Please walk me through fsck

English Install/Boot/Login
#1

I had a power failure at my house which messed up my HDD. Everything was backed up and I did a fresh install on a new HDD. However, it seems many of my konact files are missing from the backup. I want to see if I can restore the corrupted HDD and retrieve them. I am not having much luck and I don’t understand the information the terminal is returning to me. Can someone walk me through this please.

Here is what I am doing. In a terminal as root:

fdisk -l

Disk /dev/sda: 1000.2 GB, 1000204886016 bytes
255 heads, 63 sectors/track, 121601 cylinders, total 1953525168 sectors
Units = sectors of 1 * 512 = 512 bytes
Sector size (logical/physical): 512 bytes / 512 bytes
I/O size (minimum/optimal): 512 bytes / 512 bytes
Disk identifier: 0xd9aba86c

Device Boot Start End Blocks Id System
/dev/sda1 63 1953520064 976760001 83 Linux

Disk /dev/sdb: 320.1 GB, 320072933376 bytes
255 heads, 63 sectors/track, 38913 cylinders, total 625142448 sectors
Units = sectors of 1 * 512 = 512 bytes
Sector size (logical/physical): 512 bytes / 512 bytes
I/O size (minimum/optimal): 512 bytes / 512 bytes
Disk identifier: 0x00062039

Device Boot Start End Blocks Id System
/dev/sdb1 * 2048 625141759 312569856 f W95 Ext’d (LBA)
/dev/sdb5 4096 321535 158720 83 Linux
/dev/sdb6 323584 4530175 2103296 82 Linux swap / Solaris
/dev/sdb7 4532224 88422399 41945088 83 Linux
/dev/sdb8 88424448 625121279 268348416 83 Linux

Disk /dev/sdc: 150.0 GB, 150039945216 bytes
255 heads, 63 sectors/track, 18241 cylinders, total 293046768 sectors
Units = sectors of 1 * 512 = 512 bytes
Sector size (logical/physical): 512 bytes / 512 bytes
I/O size (minimum/optimal): 512 bytes / 512 bytes
Disk identifier: 0x00087d8a

Device Boot Start End Blocks Id System
/dev/sdc1 * 2048 293046271 146522112 f W95 Ext’d (LBA)
/dev/sdc5 4096 321535 158720 83 Linux
/dev/sdc6 323584 4530175 2103296 82 Linux swap / Solaris
/dev/sdc7 4532224 88422399 41945088 83 Linux
/dev/sdc8 88424448 293025791 102300672 83 Linux

It is /dev/sdb that I am trying to recover.

umount /dev/sdb

umount: /dev/sdb: not mounted

fsck /dev/sdb

fsck from util-linux 2.20.1
e2fsck 1.41.14 (22-Dec-2010)
fsck.ext2: Device or resource busy while trying to open /dev/sdb
Filesystem mounted or opened exclusively by another program?

So what am I doing wrong and what do I need to do? Thanks.

#2

thedarpone wrote:
> IT IS /DEV/SDB THAT I AM TRYING TO RECOVER.
> # umount /dev/sdb
> umount: /dev/sdb: not mounted
>
> # fsck /dev/sdb
> fsck from util-linux 2.20.1
> e2fsck 1.41.14 (22-Dec-2010)
> fsck.ext2: Device or resource busy while trying to open /dev/sdb
> Filesystem mounted or opened exclusively by another program?
>
> So what am I doing wrong and what do I need to do? Thanks.

It sounds as though you have a partition from sdb mounted. Remember it
is PARTITIONS not whole disks that you mount. Please show the output of

df -h
cat /proc/mounts

and you will want to unmount any partitions from sdb that are shown

#3

Please, when you post computer output like the fdisk -l one, use CODE tgas around the copied/pasted text: http://forums.opensuse.org/english/information-new-users/advanced-how-faq-read-only/451526-posting-code-tags-guide.html . It is not easily readable now.

