"Extra" Internal Hard Drives won't mount - swap, /, and /home work.

The Avatar of Time · 2010-09-20 08:54:51

Hello All,

*** I'm updating my post finally. Sorry it's so long, I'm not entirely sure what is and isn't relevant.

*** Also I think it's a mix up with RAID causing the issue, as I'll address below, but in case I'm wrong I wanted to include the other information that came to mind before I noticed the RAID issue (because the RAID issue doesn't make a lot of sense to me).

First off, I'm running Arch64 and KDE4 with my system up to date. Also I have Gamin installed and not FAM. I use ReiserFS on all of my partitions and drives.

I have six hard drives in my computer, all of which are SATA drives. One is a WD 150GB Raptor with my swap, /, and /home partitions on it, the other five are for extra storage on what I call my media-server (they mount to /media/media-server/1 ,2 ,and so on). They only have one partition each. Of those five the first (/media/media-server/1) is a WD 1TB drive, the other four are WD 500GB drives of identical model.

All of the Raptor's patitions, swap, /, and /home mount just fine.

On top of that the 1TB drive mounts just fine. But the four 500GB drives will not mount at all. They show up in Dolphin, Disk Utility, KwikDisk, KDiskFree, and Partition Manager. However, their size is not shown and/or marked N/A, except in Partition Manager, where their size is shown. They are not seen by 'df'.

Output from 'df':

df -h
Filesystem            Size  Used Avail Use% Mounted on
udev                   10M  308K  9.7M   4% /dev
/dev/sda2              25G  7.7G   17G  32% /
none                  4.0G     0  4.0G   0% /dev/shm
/dev/sda3             112G   90G   22G  81% /home
/dev/sdb1             932G  919G   13G  99% /media/media-server/1

*** As for removable media, it's being a bit odd. When I first wrote this post removable media, specifically USB flash drives and SD media cards mounted fine. A while after that they'd show up in Device Notifier, but wouldn't mount. Just a little while ago they wouldn't show up in Device Notifier and wouldn't mount. Now they show up fine and mount fine again. This is possibly unrelated, I'm not sure.

Depending on what method I try to mount my hard drives I either get an 'only root can mount that' error or else a HAL drive busy or already mounted error, followed by a 'drive is not mounted' error when I use umount to double check.

*** Permissions are working better now , Dolphin isn't giving me the 'only root can mount' errors anymore. Still can't mount the drives though.

All of this output is from when the drives are listed in /etc/fstab:
Konsole output:

mount /dev/sdf1
mount: only root can mount /dev/sdf1 on /media/media-server/5

sudo mount /dev/sdf1
mount: /dev/sdf1 already mounted or /media/media-server/5 busy

umount /dev/sdf1
umount: /dev/sdf1 is not mounted (according to mtab)

sudo umount /dev/sdf1
umount: /dev/sdf1: not mounted

Dolphin output:

An error occurred while accessing 'MediaServer5', the system responded: org.freedesktop.Hal.Device.Volume.PermissionDenied: Device /dev/sdf1 is listed in /etc/fstab. Refusing to mount.

Device Notifier output (pops up when I try to mount the drives in Dolphin):

org.freedesktop.Hal.Device.Volume.PermissionDenied: Device /dev/sdf1 is listed in /etc/fstab. Refusing to mount.

Partition Manager:

The file system on partition /dev/sdf1 could not be mounted.

Details

==========================================================================================
Command: mount -v /dev/sdf1 /media/media-server/5
==========================================================================================
mount: /dev/sdf1 already mounted or /media/media-server/5 busy
mount: you didn't specify a filesystem type for /dev/sdf1
       I will try type reiserfs

At boot, when the daemons are starting I get:

mount: /dev/sdc1 already mounted or /media/media-server/2 busy
mount: /dev/sdd1 already mounted or /media/media-server/3 busy
mount: /dev/sde1 already mounted or /media/media-server/4 busy
mount: /dev/sdf1 already mounted or /media/media-server/5 busy

This is from when the drives are not listed in /etc/fstab:
Dolphin output:

An error occurred while accessing 'MediaServer5', the system responded: org.freedesktop.Hal.Device.InterfaceLocked: The enclosing drive for the volume is locked

Device Notifier output (pops up when I try to mount the drives in Dolphin):

An error occurred while accessing 'MediaServer5', the system responded: org.freedesktop.Hal.Device.InterfaceLocked: The enclosing drive for the volume is locked

I have tried listing (with the UUID method and the /dev method) and not listing the drives in /etc/fstab, and I have 'hal' and 'dbus' daemons in /etc/rc.conf (and I've tried using only the hal daemon and not the dbus in /etc/rc.conf). I've also tried not listing 'hal', 'dbus', and 'kdm' and only booting to terminal and then mounting the drives manually, that doesn't work either.

