Arch Linux

First of all "After the latest update" does not really say much.
Please provide the relevant part of pacman.log so we could see what packages were upgraded and their versions (kernel alone is not enough).
Also menu.lst and mkinitcpio.conf could be useful to see.

Well I just tried to downgrade the kernel to 2.6.30.2 and that didn't work still the same error. Where is pacman.log located?

Regards

André

/var/log/pacman.log

Here is the pacman.log from the 29th of july until today

[2009-07-29 18:50] Upgraded bzr successfully (1.16.1-1 -> 1.17-1)
[2009-07-29 18:50] Upgraded ed successfully (1.3-1 -> 1.4-1)
[2009-07-29 18:50] Upgraded git successfully (1.6.3.3-1 -> 1.6.4-1)
[2009-07-29 18:50] Upgraded man-pages successfully (3.21-1 -> 3.22-1)
[2009-07-29 18:50] Upgraded sqlite3 successfully (3.6.15-2 -> 3.6.16-1)
[2009-07-29 18:50] Upgraded tzdata successfully (2009j-1 -> 2009k-1)
[2009-07-30 18:52] Upgraded kdelibs3 successfully (3.5.10-4 -> 3.5.10-5)
[2009-07-30 18:52] Upgraded less successfully (429-1 -> 436-1)
[2009-07-30 18:52] Upgraded miro successfully (2.0.5-1 -> 2.5.1-1)
[2009-07-30 18:52] Upgraded vlc successfully (1.0.1-1 -> 1.0.1-2)
[2009-07-31 18:51] Upgraded abs successfully (2.3.2-2 -> 2.3.3-1)
[2009-07-31 18:51] Upgraded flashplugin successfully (10.0.22.87-1 -> 10.0.32.18-1.1)
[2009-08-01 08:42] synchronizing package lists
[2009-08-01 08:42] starting full system upgrade
[2009-08-01 08:42] starting full system upgrade
[2009-08-01 08:43] upgraded libmysqlclient (5.1.36-1 -> 5.1.37-1)
[2009-08-01 08:43] upgraded mysql-clients (5.1.36-1 -> 5.1.37-1)
[2009-08-01 08:43] upgraded mysql (5.1.36-1 -> 5.1.37-1)
[2009-08-01 08:43] upgraded amarok (2.1-1 -> 2.1.1-1)
[2009-08-01 08:43] upgraded dosfstools (3.0.1-1 -> 3.0.5-1)
[2009-08-01 08:43] upgraded gpgme (1.1.8-1 -> 1.2.0-1)
[2009-08-01 08:43] upgraded kipi-plugins (0.3.0-2 -> 0.5.0-1)
[2009-08-01 08:43] upgraded tdb (3.3.6-1 -> 3.3.7-1)
[2009-08-01 08:43] upgraded smbclient (3.3.6-2 -> 3.3.7-1)
[2009-08-01 08:43] upgraded samba (3.3.6-2 -> 3.3.7-1)
[2009-08-01 13:19] synchronizing package lists
[2009-08-01 13:19] starting full system upgrade
[2009-08-01 13:19] starting full system upgrade
[2009-08-01 15:53] synchronizing package lists
[2009-08-01 15:53] starting full system upgrade
[2009-08-01 15:53] starting full system upgrade
[2009-08-01 15:59] synchronizing package lists
[2009-08-01 15:59] starting full system upgrade
[2009-08-01 15:59] starting full system upgrade
[2009-08-01 16:00] >>>
[2009-08-01 16:00] >>> You appear to be using the LILO bootloader. You should run
[2009-08-01 16:00] >>> 'lilo' before rebooting.
[2009-08-01 16:00] >>>
[2009-08-01 16:00] >>> Updating module dependencies. Please wait ...
[2009-08-01 16:00] >>> MKINITCPIO SETUP
[2009-08-01 16:00] >>> ----------------
[2009-08-01 16:00] >>> If you use LVM2, Encrypted root or software RAID,
[2009-08-01 16:00] >>> Ensure you enable support in /etc/mkinitcpio.conf .
[2009-08-01 16:00] >>> More information about mkinitcpio setup can be found here:
[2009-08-01 16:00] >>> http://wiki.archlinux.org/index.php/Mkinitcpio
[2009-08-01 16:00]
[2009-08-01 16:00] >>> Generating initial ramdisk, using mkinitcpio. Please wait...