As fdisk works on file systems, you must give it as a parameter a block device special file that contains a file system. As you fdisk -l shows, sdb is partitioned. Thus file systems are on the partitions (/dev/sdb5, or /dev/sdb7, or /dev/sdb8) and not on* /dev /sdb*. Thus you have to find out which one of them (or all of them) you want to check. and then use e.g.

fsck /dev/sdb7
#4

What happens if you boot from the old disk ? Add “1” to the boot options after the vga= statement (plus a space). If on 12.1 hit F5 and pick “System V” then boot into single user mode. You will only have to enter the rootpassword. After doing so you can fsck the filesystems.

Another option is to boot from a liveCD, or USBstick, and run fsck. You will be certain that none of the filesystems are mounted.

#5

Thanks for the replies. Sorry about not using the code thing.

So I umounted each partition and this is what I got. I still don’t understand if there is something here that is relevant. For clarity I will re-post the the fdisk -l output.

Disk /dev/sdb: 320.1 GB, 320072933376 bytes
255 heads, 63 sectors/track, 38913 cylinders, total 625142448 sectors
Units = sectors of 1 * 512 = 512 bytes
Sector size (logical/physical): 512 bytes / 512 bytes
I/O size (minimum/optimal): 512 bytes / 512 bytes
Disk identifier: 0x00062039

   Device Boot      Start         End      Blocks   Id  System
/dev/sdb1   *        2048   625141759   312569856    f  W95 Ext'd (LBA)
/dev/sdb5            4096      321535      158720   83  Linux
/dev/sdb6          323584     4530175     2103296   82  Linux swap / Solaris
/dev/sdb7         4532224    88422399    41945088   83  Linux
/dev/sdb8        88424448   625121279   268348416   83  Linux

Then I did the following and got the corresponding output.

~ # umount /dev/sdb1
umount: /dev/sdb1: not mounted

~ #  fsck /dev/sdb1
fsck from util-linux 2.20.1
e2fsck 1.41.14 (22-Dec-2010)
fsck.ext2: Attempt to read block from filesystem resulted in short read while trying to open /dev/sdb1
Could this be a zero-length partition?

~ # umount /dev/sdb5
umount: /dev/sdb5: not mounted

~ # fsck /dev/sdb5
fsck from util-linux 2.20.1
e2fsck 1.41.14 (22-Dec-2010)
/dev/sdb5: clean, 42/39680 files, 37333/158720 blocks

~ # umount /dev/sdb6
umount: /dev/sdb6: not mounted

~ # fsck /dev/sdb6
fsck from util-linux 2.20.1
fsck: fsck.swap: not found
fsck: error 2 while executing fsck.swap for /dev/sdb6

~ # umount /dev/sdb7
umount: /dev/sdb7: not mounted

~ # fsck /dev/sdb7
fsck from util-linux 2.20.1

~ # umount /dev/sdb8
umount: /dev/sdb8: not mounted

~ # fsck /dev/sdb8
fsck from util-linux 2.20.1

I assume /dev/sdb1 is ok, being an extended partition.

/dev/sdb5 would be /home which gives no errors. However, if I try to navigate to it using dolphin it is blank. No files or directories.

#6

You still does not understand what fsck does. It checks the consistency of file systems. As */dev/sdb1 *is the extended partition, it does NOT contain a file system. Thus trying to fsck it is nonsence. And when you understand this (as you say above), why do you then spend time on it?

Same for /dev/sdb6, which is a Swap partition. No fil esystem there!

And when fsck says that /dev/sdb5 is a correct file system, then it is correct. It means that the structure, chaining of blocks, inodes, etc. are correct. It is also correct when it does not contain any directories or files. And please, do not use those GUI tools when you want to now things about your system. When you want to know what is in /home, you do

ls -l /home

But first prove it wheny ou say: “/dev/sdb5 would be /home”, because we do not believe anything from saying. We only believe the computer itself:

mount | grep /home
#7

Thanks for confirming my suspicions.


mount | grep /home

just confirms where my present /home is. Not helpful. How do I not use the the GUI tools to navigate to an umounted HDD and find out what the /home was on that?

Then what do I do to recover my files? Just to reiterate, after my ( now replaced ) UPS failed during a wind storm I did a fresh install on a new HDD. Strangely a bunch of Kontact files are missing from the back up. I am trying to locate them on the damaged HDD.

And why do I have those partitions if there is no file system in them? I used btrfs on the install.