Here is my /etc/fstab (back to using UUID's):

# 
# /etc/fstab: static file system information
#
# <file system>        <dir>         <type>    <options>          <dump> <pass>
none                   /dev/pts      devpts    defaults            0      0
none                   /dev/shm      tmpfs     defaults            0      0

UUID=699cdd5e-3f4a-4b85-8dcb-5d00c00c94a9 swap swap defaults 0 0
UUID=ce94e7a3-3e07-41e5-90d7-1116e5da3121 / reiserfs defaults 0 0
UUID=a1dce435-df90-4f0d-892c-61ff3ea0c8d1 /home reiserfs defaults 0 0

## Media-Server

# media-server/1
UUID=81998a75-f887-4a35-b3fa-e159190f25a3 /media/media-server/1 reiserfs defaults 0 0
#/dev/sdb1  /media/media-server/1 reiserfs defaults 0 0

# media-server/2 
UUID=8cdd114a-b190-44f6-9240-1910a5d048f7 /media/media-server/2 reiserfs defaults 0 0
#/dev/sdc1  /media/media-server/2 reiserfs defaults 0 0

# media-server/3
UUID=59164593-b56d-4c49-8837-277b3c392dfb /media/media-server/3 reiserfs defaults 0 0
#/dev/sdd1  /media/media-server/3 reiserfs defaults 0 0 

# media-server/4
UUID=ba6007bd-1e36-431c-85f4-857c83845c86 /media/media-server/4 reiserfs defaults 0 0
#/dev/sde1  /media/media-server/4 reiserfs defaults 0 0

# media-server/5 
UUID=6fab5364-1acb-4ba3-9f83-bb171a6d398a /media/media-server/5 reiserfs defaults 0 0
#/dev/sdf1  /media/media-server/5 reiserfs defaults 0 0

## VirtualBox USB
none /proc/bus/usb usbfs auto,busgid=108,busmode=0775,devgid=108,devmode=0664 0 0

Here's the daemons line from /etc/rc.conf:

DAEMONS=(syslog-ng !dbus hal network netfs crond wicd openntpd sshd avahi-daemon avahi-dnsconfd samba alsa mysqld lircd mythbackend !mpd !sensors !sensord !fancontrol cups keytouch kdm)

Also, /etc/mtab because only the first media-server drive is listed in it (the one that does mount):

proc /proc proc rw,relatime 0 0
sys /sys sysfs rw,relatime 0 0
udev /dev devtmpfs rw,nosuid,relatime,size=10240k,nr_inodes=1014438,mode=755 0 0
/dev/sda2 / reiserfs rw,relatime 0 0
fusectl /sys/fs/fuse/connections fusectl rw,relatime 0 0
none /dev/pts devpts rw 0 0
none /dev/shm tmpfs rw 0 0
/dev/sda3 /home reiserfs rw 0 0
/dev/sdb1 /media/media-server/1 reiserfs rw 0 0
none /proc/bus/usb usbfs rw,busgid=108,busmode=0775,devgid=108,devmode=0664 0 0

I attempted to add the other drives to /etc/mtab but they get overwritten upon a reboot. I've drived mounting the drives from konsole and dolphin.

All my drives used to work. At one point I had Chakra installed and had this issue after an update, however after thinking about it the problem may have actually started after the power flashed on and off here. I couldn't fix it so I installed Arch64 and KDE4 (the normal KDE from the Arch repos, not KDEMod) thinking maybe it was a Chakra specific issue. That fixed it at the time. It later happened on my Arch install. I reinstalled Arch again and the issue was gone. However the issue is back. Again I think it was after a power outage/flash but it was also after an update that included a kernel update. Best I can remember each time it's happened has been after a kernel update and power flash. So I can't say for sure which is the cause.

Oddly though if I boot from a live disc (in my case a Chakra Live DVD) I can mount and access all of the drives just fine.

Another oddity, is that in 'Disk Utility' (palimpsest) there is a 1.5TB RAID-5 array listed. The only thing I can figure is that that is what is making the individual drives busy.

Here's a screenshot from Disk Utility of the RAID-5 array:
http://theavataroftime.com/sites/defaul … enshot.png

After talking to a friend I remembered that quite a while back (over a year) we had messed around trying to setup a RAID array and dual boot WindowsXP. However, we never could get it working quite right, not even just in Linux without dual booting. I really can't say what all we did/tried, it's been too long. I realize that's not much help but...

Anyway I'm pretty certain I deleted everything RAID related at the time and turned it off. However I must be mistaken or I guess the array wouldn't be showing up.

I apologize, I really know nothing about RAID or RAID controllers. I do have an onboard RAID controller, but it's not a full fledged hardware RAID controller. It's onboard, but doesn't have it's on proccessor like good hardware RAID would. Or that's the way I understand it. Perhaps 'hybrid' is the proper term?

I checked in BIOS and my onboard RAID controller is off.

I take it that my attempts at setting up a RAID array before are still in the MBR (on that note I use GRUB and have never used LILO)? That's the only place I can think of that any leftover RAID configurations wouldn't have been wiped during reinstalling and reformatting the drives and partitions? I've reformatted my drives completely and changed the size of swap, /, and /home on several occasions since then, as well as formatting the media-server drives to ReiserFS (not sure what FS I was using when trying to setup RAID before. Perhaps FAT32 since I was trying to dual boot).

I clicked the stop array button in Disk utility, and then all the options except start array went away. Then I clicked start, and instead of my four media-server drives being listed at the side of Dolphin there was one 1.4TB drive. I rebooted and now it's back to the 4 500GB media-server drives.

