3 defaults 1 2

none /mnt/cdrom supermount dev=/dev/hdc,fs=udf:iso9660,ro,--,iocharset=iso8859-1 0 0

none /mnt/cdrom2 supermount dev=/dev/scd0,fs=udf:iso9660,ro,--,iocharset=iso8859-1 0 0

none /mnt/floppy supermount dev=/dev/fd0,fs=ext2:vfat,--,sync,codepage=850,iocharset=iso8859-1 0 0

none /proc proc defaults 0 0

/dev/hdb5 /usr ext3 defaults 1 2

/dev/hdb6 /var ext3 defaults 1 2

/dev/hda6 /mnt/fedora_home ext3 defaults 1 2

/dev/hda2 swap swap noatime,defaults 0 0

/dev/hdb3 swap swap noatime,defaults 0 0

[bvc]

6      13099.909  16096.376  logical  ext3   

7      16096.408  38162.219  logical             

 

 

Error: Unable to open /dev/hdb7 - unrecognised disk label.

 

yuk!

 

I wouldn't know how to fix that or why it would happen. I'd try and stop some of the checks at boot (at the risk of further corruption?maybe? I don't know). Change fstab to look like this;

/dev/hdb2 / ext3 defaults 0 0

/dev/hdb1 /boot ext3 defaults 0 0

none /dev/pts devpts mode=0620 0 0

/dev/hdb7 /home ext3 defaults 0 0

none /mnt/cdrom supermount dev=/dev/hdc,fs=udf:iso9660,ro,--,iocharset=iso8859-1 0 0

none /mnt/cdrom2 supermount dev=/dev/scd0,fs=udf:iso9660,ro,--,iocharset=iso8859-1 0 0

none /mnt/floppy supermount dev=/dev/fd0,fs=ext2:vfat,--,sync,codepage=850,iocharset=iso8859-1 0 0

none /proc proc defaults 0 0

/dev/hdb5 /usr ext3 defaults 0 0

/dev/hdb6 /var ext3 defaults 0 0

/dev/hda6 /mnt/fedora_home ext3 defaults 0 0

/dev/hda2 swap swap noatime,defaults 0 0

/dev/hdb3 swap swap noatime,defaults 0 0

which is how I run mine.....I hate the checks :P

[Cannonfodder]

Just a guess, the ext3 may have been disabled in this fashion so that the recovery process can restore data from the journal. However, the flags are not set to automate this process. Once you recover the journaled data, it may be ready to use again. I have no experience with ext3 perferring reiserfs, but they are both journeled filling systems.. best to do a good read of

 

man ext3

man mount

man whatever you can think of!

[bvc]

man!...

....why didn't I think of that :lol: sounds good!

[blackbird]

I seem to be fairing slightly better in rh

[root@localhost ben]# /sbin/e2fsck -b 32768 /dev/hdb7

e2fsck 1.32 (09-Nov-2002)

ext3 recovery flag clear, but journal has data.

Recovery flag not set in backup superblock, so running journal anyway.

/home1: recovering journal

Pass 1: Checking inodes, blocks, and sizes

Inodes that were part of a corrupted orphan linked list found.  Fix<y>?

I've never seen this in mdk when I try e2fsck, I wonder if I have mke2fs as well. Anyways, what should I answer? y or n?

Edited April 4, 2004 by blackbird

[blackbird]

yes I do have mke2fs, but not to create a new filesystem of course but just to check out backup superblocks and I also wanted to check my blocksizes just in case I need that info.

[root@localhost ben]# /sbin/mke2fs -nb /dev/hdb7

mke2fs 1.32 (09-Nov-2002)

mke2fs: bad block size - /dev/hdb7

[root@localhost ben]# /sbin/mke2fs -n /dev/hdb7

mke2fs 1.32 (09-Nov-2002)

Filesystem label=

OS type: Linux

Block size=4096 (log=2)

Fragment size=4096 (log=2)

2828896 inodes, 5648847 blocks

282442 blocks (5.00%) reserved for the super user

First data block=0

173 block groups

32768 blocks per group, 32768 fragments per group

16352 inodes per group

Superblock backups stored on blocks:

        32768, 98304, 163840, 229376, 294912, 819200, 884736, 1605632, 2654208,

        4096000

 

Oh almost forgot to mention that the -b option returns a bad block size, does this really mean I'm screwed or is it possible to use an alternate superblock and fix the damage?

Edited April 4, 2004 by blackbird