[2009-08-01 16:00] ==> Building image "default"
[2009-08-01 16:00] ==> Running command: /sbin/mkinitcpio -k 2.6.30-ARCH -c /etc/mkinitcpio.conf -g /boot/kernel26.img
[2009-08-01 16:00] :: Begin build
[2009-08-01 16:00] :: Parsing hook [base]
[2009-08-01 16:00] :: Parsing hook [udev]
[2009-08-01 16:00] :: Parsing hook [autodetect]
[2009-08-01 16:00] :: Parsing hook [pata]
[2009-08-01 16:00] :: Parsing hook [scsi]
[2009-08-01 16:00] :: Parsing hook [sata]
[2009-08-01 16:00] :: Parsing hook [usb]
[2009-08-01 16:00] :: Parsing hook [fw]
[2009-08-01 16:00] :: Parsing hook [pcmcia]
[2009-08-01 16:00] :: Parsing hook [net]
[2009-08-01 16:00] :: Parsing hook [usbinput]
[2009-08-01 16:00] :: Parsing hook [keymap]
[2009-08-01 16:00] :: Parsing hook [encrypt]
[2009-08-01 16:00] :: Parsing hook [dmraid]
[2009-08-01 16:00] :: Parsing hook [filesystems]
[2009-08-01 16:00] :: Parsing hook [splashy]
[2009-08-01 16:00] :: Generating module dependencies
[2009-08-01 16:00] :: Generating image '/boot/kernel26.img'...SUCCESS
[2009-08-01 16:00] ==> SUCCESS
[2009-08-01 16:00] ==> Building image "fallback"
[2009-08-01 16:00] ==> Running command: /sbin/mkinitcpio -k 2.6.30-ARCH -c /etc/mkinitcpio.conf -g /boot/kernel26-fallback.img -S autodetect
[2009-08-01 16:00] :: Begin build
[2009-08-01 16:00] :: Parsing hook [base]
[2009-08-01 16:00] :: Parsing hook [udev]
[2009-08-01 16:00] :: Parsing hook [pata]
[2009-08-01 16:00] :: Parsing hook [scsi]
[2009-08-01 16:00] :: Parsing hook [sata]
[2009-08-01 16:01] :: Parsing hook [usb]
[2009-08-01 16:01] :: Parsing hook [fw]
[2009-08-01 16:01] :: Parsing hook [pcmcia]
[2009-08-01 16:01] :: Parsing hook [net]
[2009-08-01 16:01] :: Parsing hook [usbinput]
[2009-08-01 16:01] :: Parsing hook [keymap]
[2009-08-01 16:01] :: Parsing hook [encrypt]
[2009-08-01 16:01] :: Parsing hook [dmraid]
[2009-08-01 16:01] :: Parsing hook [filesystems]
[2009-08-01 16:01] :: Parsing hook [splashy]
[2009-08-01 16:01] :: Generating module dependencies
[2009-08-01 16:01] :: Generating image '/boot/kernel26-fallback.img'...SUCCESS
[2009-08-01 16:01] ==> SUCCESS
[2009-08-01 16:01] upgraded kernel26 (2.6.30.2-1 -> 2.6.30.4-1)
[2009-08-01 16:01] upgraded winetricks (20090607-1 -> 20090716-1)
[2009-08-01 16:01] upgraded pciutils (3.1.2-1 -> 3.1.3-1)
[2009-08-01 16:16]
[2009-08-01 16:16] NOTE
[2009-08-01 16:16] ----
[2009-08-01 16:16] # If you want to use this driver with sane:
[2009-08-01 16:16] # echo "hpaio" >> /etc/sane.d/dll.conf
[2009-08-01 16:16]
[2009-08-01 16:16]
[2009-08-01 16:16] UPGRADING
[2009-08-01 16:16] ----
[2009-08-01 16:16] # This version no longer uses an init script. You should remove hplip
[2009-08-01 16:16] # from the /etc/rc.conf daemon list.
[2009-08-01 16:16]
[2009-08-01 16:16] upgraded hplip (3.9.4b-2 -> 3.9.6b-1)
[2009-08-01 16:16] upgraded ocaml (3.11.0-2 -> 3.11.1-1)
[2009-08-01 18:04] Upgraded libnice successfully (0.0.8-2 -> 0.0.9-1)
[2009-08-01 18:04] Upgraded python-telepathy successfully (0.15.8-1 -> 0.15.10-1)
[2009-08-01 18:04] papyon (0.4.1-1) installed successfully!
[2009-08-01 18:04] Upgraded telepathy-butterfly successfully (0.3.4-1 -> 0.5.0-1)
[2009-08-01 18:04] Upgraded telepathy-sofiasip successfully (0.4.5-4 -> 0.5.17-1)