I'm worried about losing a bunch of data, so I'm hesitant to start experimenting. How should I go about getting rid of the remnants of the RAID array safely?

Also here's a screenshot of MediaServer5 from Disk Utility. The other 500GB drives have the same info minus the names and labels of course:
http://theavataroftime.com/sites/defaul … enshot.png

*** I had forgotten about dmesg output. It's quite long so instead of copying it, here's a link to it:
http://theavataroftime.com/sites/defaul … /dmesg.txt

Thanks for advice on the matter.

Last edited by The Avatar of Time (2010-10-07 08:27:06)

delcypher · 2010-09-20 11:29:29

I believe the default mount options (see man mount) is only to allow root to mount (this can be changed by adding the user, owner or group option). Although this probably isn't your problem as you can't mount as root.

Have you checked the following...

- That your mount points (e.g. /media/media-server/2 ) exist and are not is use by something else

to check run

fuser -v -m /media/media-server/2

- Check that your block devices aren't being used by something else... to check run

fuser -v -m /dev/sdc1

- Running reiserfsck on the partitions that won't mount to see if there is a problem with them.

I see that you're using KDE daemon and not using initab. You could try not starting KDE and possible HAL & dbus and try manually mounting from the console and if it fails there looking at the output of dmesg to see if there's a hardware problem somewhere.

Sorry not sure if I can be much more help than that. Good luck.

Last edited by delcypher (2010-09-20 11:29:51)

The Avatar of Time · 2010-09-23 05:32:23

Thanks for the reply.

Here is my output from the commands you suggested:

fuser -v -m /media/media-server/2
                     USER        PID ACCESS COMMAND
/media/media-server/2:
                     root     kernel mount /
                     theavataroftime   2518 fr.e. startkde
                     theavataroftime   2561 Frce. gpg-agent
                     theavataroftime   2575 .rce. dbus-launch
                     theavataroftime   2576 .rce. dbus-daemon
                     theavataroftime   2584 Fr.e. kdeinit4
                     theavataroftime   2585 Fr.e. klauncher
                     theavataroftime   2587 fr.e. kded4
                     theavataroftime   2592 .r.e. kglobalaccel
                     theavataroftime   2605 .r.e. kwrapper4
                     theavataroftime   2613 Fr.e. ksmserver
                     theavataroftime   2622 .r.e. kwin
                     theavataroftime   2656 fr.e. knotify4
                     theavataroftime   2658 fr.e. plasma-desktop
                     theavataroftime   2662 .rce. kuiserver
                     theavataroftime   2665 .r.e. akonadi_control
                     theavataroftime   2667 .r.e. akonadiserver
                     theavataroftime   2670 Fr.e. mysqld
                     theavataroftime   2714 .r.e. akonadi_contact
                     theavataroftime   2715 .r.e. akonadi_contact
                     theavataroftime   2716 .r.e. akonadi_ical_re
                     theavataroftime   2717 .r.e. akonadi_ical_re
                     theavataroftime   2718 .r.e. akonadi_maildir
                     theavataroftime   2719 .r.e. akonadi_maildis
                     theavataroftime   2720 .r.e. akonadi_nepomuk
                     theavataroftime   2721 .r.e. akonadi_vcard_r
                     theavataroftime   2739 .r.e. nepomukserver
                     theavataroftime   2741 .r.e. nepomukservices
                     theavataroftime   2747 .r.e. kaccess
                     theavataroftime   2759 fr.e. krunner
                     theavataroftime   2765 Fr.e. virtuoso-t
                     theavataroftime   2772 Fr.e. kio_http_cache_
                     theavataroftime   2774 .r.e. kmix
                     theavataroftime   2782 Fr.e. konsole
                     theavataroftime   2785 .r.e. kjots
                     theavataroftime   2790 .r.e. bash
                     theavataroftime   2807 .r.e. kalarm
                     theavataroftime   2809 fr.e. kopete
                     theavataroftime   2812 .r.e. polkit-kde-auth
                     theavataroftime   2815 .r.e. kwin
                     theavataroftime   2821 .r.e. dolphin
                     theavataroftime   2822 Fr.e. firefox
                     theavataroftime   2834 .r.e. kwalletd
                     theavataroftime   2839 .r.e. kwalletmanager
                     theavataroftime   2842 Frce. gconfd-2
                     theavataroftime   2850 .r.e. python
                     theavataroftime   2854 .rce. wicd-client
                     theavataroftime   2858 .r.e. klipper
                     theavataroftime   2861 .r.e. nepomukservices
                     theavataroftime   2862 .r.e. nepomukservices
                     theavataroftime   2865 .r.e. nepomukservices
                     theavataroftime   2867 .r.e. nepomukservices
                     theavataroftime   2868 .r.e. nepomukservices
                     theavataroftime   2869 .r.e. nepomukservices
                     theavataroftime   2870 .r.e. nepomukservices
                     theavataroftime   2871 .r.e. korgac
                     theavataroftime   2883 fr.e. kmozillahelper
                     theavataroftime   2979 .r.e. amarokapp
                     theavataroftime   2989 .r.e. plugin-containe
                     theavataroftime   3010 .r.e. npviewer.bin
                     theavataroftime   3048 Fr.e. kdeinit
                     theavataroftime   3051 Fr.e. dcopserver
                     theavataroftime   3053 Fr.e. klauncher
                     theavataroftime   3055 fr.e. kded
                     theavataroftime   3075 .r.e. ruby
                     theavataroftime   3081 fr.e. kio_file
                     theavataroftime  15311 .r.e. kio_http
                     theavataroftime  15312 .r.e. kio_http
                     theavataroftime  15313 .r.e. kio_http
                     theavataroftime  15314 .r.e. kio_http
                     theavataroftime  15315 .r.e. kio_http
                     theavataroftime  15316 .r.e. kio_http