#8

You seem not to understand that when we ask for a command to be executed, that we then want to see that command and it’s output posted here (between CODE tags). We do NOT ask for you to tell “just confirms where my present /home is. Not helpful.”. That is for us us to decide.

I give you one more chance. Please post the output of

mount | grep /home

and

ls -l /home
#9

You don’t seen to understand what I am doing. Read my post.

~ # mount | grep /home
/dev/sdc8 on /home type btrfs (rw,relatime,space_cache)
gvfs-fuse-daemon on /home/XXX/.gvfs type fuse.gvfs-fuse-daemon (rw,nosuid,nodev,relatime,user_id=1000,group_id=100)

Totally useless information.

~ # ls -l /home
total 0
drwxr-xr-x 1 xxx   users 966 Jun 21 13:02 xxx
drwxr-xr-x 1 xxxx users 828 Jun 20 20:56 xxxx

Not even the HDD that I am interested in.

Again, so what? Useless.

Does anybody here know what they are talking about? This guy who has no manners or social skills keeps highjacking my post with useless advice. I’m a busy man and need results. Perhaps Linux is too primitive as a real business solution. I would gladly pay good money for snap solutions rather than go through this.

#10

The first output shows that your /home is mounted from /dev/sdc8. Why is that useless?

And the second output was crippled by you, but nevertheless shows that indeed nothing is there.
And it also shop that it’s group is wrong. As you criplled the user, I can only guess that it is also wrong.

But it seems that you are not interested in my information. So I will quit from this post. The last thing I want is trying to help people against their will.

#11

It is useless because the I am interested in the files that are on /dev/sdb not /dev/sdc8.

And I am not going to post user names to the world. Stop wasting my time! I feel I need to file a complaint here if this nonsense continues. I have posted a simple request.

Do I have to re-post and ask once again - As I have no idea about Unix command lines, but have a vague idea that the damaged drive can be recovered using certain esoteric commands, could someone kindly walk me through the process step by step. Thank you.

#12

On 2012-06-21 19:16, thedarpone wrote:
>
> You don’t seen to understand what I am doing. Read my post.
>
>
> Code:
> --------------------
> ~ # mount | grep /home
> /dev/sdc8 on /home type btrfs (rw,relatime,space_cache)
> gvfs-fuse-daemon on /home/XXX/.gvfs type fuse.gvfs-fuse-daemon (rw,nosuid,nodev,relatime,user_id=1000,group_id=100)
>
> --------------------
>
>
> Totally useless information.

Not for us. It is very interesting. For example, you are using btrfs, and
many of us consider that experimental. It does fail and leaves you in the
water!

–
Cheers / Saludos,

Carlos E. R.
(from 11.4 x86_64 “Celadon” at Telcontar)

#13

On 2012-06-21 19:56, thedarpone wrote:
>
> It is useless because the I am interested in the files that are on
> /dev/sdb not /dev/sdc8.

Getting adequate answers need posting adequate questions.

You did not say clearly that you were interested in sdb only. Now, I see
that you did fscks on those partitions, but I do not see them mounted and
their content. For that, I would need to see their listing in command line:


mount | grep whatever
ls -l /whatever
df -h /whatever

And yes, I also want to see the commands and the results, not your
interpretation of them. GUI programs are useless for diagnostics here.

>
> And I am not going to post user names to the world. Stop wasting my
> time! I feel I need to file a complaint here if this nonsense continues.
> I have posted a simple request.

Then use a replacement name and say clearly that you are using replacement
names.

> Do I have to re-post and ask once again - As I have no idea about Unix
> command lines, but have a vague idea that the damaged drive can be
> recovered using certain esoteric commands, could someone kindly walk me
> through the process step by step. Thank you.

If the fsck recovered nothing, there is nothing. You might try “photorec”
on them, but I doubt it will work on btrfs partitions.

–
Cheers / Saludos,

Carlos E. R.
(from 11.4 x86_64 “Celadon” at Telcontar)

#14

On 06/21/2012 12:56 PM, thedarpone wrote:
>
> It is useless because the I am interested in the files that are on
> /dev/sdb not /dev/sdc8.
>
> And I am not going to post user names to the world. Stop wasting my
> time! I feel I need to file a complaint here if this nonsense continues.
> I have posted a simple request.
>
> Do I have to re-post and ask once again - As I have no idea about Unix
> command lines, but have a vague idea that the damaged drive can be
> recovered using certain esoteric commands, could someone kindly walk me
> through the process step by step. Thank you.