[2009-08-01 21:03] Upgraded aria2 successfully (1.3.0-1 -> 1.5.1-1)
[2009-08-01 21:03] Upgraded autoconf successfully (2.63-1 -> 2.64-1)
[2009-08-01 21:03] Upgraded hal-info successfully (0.20090414-1 -> 0.20090716-1)
[2009-08-01 21:03] Upgraded pinentry successfully (0.7.5-5 -> 0.7.6-1)
[2009-08-01 21:03] Upgraded sbcl successfully (1.0.28-1 -> 1.0.30-1)
[2009-08-02 13:13] >>>
[2009-08-02 13:13] >>> You appear to be using the LILO bootloader. You should run
[2009-08-02 13:13] >>> 'lilo' before rebooting.
[2009-08-02 13:13] >>>
[2009-08-02 13:13] >>> Updating module dependencies. Please wait ...
[2009-08-02 13:13] >>> MKINITCPIO SETUP
[2009-08-02 13:13] >>> ----------------
[2009-08-02 13:13] >>> If you use LVM2, Encrypted root or software RAID,
[2009-08-02 13:13] >>> Ensure you enable support in /etc/mkinitcpio.conf .
[2009-08-02 13:13] >>> More information about mkinitcpio setup can be found here:
[2009-08-02 13:13] >>> http://wiki.archlinux.org/index.php/Mkinitcpio
[2009-08-02 13:13]
[2009-08-02 13:13] >>> Generating initial ramdisk, using mkinitcpio. Please wait...
[2009-08-02 13:13] ==> Building image "default"
[2009-08-02 13:13] ==> Running command: /sbin/mkinitcpio -k 2.6.30-ARCH -c /etc/mkinitcpio.conf -g /boot/kernel26.img
[2009-08-02 13:13] :: Begin build
[2009-08-02 13:13] :: Parsing hook [base]
[2009-08-02 13:13] :: Parsing hook [udev]
[2009-08-02 13:13] :: Parsing hook [autodetect]
[2009-08-02 13:13] :: Parsing hook [pata]
[2009-08-02 13:13] :: Parsing hook [scsi]
[2009-08-02 13:13] :: Parsing hook [sata]
[2009-08-02 13:13] :: Parsing hook [usb]
[2009-08-02 13:13] :: Parsing hook [fw]
[2009-08-02 13:13] :: Parsing hook [pcmcia]
[2009-08-02 13:13] :: Parsing hook [net]
[2009-08-02 13:13] :: Parsing hook [usbinput]
[2009-08-02 13:13] :: Parsing hook [keymap]
[2009-08-02 13:13] :: Parsing hook [encrypt]
[2009-08-02 13:13] :: Parsing hook [dmraid]
[2009-08-02 13:13] :: Parsing hook [filesystems]
[2009-08-02 13:13] :: Parsing hook [splashy]
[2009-08-02 13:13] :: Generating module dependencies
[2009-08-02 13:13] :: Generating image '/boot/kernel26.img'...SUCCESS
[2009-08-02 13:13] ==> SUCCESS
[2009-08-02 13:13] ==> Building image "fallback"
[2009-08-02 13:13] ==> Running command: /sbin/mkinitcpio -k 2.6.30-ARCH -c /etc/mkinitcpio.conf -g /boot/kernel26-fallback.img -S autodetect
[2009-08-02 13:13] :: Begin build
[2009-08-02 13:13] :: Parsing hook [base]
[2009-08-02 13:13] :: Parsing hook [udev]
[2009-08-02 13:13] :: Parsing hook [pata]
[2009-08-02 13:13] :: Parsing hook [scsi]
[2009-08-02 13:13] :: Parsing hook [sata]
[2009-08-02 13:13] :: Parsing hook [usb]
[2009-08-02 13:13] :: Parsing hook [fw]
[2009-08-02 13:13] :: Parsing hook [pcmcia]
[2009-08-02 13:13] :: Parsing hook [net]
[2009-08-02 13:13] :: Parsing hook [usbinput]
[2009-08-02 13:13] :: Parsing hook [keymap]
[2009-08-02 13:13] :: Parsing hook [encrypt]
[2009-08-02 13:13] :: Parsing hook [dmraid]
[2009-08-02 13:13] :: Parsing hook [filesystems]
[2009-08-02 13:13] :: Parsing hook [splashy]
[2009-08-02 13:13] :: Generating module dependencies
[2009-08-02 13:14] :: Generating image '/boot/kernel26-fallback.img'...SUCCESS
[2009-08-02 13:14] ==> SUCCESS
[2009-08-02 13:14] upgraded kernel26 (2.6.30.4-1 -> 2.6.30.2-1)