fuser -v -m /dev/sdb1
                     USER        PID ACCESS COMMAND
/dev/sdb1:           root     kernel mount /media/media-server/1

I ran 'fsck.reiserfs --check' one /dev/sdf1, which is the /media/media-server/5 hard drive.

Here are the results:

fsck.reiserfs --check /dev/sdf1
reiserfsck 3.6.21 (2009 www.namesys.com)

*************************************************************
** If you are using the latest reiserfsprogs and  it fails **
** please  email bug reports to reiserfs-list@namesys.com, **
** providing  as  much  information  as  possible --  your **
** hardware,  kernel,  patches,  settings,  all reiserfsck **
** messages  (including version),  the reiserfsck logfile, **
** check  the  syslog file  for  any  related information. **
** If you would like advice on using this program, support **
** is available  for $25 at  www.namesys.com/support.html. **
*************************************************************

Will read-only check consistency of the filesystem on /dev/sdf1
Will put log info to 'stdout'

Do you want to run this program?[N/Yes] (note need to type Yes if you do):Yes
###########
reiserfsck --check started at Mon Sep 27 19:21:20 2010
###########
Replaying journal: Done.
Reiserfs journal '/dev/sdf1' in blocks [18..8211]: 0 transactions replayed
Checking internal tree.. finished                                
Comparing bitmaps..finished
Checking Semantic tree:
finished                                                                       
No corruptions found
There are on the filesystem:
        Leaves 77863
        Internal nodes 474
        Directories 2856
        Other files 18718
        Data block pointers 77184421 (0 of them are zero)
        Safe links 0
###########
reiserfsck finished at Mon Sep 27 19:29:03 2010
###########

Also I just used Partition Manager to check /media/media-server/5 as well.

Check and repair partition '/dev/sdf1' (465.76 GiB, reiser) 
Job: Check file system on partition '/dev/sdf1' 
Command: fsck.reiserfs --fix-fixable -q -y /dev/sdf1 
reiserfsck 3.6.21 (2009 www.namesys.com)

*************************************************************
** If you are using the latest reiserfsprogs and  it fails **
** please  email bug reports to reiserfs-list@namesys.com, **
** providing  as  much  information  as  possible --  your **
** hardware,  kernel,  patches,  settings,  all reiserfsck **
** messages  (including version),  the reiserfsck logfile, **
** check  the  syslog file  for  any  related information. **
** If you would like advice on using this program, support **
** is available  for $25 at  www.namesys.com/support.html. **
*************************************************************

Will check consistency of the filesystem on /dev/sdf1
and will fix what can be fixed without --rebuild-tree
Will put log info to 'stdout'
###########
reiserfsck --fix-fixable started at Wed Oct  6 22:29:02 2010
###########
Replaying journal: Replaying journal: Done.
Reiserfs journal '/dev/sdf1' in blocks [18..8211]: 0 transactions replayed
Checking internal tree.. 

***I'm cutting most of this out, it's insanely long: \|/-\|/-\|finished

Comparing bitmaps..Bad nodes were found, Semantic pass skipped
3 found corruptions can be fixed only when running with --rebuild-tree
###########
reiserfsck finished at Wed Oct  6 22:36:43 2010
###########
block 32772: The level of the node (48287) is not correct, (1) expected
 the problem in the internal node occured (32772), whole subtree is skipped
block 262149: The level of the node (28423) is not correct, (1) expected
 the problem in the internal node occured (262149), whole subtree is skipped
block 11326: The level of the node (63701) is not correct, (1) expected
 the problem in the internal node occured (11326), whole subtree is skipped
vpf-10630: The on-disk and the correct bitmaps differs. Will be fixed later. 
Check file system on partition '/dev/sdf1': Error
Check and repair partition '/dev/sdf1' (465.76 GiB, reiser): Error

Thanks for the advice.

Last edited by The Avatar of Time (2010-10-07 05:53:05)

The Avatar of Time · 2010-09-26 10:05:04

I noticed that both the RAID array screenshot from Disk Utility and the output from checking /media/media-server/5 with Partition Manager mention problems.

In Disk Utility about the RAID array:

WARNING: The partition is misaligned by 33280 bytes. This may result in very poor performance. Repartitioning is suggested.ARNING: The partition is misaligned by 33280 bytes. This may result in very poor performance. Repartitioning is suggested.