I will try to help you, but if you get at all nasty, I will also dump you.
Remember that we are all volunteers. If you want help from someone that has to
take your guff, then you will need to find someone to pay.

I do not care whether you want to publish user names or not, but at least
indicate that the user names have been obscured. That would just be common courtesy.

As you do not seem to understand, /dev/sdb is the ENTIRE disk. From your fdisk
listings, it has a single primary partition, which is the extended container. It
contains no files directly, cannot be mounted, and thus cannot be fscked. One of
the extended partitions, /dev/sdb6, is a swap partition. It also has no files.
This one has no Windows equivalent.

The rest of the partitions have the form that “might” contain a file system.
From the fdisk listing, we cannot tell. The partitions in question are
/dev/sda1, /dev/sdb5, /dev/sdb7. /dev/sdb8, /dev/sdc5, /dev/sdc7, and /dev/sdc8.

The first step is to see if any of those are mounted in the boot process by
using the “df” command. If they are already mounted, you can browse them
yourself. For the others, do the following (in a terminal as root):


mount /dev/sdXN /mnt
ls -l
umount /mnt

Of course, X and N are from the list of partitions that we want to investigate.
Do each in turn.

The easy triage is as follows:

If the partition contains directories “bin”, “etc”, and so forth, it is a system
disk. If it contains directories with the names of your users, it is a “home”
disk. I guess that is what you are looking for. If it has neither of these
conditions, it is likely a user data disk.

#15

On 2012-06-21 20:38, Larry Finger wrote:

>


> mount /dev/sdXN /mnt
> ls -l
> umount /mnt
>

A “df -h /mnt” in there would also be interesting.

–
Cheers / Saludos,

Carlos E. R.
(from 11.4 x86_64 “Celadon” at Telcontar)

#16

Thank you for your useful answers. Following is the output.

Carlos, does this help? Probably not. It has nothing to do with sdb, which I clearly stated was** the only thing I was interested in in bold type at the outset - my first post.** I also announced that I had installed the system using btrfs at the outset.

~ # df -h /mnt
Filesystem      Size  Used Avail Use% Mounted on
/dev/sdc7        41G  8.4G   29G  23% /


~ # mount /dev/sdb5 /mnt
~ # ls -l
total 112
-rw------- 1 root root     0 Jun 15 09:28 .Xauthority
-rw------- 1 root root  6086 Jun 21 21:43 .bash_history
drwxr-xr-x 1 root root   132 Jun 15 09:58 .config
drwx------ 1 root root    22 Jun 15 09:18 .dbus
-rw------- 1 root root    26 Jun 15 09:23 .dmrc
-rw------- 1 root root    16 Jun 15 09:23 .esd_auth
drwx------ 1 root root    14 Jun 17 22:38 .gconf
drwx------ 1 root root     0 Jun 15 09:23 .gnome2
drwx------ 1 root root   144 Jun 15 13:11 .gnupg
drwxr-xr-x 1 root root    38 Jun 17 22:38 .gstreamer-0.10
-rw-r--r-- 1 root root   328 Jun 15 09:23 .gtkrc-2.0-kde4
drwx------ 1 root root     0 Jun 15 09:23 .gvfs
drwxr-xr-x 1 root root    20 Jun 15 09:09 .kbd
drwxr-xr-x 1 root root   122 Jun 15 09:23 .kde4
drwxr-xr-x 1 root root   198 Jun 15 09:16 .kdm
drwxr-xr-x 1 root root    12 Jun 15 13:13 .libreoffice
drwxr-xr-x 1 root root    10 Jun 15 09:23 .local
-rw------- 1 root root    89 Jun 17 22:08 .mysql_history
-rw-r--r-- 1 root root  1252 Jun 15 09:28 .nvidia-settings-rc
drwx------ 1 root root   670 Jun 15 09:23 .pulse
-rw------- 1 root root   256 Jun 15 09:23 .pulse-cookie
drwxr-xr-x 1 root root   116 Jun 15 09:23 .skel
drwx------ 1 root root    22 Jun 17 21:19 .thumbnails
-rw------- 1 root root    55 Jun 21 21:45 .xauth7Y3BAS
-rw------- 1 root root 76845 Jun 15 09:28 .xsession-errors
drwxr-xr-x 1 root root   182 Jun 15 09:23 Desktop
drwxr-xr-x 1 root root    20 Jun 15 09:23 Documents
drwxr-xr-x 1 root root     0 Jun 15 09:23 Downloads
drwxr-xr-x 1 root root     0 Jun 15 09:23 Music
drwxr-xr-x 1 root root     0 Jun 15 09:23 Pictures
drwxr-xr-x 1 root root     0 Jun 15 09:23 Public
drwxr-xr-x 1 root root     0 Jun 15 09:23 Templates
drwxr-xr-x 1 root root     0 Jun 15 09:23 Videos
drwxr-xr-x 1 root root     0 Oct 25  2011 bin
drwxr-xr-x 1 root root    26 Jun 15 08:52 inst-sys