In Parition Manager about /media/media-server/5:

block 32772: The level of the node (48287) is not correct, (1) expected
 the problem in the internal node occured (32772), whole subtree is skipped
block 262149: The level of the node (28423) is not correct, (1) expected
 the problem in the internal node occured (262149), whole subtree is skipped
block 11326: The level of the node (63701) is not correct, (1) expected
 the problem in the internal node occured (11326), whole subtree is skipped
vpf-10630: The on-disk and the correct bitmaps differs. Will be fixed later.

However, Disk Utility shows /media/media-server/5 (and all the drives, just not the RAID array) as being fine:
http://theavataroftime.com/sites/defaul … enshot.png

Also I don't understand is if there's a problem with the partitioning of the drives, why can I access them fine from a live disk?

Last edited by The Avatar of Time (2010-10-07 06:07:49)

Mr.Elendig · 2010-09-26 12:42:58

The partions should have been mounted at boot. Pay close attention to the boot process, maybe there is some error related to them? If you have had a unclean shutdown, that could eg be the case. Try to fsck the partions. Also check dmesg for any errors. Also try to mount the partions by hand as root, and see if it complains.

Last edited by Mr.Elendig (2010-09-26 12:44:29)

The Avatar of Time · 2010-09-26 19:15:37

Thanks for the reply. I've updated my main post so I'm deleting this to cut down on redundancy.

Thanks for the advice.

Last edited by The Avatar of Time (2010-10-07 06:05:40)

The Avatar of Time · 2010-09-28 04:33:54

*** Deleted by author.

Last edited by The Avatar of Time (2010-10-07 06:10:16)

The Avatar of Time · 2010-09-28 04:50:44

*** Deleted by author.

Last edited by The Avatar of Time (2010-10-07 06:09:46)

The Avatar of Time · 2010-09-28 05:39:27

*** Deleted by author.

Last edited by The Avatar of Time (2010-10-07 06:10:54)

The Avatar of Time · 2010-09-28 17:10:52

*** Deleted by author.

Last edited by The Avatar of Time (2010-10-07 06:11:21)

The Avatar of Time · 2010-10-07 07:19:46

Okay, I 'think' I finally have all of my information correctly, and I've not solved the issue yet.

The Avatar of Time · 2010-10-12 04:05:19

Well, I uninstalled 'mdadm' and now my drives mount. With the exception of media-server/4, which is apparently screwed up now.

However, uninstalling 'mdadm' isn't actually getting rid of the problem so much as hiding it. Apparently the information about the array still exists, and I'd like to be able to get rid of it.

I'm curious though, does my post just not make any sense at all, or have I not done something I was supposed to do before asking questions, or does no one know anything about the issue and how to get rid of an array?

Last edited by The Avatar of Time (2010-10-12 04:07:21)

The Avatar of Time · 2010-10-12 04:09:27

This is the output from checking /dev/sde1 (media-server/4):

 
KDE Partition Manager: Operation Report
Date:  2010-10-11 22:09
  Program version:  1.0.3
  LibParted version:  2.3
  KDE version:  4.5.1 (KDE 4.5.1)
  Machine:  Linux Avalon 2.6.35-ARCH #1 SMP PREEMPT Wed Sep 29 08:45:18 CEST 2010 x86_64
  User ID:  0

   
Check and repair partition '/dev/sde1' (465.76 GiB, reiser) 
Job: Check file system on partition '/dev/sde1' 
Command: fsck.reiserfs --fix-fixable -q -y /dev/sde1 
reiserfsck 3.6.21 (2009 www.namesys.com)

*************************************************************
** If you are using the latest reiserfsprogs and  it fails **
** please  email bug reports to reiserfs-list@namesys.com, **
** providing  as  much  information  as  possible --  your **
** hardware,  kernel,  patches,  settings,  all reiserfsck **
** messages  (including version),  the reiserfsck logfile, **
** check  the  syslog file  for  any  related information. **
** If you would like advice on using this program, support **
** is available  for $25 at  www.namesys.com/support.html. **
*************************************************************

Will check consistency of the filesystem on /dev/sde1
and will fix what can be fixed without --rebuild-tree
Will put log info to 'stdout'
###########
reiserfsck --fix-fixable started at Mon Oct 11 22:07:20 2010
###########
Replaying journal: 
Replaying journal: Done.
Reiserfs journal '/dev/sde1' in blocks [18..8211]: 0 transactions replayed
Checking internal tree..  \finished
Comparing bitmaps..Bad nodes were found, Semantic pass skipped
1 found corruptions can be fixed only when running with --rebuild-tree
###########
reiserfsck finished at Mon Oct 11 22:08:07 2010
###########
block 115043731: The level of the node (16352) is not correct, (4) expected
 the problem in the internal node occured (115043731), whole subtree is skipped
vpf-10630: The on-disk and the correct bitmaps differs. Will be fixed later.


 
Check file system on partition '/dev/sde1': Error
 
Check and repair partition '/dev/sde1' (465.76 GiB, reiser): Error

The Avatar of Time · 2010-10-12 04:46:33

What I can't understand is, since mdadm has always been installed, why would my drives mount for a while after a fresh installation, and then stop? mdadm was already there, I didn't install anything else RAID related later, so shouldn't the drives have always of been busy and refused to mount?

lilsirecho · 2010-10-12 15:18:29

Perhaps your drives cannot be mounted because they contain raid ID's and do not mount except as raid devices in an array.

I suggest that those drives might be zeroed out with DD to eliminate any residual ID data and perhaps MBR as well.

Most raid info is contained in .../etc/proc/mdstat.... which will list all raid elements present in the system.

Also, .../etc/mdadm.conf ....lists raid arrays as the last item in the file. This file could be deleted. It is used to permit raid enable during boot.

Drives which contain raid data cannot be mounted individually but can be assembled into a raid array as determined by the array initially setup.

Hope this helps ....................

steve_v · 2010-10-13 06:19:56

Have your drives ever been part of a soft raid array?

---8<---
md/raid:md127: device sdf operational as raid disk 3
md/raid:md127: device sde operational as raid disk 2
md/raid:md127: device sdc operational as raid disk 0
md/raid:md127: device sdd operational as raid disk 1
md/raid:md127: allocated 4272kB
md/raid:md127: raid level 5 active with 4 out of 4 devices, algorithm 2
RAID conf printout:
--- level:5 rd:4 wd:4
disk 0, o:1, dev:sdc
disk 1, o:1, dev:sdd
disk 2, o:1, dev:sde
disk 3, o:1, dev:sdf
md127: detected capacity change from 0 to 1500323315712
md127: p1
--->8---

If so I would sugest running mdadm --zero-superblock on them. :-)
This will remove any software raid signatures, as for dmraid, I'm not sure.

Last edited by steve_v (2010-10-13 06:40:06)

The Avatar of Time · 2010-10-13 06:58:02

Thanks for the reply.

Perhaps your drives cannot be mounted because they contain raid ID's and do not mount except as raid devices in an array.

I think that's partially correct, but I'm not sure how/why.

A couple years ago, when I first built my computer, I had tried setting up a RAID array to use the drives with. But it wasn't working the way I wanted so I formatted and repartitioned my drives to ReiserFS, with one partition on each drive, to be used as individual drives. Not in an array anymore. I also formatted and repartitioned the drive containing swap, /, and /home. That seemed to work fine. I've been using them as individual drives ever since then (couple years now). It's only been very recently that they had stopped mounting.

When I first posted this they had quit mounting and were showing as busy. So the other day I uninstalled mdadm, to see if that helped, and now I can mount them again. Minus the one drive that must have gotten damaged.

That is what I'm puzzled about. I don't see how, if I had left some remnants of the array I tried to setup a long time ago, that my drives had been working fine till recently. I don't see what made the problem pop up because I've always had mdadm installed.

I suggest that those drives might be zeroed out with DD to eliminate any residual ID data and perhaps MBR as well.

I really can't zero out the entire drives right now because they're full of data that I have no way to backup.

I don't know as I've ever zeroed them out, but they were formatted and repartitioned back when I gave up on making a RAID array. I put one partition, the entire size of the drive, on each hard drive and used ext3 for the file system. Later I changed to ReiserFS for the file system (which I what I currently use for everything). This was before my drives were full of data. I don't recall for sure what file system I was using when trying to get RAID working, but it wasn't ReiserFS.

Also the drive that contains swap, /, and /home has been repartitioned several times since I had messed with RAID. I've changed the partition sizes and the file systems on them multiple times (used to use ext3, use ReiserFS now). I've always had them in the same order: swap, /, /home, with swap at the beginning of the drive, but I have changed the size of swap, /, and /home several times, until I got them how I liked them. Again I don't know as I've ever zeroed the drive/s though.

I didn't backup any files or anything from the old OS when I gave up on the array and reinstalled. / and /home were both formatted.

I could backup the drive with swap, /, and /home and zero it out, but I can't on the others. Just no way to backup their data.

Most raid info is contained in .../etc/proc/mdstat.... which will list all raid elements present in the system.
Also, .../etc/mdadm.conf ....lists raid arrays as the last item in the file. This file could be deleted. It is used to permit raid enable during boot.

Unfortunately, neither /etc/proc or /etc/mdadm.conf exist for me.

Drives which contain raid data cannot be mounted individually but can be assembled into a raid array as determined by the array initially setup.
Hope this helps ....................

I understand that, which is why I don't understand the problem I'm having. I've been using my drives for a long time as individual drives (couple years I believe) and they've worked quite nicely until very recently (the last month or two). And after uninstalling mdadm they work again. To the best of my knowledge and memory there isn't anything left of the old RAID array I messed around with a couple years ago.

Now clearly I'm either mistaken or the trouble I'm having isn't in any way related to the old array I had messed with (which I find unlikely). Either way it is somehow RAID related or else uninstalling mdadm wouldn't have helped anything.

I realize that since the drives are mounting again the problem is more or less fixed, but I'd like to find the underlying cause as well.

As to my one drive that still won't mount (media-server/4), I'm hoping to get it working and not lose what's on it. I tried mounting it from a Live DVD, just to be sure, and I get a cannot read superblock error from trying to mount it. Still reading up on this but it looks like I need to try to rebuild the tree.