I am pretty sure that is not the output I am looking for. I think that is the output for my main HDD. So I navigated to:


~ # cd /mnt
:/mnt # ls -l
total 26396
-rw-r--r-- 1 root root  2548922 Apr 10 11:24 System.map-3.1.10-1.9-desktop
-rw------- 1 root root      512 Jan 13 23:35 backup_mbr
lrwxrwxrwx 1 root root        1 Feb 24 12:43 boot -> .
-rw-r--r-- 1 root root     1236 Oct 25  2011 boot.readme
-rw-r--r-- 1 root root   128908 Apr 10 09:55 config-3.1.10-1.9-desktop
drwxr-xr-x 2 root root     1024 Apr 25 15:08 grub
lrwxrwxrwx 1 root root       25 May 22 12:33 initrd -> initrd-3.1.10-1.9-desktop
-rw-r--r-- 1 root root 12880013 May 22 12:33 initrd-3.1.10-1.9-desktop
drwx------ 2 root root    12288 Jan 13 23:26 lost+found
-rw-r--r-- 1 root root   436224 Jan 13 23:35 message
-rw-r--r-- 1 root root   212478 Apr 10 11:43 symvers-3.1.10-1.9-desktop.gz
-rw-r--r-- 1 root root      515 Apr 10 11:43 sysctl.conf-3.1.10-1.9-desktop
-rw-r--r-- 1 root root  5817042 Apr 10 11:43 vmlinux-3.1.10-1.9-desktop.gz
lrwxrwxrwx 1 root root       26 Apr 25 15:07 vmlinuz -> vmlinuz-3.1.10-1.9-desktop
-rw-r--r-- 1 root root  4985808 Apr 10 11:25 vmlinuz-3.1.10-1.9-desktop
:/mnt # umount /mnt
umount: /mnt: device is busy.
        (In some cases useful info about processes that use
         the device is found by lsof(8) or fuser(1))
:/mnt # cd /
:/ # umount /mnt
:/ # mount /dev/sdb7 /mnt
:/ # cd /mnt
:/mnt # ls -l
total 16
drwx------ 1 root root    34 May 28 15:08 .Trash-0
drwxr-xr-x 1 root root    28 Jan 13 23:41 .config
-rw-r--r-- 1 root root 10152 Jun  8 08:29 .readahead
drwx------ 1 root root  1062 Jun  8 01:30 .snapshots
-rw-r--r-- 1 root root    24 Apr 16 13:53 C:
ppdf32Log\debuglog.txt
drwxr-xr-x 1 root root  1998 May 31 21:23 bin
drwxr-xr-x 1 root root     0 Jan 13 23:26 boot
drwxr-xr-x 1 root root   208 Jan 13 23:35 dev
drwxr-xr-x 1 root root  5596 Jun  8 01:31 etc
drwxr-xr-x 1 root root     0 Jan 13 23:26 home
drwxr-xr-x 1 root root  3064 Jun  4 18:27 lib
drwxr-xr-x 1 root root  5910 Jun  4 18:27 lib64
drwxr-xr-x 1 root root     0 Oct 25  2011 media
drwxr-xr-x 1 root root    24 Jan 15 16:04 mnt
drwxr-xr-x 1 root root    34 Apr 12 15:45 opt
drwxr-xr-x 1 root root     6 Jan 13 23:26 proc
drwx------ 1 root root   810 Jun  7 09:57 root
drwxr-xr-x 1 root root     0 Jan 13 23:26 run
drwxr-xr-x 1 root root  5070 May 31 21:23 sbin
drwxr-xr-x 1 root root     0 Oct 25  2011 selinux
drwxr-xr-x 1 root root    20 Feb  9 17:47 srv
drwxr-xr-x 1 root root    12 Jan 13 23:26 sys
drwxrwxrwt 1 root root  4944 Jun  8 01:31 tmp
drwxr-xr-x 1 root root   130 Nov 10  2011 usr
drwxr-xr-x 1 root root   106 Nov 10  2011 var
:/mnt # cd /
:/ # umount /mnt