On a somewhat odd note, if I unmount media-server/5 and run a check on it I get an error also (different error from media-server/4), yet it will mount.

Anyway, thank you for the advice. I wish I could be a bit more clear on the problem, but I really don't understand why it's happening.

The Avatar of Time · 2010-10-13 07:02:29

steve_v wrote:

Have your drives ever been part of a soft raid array?
---8<---
md/raid:md127: device sdf operational as raid disk 3
md/raid:md127: device sde operational as raid disk 2
md/raid:md127: device sdc operational as raid disk 0
md/raid:md127: device sdd operational as raid disk 1
md/raid:md127: allocated 4272kB
md/raid:md127: raid level 5 active with 4 out of 4 devices, algorithm 2
RAID conf printout:
--- level:5 rd:4 wd:4
disk 0, o:1, dev:sdc
disk 1, o:1, dev:sdd
disk 2, o:1, dev:sde
disk 3, o:1, dev:sdf
md127: detected capacity change from 0 to 1500323315712
md127: p1
--->8---
If so I would sugest running mdadm --zero-superblock on them. :-)
This will remove any software raid signatures, as for dmraid, I'm not sure.

Thanks for the reply.

Yes, they were a long time ago, but have been used as individual drives without any trouble (until just recently).

mdadm --zero-superblock

I'll check into that, thanks. Is that going to wipe all my data as well though?

Thanks for the advice.

The Avatar of Time · 2010-10-13 07:12:25

Also, back when I was trying to setup a RAID array, it wasn't just me working on it. A friend of mine was helping. He messed around with it quite a bit by himself, after I had decided it wasn't going to work for me, which is why I'm not 100% sure what all was done.

I thought I had completely gotten rid of it when I reinstalled/reformatted/etc., and since my drives worked fine individually until very recently I had no reason to think I'd overlooked anything.

steve_v · 2010-10-14 04:53:51

I was game enough to run it on a 5TB array full of data & I didn't loose anything - it will only remove the soft raid superblock, not the data.
The drives involved had once been part of another array, one where the md device was partitioned ie raid5 of /dev/sda /dev/sdb to create /dev/md0 then partitioned to create /dev/md0p1
I changed the setup to an array of partitions ie raid5 of /dev/sda1 /dev/sdb1 etc. to give /dev/md0.
All was fine for some months until an mdadm update changed the order in which it checks for raid superblocks... all of a sudden I had 2 failed arrays, neither of which would mount.
Running mdadm --zero-superblock on /dev/sda, /dev/sdb etc. was the solution, however running it on /dev/sda1, /dev/sdb1 would have meant a manual array reconstruction.

The Avatar of Time · 2010-10-14 07:07:36

steve_v wrote:

I was game enough to run it on a 5TB array full of data & I didn't loose anything - it will only remove the soft raid superblock, not the data.
The drives involved had once been part of another array, one where the md device was partitioned ie raid5 of /dev/sda /dev/sdb to create /dev/md0 then partitioned to create /dev/md0p1
I changed the setup to an array of partitions ie raid5 of /dev/sda1 /dev/sdb1 etc. to give /dev/md0.
All was fine for some months until an mdadm update changed the order in which it checks for raid superblocks... all of a sudden I had 2 failed arrays, neither of which would mount.
Running mdadm --zero-superblock on /dev/sda, /dev/sdb etc. was the solution, however running it on /dev/sda1, /dev/sdb1 would have meant a manual array reconstruction.

You're my hero!

I ran 'mdadm --zero-superblock /dev/sdc' and on the other three drives. It worked. I also know it was needed because I tried the command on a drive that was never part of an array and got an error "mdadm: Unrecognised md component device - /dev/sde". Also got that error when I tried to run it twice on one of the drives it worked on (meaning it did what it was supposed to the first time).

You did mean for me to only use the command on /dev/sdc and not on /dev/sdc1 right?

For the sake of completeness, in case someone else has a similar problem, I'd like to mention the full process.

I referenced this page also:
http://www.ducea.com/2009/03/08/mdadm-cheat-sheet/

It wasn't quite the same for me as the page says, and a couple of the things I did may have been pointless, I'm not sure, but here it is anyway:

I had removed mdadm the other day so I had to reinstall it and restart before doing this.

Then I had to stop the array. The page I linked to said to use:

mdadm --stop /dev/md0

However, for me I actually had to use:

mdadm --stop /dev/md/0_0

Once I did that, /dev/md/ disappeared entirely from /dev.

Not sure it was needed by I also used:

mdadm --stop /dev/md127

/dev/md127 did not disappear.

Those were the only two 'md' related items in /dev.

Then, again referencing the page I linked to I used:

mdadm --remove /dev/md127

That didn't seem to do anything, and nothing disappeared but...

Now for the more relevant part. As suggested by "steve_v" and the linked page:

mdadm --zero-superblock /dev/sdc

And repeat for any other drives in the array.

I restarted my computer again and now my drives mount individually.

Well, at least three of the four do. However, /dev/sde1 still won't mount. I'm still getting the error:

mount: /dev/sde1: can't read superblock

So I have to get it back to working, if possible. I've been reading the man page for 'reiserfsck'. Trying to decide whether I need to use "--rebuild-sb" or "--rebuild-tree" and/or which order to run them in.

I've tried moving some stuff around and freeing up enough space on another drive to use "dd_rescue" on /dev/sde1, but I can't come up with enough space to do so.

I've referenced the following pages:
http://linux.die.net/man/8/reiserfsck
http://www.cyberciti.biz/tips/repairing … rfsck.html

The man page for reiserfsck says:

--rebuild-sb
    This option recovers the superblock on a Reiserfs partition. Normally you only need this option if mount reports "read_super_block: can't find a reiserfs file system" and you are sure that a Reiserfs file system is there. But remember that if you have used some partition editor program and now you cannot find a filesystem, probably something has gone wrong while repartitioning and the start of the partition has been changed. If so, instead of rebuilding the super block on a wrong place you should find the correct start of the partition first.

Also, the second one warns against running "reiserfsck --rebuild-sb" more then once on the same drive.

So it seems I should try to rebuild the tree first.

I ran "reiserfsck --fix-fixable --logfile FixFixable.log /dev/sde1". I'd done so before, but thought I'd check again after fixing the mdadm issue. I still get the same thing though:

block 115043731: The level of the node (16352) is not correct, (4) expected
 the problem in the internal node occured (115043731), whole subtree is skipped
vpf-10630: The on-disk and the correct bitmaps differs. Will be fixed later.

Anyone have any advice on which to run, or in what order? Or something else I'm entirely overlooking?

Also, since I'd mentioned originally noticing the array in 'Disk Utility' (palimpsest), I looked there again to make sure, and no RAID array shows up any longer.

Thanks to everyone for the help on this. If I'm lucky perhaps I can get the last drive mounting without losing any data. I really don't think it's a matter of the drive failing because until I had removed 'mdadm' the first time I could boot to a Live DVD and mount the drive (I can't now, I get the same superblock error). Just seems unlikely to me it would have started failing at the same time I had the raid issue.

The Avatar of Time · 2010-10-14 07:13:06

I forgot to ask in the previous post, do I need to run 'mdadm --zero-superblock' on my main drive, /dev/sda? It wasn't part of the array, but it's the drive that contains swap, /, and home as well as my MBR. So I didn't know if there is RAID array related info stored on it as well or not.

Thanks.

Arch Linux

#1 2010-09-20 08:54:51

"Extra" Internal Hard Drives won't mount - swap, /, and /home work.

#2 2010-09-20 11:29:29

Re: "Extra" Internal Hard Drives won't mount - swap, /, and /home work.

#3 2010-09-23 05:32:23

Re: "Extra" Internal Hard Drives won't mount - swap, /, and /home work.

#4 2010-09-26 10:05:04

Re: "Extra" Internal Hard Drives won't mount - swap, /, and /home work.

#5 2010-09-26 12:42:58

Re: "Extra" Internal Hard Drives won't mount - swap, /, and /home work.

#6 2010-09-26 19:15:37

Re: "Extra" Internal Hard Drives won't mount - swap, /, and /home work.

#7 2010-09-28 04:33:54

Re: "Extra" Internal Hard Drives won't mount - swap, /, and /home work.

#8 2010-09-28 04:50:44

Re: "Extra" Internal Hard Drives won't mount - swap, /, and /home work.

#9 2010-09-28 05:39:27

Re: "Extra" Internal Hard Drives won't mount - swap, /, and /home work.

#10 2010-09-28 17:10:52

Re: "Extra" Internal Hard Drives won't mount - swap, /, and /home work.

#11 2010-10-07 07:19:46

Re: "Extra" Internal Hard Drives won't mount - swap, /, and /home work.

#12 2010-10-12 04:05:19

Re: "Extra" Internal Hard Drives won't mount - swap, /, and /home work.

#13 2010-10-12 04:09:27

Re: "Extra" Internal Hard Drives won't mount - swap, /, and /home work.

#14 2010-10-12 04:46:33

Re: "Extra" Internal Hard Drives won't mount - swap, /, and /home work.

#15 2010-10-12 15:18:29

Re: "Extra" Internal Hard Drives won't mount - swap, /, and /home work.

#16 2010-10-13 06:19:56

Re: "Extra" Internal Hard Drives won't mount - swap, /, and /home work.

#17 2010-10-13 06:58:02

Re: "Extra" Internal Hard Drives won't mount - swap, /, and /home work.

#18 2010-10-13 07:02:29

Re: "Extra" Internal Hard Drives won't mount - swap, /, and /home work.

#19 2010-10-13 07:12:25

Re: "Extra" Internal Hard Drives won't mount - swap, /, and /home work.

#20 2010-10-14 04:53:51

Re: "Extra" Internal Hard Drives won't mount - swap, /, and /home work.

#21 2010-10-14 07:07:36

Re: "Extra" Internal Hard Drives won't mount - swap, /, and /home work.

#22 2010-10-14 07:13:06

Re: "Extra" Internal Hard Drives won't mount - swap, /, and /home work.

Board footer