This is the interesting one.


:/ # mount /dev/sdb8 /mnt
mount: wrong fs type, bad option, bad superblock on /dev/sdb8,
       missing codepage or helper program, or other error
       In some cases useful info is found in syslog - try
       dmesg | tail  or so

linux-sgbb:/ # fsck /dev/sdb8
fsck from util-linux 2.20.1

No doubt this is where the user info is and fsck seem to have no affect. Is there any chance of recovering anything?

#17

On 06/21/2012 09:46 PM, thedarpone wrote:
>
> Thank you for your useful answers. Following is the output.
>
> Carlos, does this help? Probably not. It has nothing to do with sdb,
> which I clearly stated was* the only thing I was interested in in bold
> type at the outset - my first post.* I also announced that I had
> installed the system using btrfs at the outset.
>
>
> Code:
> --------------------
> ~ # df -h /mnt
> Filesystem Size Used Avail Use% Mounted on
> /dev/sdc7 41G 8.4G 29G 23% /
> --------------------
>
>
>
> Code:
> --------------------
>
> ~ # mount /dev/sdb5 /mnt
> ~ # ls -l
> total 112

By now, you realize that this does not get the directory on /mnt, but is the
directory for the current working directory. Sorry that I gave you the wrong
command. After the mount, you should have done ‘ls -l /mnt’.

> -rw------- 1 root root 0 Jun 15 09:28 .Xauthority
> -rw------- 1 root root 6086 Jun 21 21:43 .bash_history
–snip–

>
> I am pretty sure that is not the output I am looking for. I think that
> is the output for my main HDD. So I navigated to:
>
>
> Code:
> --------------------
>
> ~ # cd /mnt
> :/mnt # ls -l
> total 26396
> -rw-r–r-- 1 root root 2548922 Apr 10 11:24 System.map-3.1.10-1.9-desktop
> -rw------- 1 root root 512 Jan 13 23:35 backup_mbr
> lrwxrwxrwx 1 root root 1 Feb 24 12:43 boot → .
> -rw-r–r-- 1 root root 1236 Oct 25 2011 boot.readme
> -rw-r–r-- 1 root root 128908 Apr 10 09:55 config-3.1.10-1.9-desktop
> drwxr-xr-x 2 root root 1024 Apr 25 15:08 grub
> lrwxrwxrwx 1 root root 25 May 22 12:33 initrd → initrd-3.1.10-1.9-desktop
> -rw-r–r-- 1 root root 12880013 May 22 12:33 initrd-3.1.10-1.9-desktop
> drwx------ 2 root root 12288 Jan 13 23:26 lost+found
> -rw-r–r-- 1 root root 436224 Jan 13 23:35 message
> -rw-r–r-- 1 root root 212478 Apr 10 11:43 symvers-3.1.10-1.9-desktop.gz
> -rw-r–r-- 1 root root 515 Apr 10 11:43 sysctl.conf-3.1.10-1.9-desktop
> -rw-r–r-- 1 root root 5817042 Apr 10 11:43 vmlinux-3.1.10-1.9-desktop.gz
> lrwxrwxrwx 1 root root 26 Apr 25 15:07 vmlinuz → vmlinuz-3.1.10-1.9-desktop
> -rw-r–r-- 1 root root 4985808 Apr 10 11:25 vmlinuz-3.1.10-1.9-desktop

These files come from a partition that would be mounted as /boot.

> :/mnt # umount /mnt
> umount: /mnt: device is busy.
> (In some cases useful info about processes that use
> the device is found by lsof(8) or fuser(1))
> :/mnt # cd /
> :confused: # umount /mnt
> :confused: # mount /dev/sdb7 /mnt
> :confused: # cd /mnt
> :/mnt # ls -l
> total 16
> drwx------ 1 root root 34 May 28 15:08 .Trash-0
> drwxr-xr-x 1 root root 28 Jan 13 23:41 .config
> -rw-r–r-- 1 root root 10152 Jun 8 08:29 .readahead
> drwx------ 1 root root 1062 Jun 8 01:30 .snapshots
> -rw-r–r-- 1 root root 24 Apr 16 13:53 C:
ppdf32Log\debuglog.txt
> drwxr-xr-x 1 root root 1998 May 31 21:23 bin
> drwxr-xr-x 1 root root 0 Jan 13 23:26 boot
> drwxr-xr-x 1 root root 208 Jan 13 23:35 dev
> drwxr-xr-x 1 root root 5596 Jun 8 01:31 etc
> drwxr-xr-x 1 root root 0 Jan 13 23:26 home
> drwxr-xr-x 1 root root 3064 Jun 4 18:27 lib
> drwxr-xr-x 1 root root 5910 Jun 4 18:27 lib64
> drwxr-xr-x 1 root root 0 Oct 25 2011 media
> drwxr-xr-x 1 root root 24 Jan 15 16:04 mnt
> drwxr-xr-x 1 root root 34 Apr 12 15:45 opt
> drwxr-xr-x 1 root root 6 Jan 13 23:26 proc
> drwx------ 1 root root 810 Jun 7 09:57 root
> drwxr-xr-x 1 root root 0 Jan 13 23:26 run
> drwxr-xr-x 1 root root 5070 May 31 21:23 sbin
> drwxr-xr-x 1 root root 0 Oct 25 2011 selinux
> drwxr-xr-x 1 root root 20 Feb 9 17:47 srv
> drwxr-xr-x 1 root root 12 Jan 13 23:26 sys
> drwxrwxrwt 1 root root 4944 Jun 8 01:31 tmp
> drwxr-xr-x 1 root root 130 Nov 10 2011 usr
> drwxr-xr-x 1 root root 106 Nov 10 2011 var
> :/mnt # cd /
> :confused: # umount /mnt

This is a system disk.

>
> --------------------
>
>
> This is the interesting one.
>
>
> Code:
> --------------------
>
> :confused: # mount /dev/sdb8 /mnt
> mount: wrong fs type, bad option, bad superblock on /dev/sdb8,
> missing codepage or helper program, or other error
> In some cases useful info is found in syslog - try
> dmesg | tail or so
>
> linux-sgbb:/ # fsck /dev/sdb8
> fsck from util-linux 2.20.1
>
> --------------------
>
>
> No doubt this is where the user info is and fsck seem to have no
> affect. Is there any chance of recovering anything?

If it has a file system on it, it has been so trashed that mount does not know
how to handle it. Too bad you did not post the output from fsck. without that,
there is no way to tell if there is any hope for recovery.

What sort of data is missing? If it were text files, you might see what is there
by using something like “dd if=/dev/sdb8 | strings | less”. This command must be
run as root. You will see the text data strings in that partition one page at a
time. Press the space key to advance to a new page.

If you see hints of the files you are missing, then you may need to hire the
services of a data-recovery expert. That will be expensive, and you will need to
decide if it is cost effective. Do you think the missing files were also on a
btrfs file system? The expert will need to know that info as well.

If you were using a btrfs file system, you are at a disadvantage. Not only is
that file system new and fragile, but the error recovery tools for it are not as
good as those for other file systems. I, for one, would never trust anything
that I wanted to keep to btrfs, even though it has some features that would be
highly desirable.

#18

Thank you. The missing data is from Kontact and Kmail. I had backed up my entire /home directory and strangely not all of the data was there when I went to restore it. Fortunately my provider archives my email for years, but the Kontact data is gone. I will have to restore it from other sources.

I apologize for getting miffed. I stated that the only data I was interested in was in sdb. I stated that in bold lettering so no one would miss it. I also stated that I had used btfrs, so all the help I was getting was useless, and someone had the audacity to tell me to shut up and do as they say, when they were clearly talking nonsense and not paying attention. Just to remind you here is a quote, “We do NOT ask for you to tell “just confirms where my present /home is. Not helpful.”. That is for us us to decide. I give you one more chance.” What kind of condescending nonsense is that? I am overwhelmed by the gall of someone that would say such a thing! Who do they think they are!

I understand that you are volunteers, but that does not give anybody the right to be careless with other people’s time and resources. People who do not have the skills to pay attention and communicate with grace and dignity should not be allowed to volunteer.

Once again, thank you for the help.

BTW, the last post says, “Too bad you did not post the output from fsck. without that,
there is no way to tell if there is any hope for recovery.”

Once again my ignorance of these commands has me confused. I thought I did post the output.

:/ # fsck /dev/sdb8
  fsck from util-linux 2.20.1

It is probably a moot point now, but just as a matter of interest and for the benefit of others that might be having trouble with fsck, is the syntax wrong? That is all the output I got.

#19

On 2012-06-22 04:46, thedarpone wrote:
>
> Thank you for your useful answers. Following is the output.
>
> Carlos, does this help? Probably not. It has nothing to do with sdb,
> which I clearly stated was* the only thing I was interested in in bold
> type at the outset - my first post.* I also announced that I had
> installed the system using btrfs at the outset.
>
>
> Code:
> --------------------
> ~ # df -h /mnt
> Filesystem Size Used Avail Use% Mounted on
> /dev/sdc7 41G 8.4G 29G 23% /
> --------------------

But you are not doing what I asked. I asked for the output of “df” after
mounting, about the same time you do the “ls” that Larry asked you about.
Obviously, the output above is totally useless.

By the way, the word “btrfs” is not in the first post, I just searched and
it appears in number three (#7 via flat web display).

> I am pretty sure that is not the output I am looking for. I think that
> is the output for my main HDD. So I navigated to:

True. You should have done “ls -l /mnt”.

> This is the interesting one.
>
> Code:
> --------------------
>
> :confused: # mount /dev/sdb8 /mnt
> mount: wrong fs type, bad option, bad superblock on /dev/sdb8,
> missing codepage or helper program, or other error
> In some cases useful info is found in syslog - try
> dmesg | tail or so
>
> linux-sgbb:/ # fsck /dev/sdb8
> fsck from util-linux 2.20.1
>
> --------------------
>
>
> No doubt this is where the user info is and fsck seem to have no
> affect. Is there any chance of recovering anything?

That is all that fsck says? I’m very surprised, as Larry, and forgive us
for asking for the output again, but this is not normal.

You did not do the “dmesg | tail” it is suggested above. Please repeat the
check and post that dmesg thing.

Let me guess: that is a “btrfs” partition. Do you remember me saying that
many people consider btrfs as experimental? You are discovering now why.

Sigh… Ok, let us confirm by doing


file -s /dev/sdc7

Or check your fstab to see how the device was mounted.

If it confirms it is “btrfs”, then please make sure you have installed the
“btrfsprogs” package. As I don’t have it installed (and not going to, sorry
(I’m not using that filesystem type)), please also get us the listing of
files in there by doing:


rpm -ql btrfsprogs

And lets hope there is something interesting in there.

Also, you may try to use “photorec” on that partition. If it finds
anything, you need to tell it another partition where to copy the files it
finds.

–
Cheers / Saludos,

Carlos E. R.
(from 11.4 x86_64 “Celadon” at Telcontar)

#20

Carlos E. R. wrote:
> That is all that fsck says? I’m very surprised, as Larry, and forgive us
> for asking for the output again, but this is not normal.

What do you expect it to do when it has no fsck for the filesystem type?

There is a new one now, apparently, but it is recommended not to use it!
http://www.phoronix.com/scan.php?page=news_item&px=MTA2MDI