and menu.lst

# general configuration:
timeout 5
default 0
color light-blue/black light-cyan/blue
splashimage /boot/grub/splash.xpm.gz

# boot sections follow
# each is implicitly numbered from 0 in the order of appearance below
#
# TIP: If you want a 1024x768 framebuffer, add "vga=773" to your kernel line.
#
#-*

# (0) Arch Linux
title Arch Linux
root (hd0,0)
kernel /boot/vmlinuz26 root=/dev/mapper/isw_baaggideei_Volume01 ro quiet vga=791 splash
initrd /boot/kernel26.img

# (1) Arch Linux
title Arch Linux Fallback
root (hd0,0)
kernel /boot/vmlinuz26 root=/dev/mapper/isw_baaggideei_Volume01 ro
initrd /boot/kernel26-fallback.img

# (1) Windows
#title Windows
#rootnoverify (hd0,0)
#makeactive
#chainloader +1

@Thomas: any ideas what might happen here?

Nothing to indicate failure, my best guess is dmraid failing.

If its dmraid that is failing, that I don't understand because I updated dmraid from 1.0.0.rc14 to 1.0.0.rc15 and had no problems there. The issue I'm observing now started after yesterdays updates.

Regards

André

dmraid changed device naming to name0p1 and no more name01.
Partitions start now with <p>. You need to change that.

Where do I need to do those changes?

Regards

André

grub config and fstab needs to be changed

I did the changes now in menu.lst and fstab and now it boots a bit further but when it gets to checking filesystems it fails. It gives the error

/dev/mapper/isw_baaggideei_Volume0p1: clean (root partition)
/dev/mapper/isw_baaggideei_Volume0p3: (home partition)

the superblock could not be read or does not describe a correct ext2 filesystem. If the device is valid and it really contains an ext2 filesystem (and not swap or ufs or something else), then the superblock is corrupt, and you might try running e2tsck with an alternative superblock:
e2fsck -b 8193 <device>

then it wants me to give my root password in order to do the maintenance. My system has 3 partitions p1 is root, p2 is swap and p3 is /home. So what now? Is my home partition gone or do I do what is suggested?

Regards

André

I updated dmraid almost a week ago and I didn't need to do those changes to the partitions until now when the kernel update came out and I've restating my machine several times since then.

I got my machine running again by downgrading dmraid to 1.0.0.rc14-4 (from March) and changing the naming scheme back to the old one for the partitions. So dmraid 1.0.0.rc15 has some sort of bug still.

Regards

André

Any progress on this? Is there anything I can do figure out why dmraid-1.0.0.rc15 isn't working for me? Is anybody else having problems with it? Should report it upstream?

Regards

André

> Is anybody else having problems with it?
It's better to ask this on forums where more people can see your post.

> Should report it upstream?
You can. If it's upstream bug - you'll help fixing it. If it's not - upstream devs will close it as 'Not a bug'.

dmraid is not perfect, 1.0.15 has some nice features 1.0.14 lacks, please report upstream i cannot do more then adding patches from others distros.
At the moment only fedora patches are applied.

OK, I'll do that and make a post in the forum to if others are having this issue.

Regards

André

Here is the thread I started in the forum

http://bbs.archlinux.org/viewtopic.php?id=77329

sofar nobody is experiencing the issue I'm having :(. Is there a way to bypass the "checking filesystems" during boot completely? It would be a good way to test my installation, if it can't boot the machine at all.

Regards

André

in rc.sysinit replace
if [ -x /sbin/fsck ]; then
stat_busy "Checking Filesystems"
with something like
if [ /bin/false ]; then

(or temporary chmod -x /sbin/fsck but don't forget to revert it later)

OK, thanks I'll try that.

chmod -x /sbin/fsck

How do I revert back? Just by running the above command again?

Regards

André

to revert back use +x instead of -x (both as root)

I tried again using the latest dmraid and did what the people in the forum mentioned but still I get the error. I noticed now that right before my Arch starts to boot I get this error

mknod: File exists
Error: dos: partition address past end of RAID device
Commandline read: splashy
Commandline read: boot

as you can see I'm using splashy, therefore I'm using initscripts-splashy which I updated manually when the new filesystems and udev entered [core]. Could it be that my initscripts is causing this because I'm using older files? I will try next to turn off fsck and see if it will boot the machine at all and to see if my /home directory gets mounted at all. I'm really running out of ideas here...

Regards

André

You should've mentioned that you use custom initscripts right from the start ;).
Yes, it could be. I suggest trying default initscripts.
Also, there may be possibility that something is wrong with your partition table. You may use testdisk (carefully!) to examine it.

OK, I'll try to switch back to initscripts. If it is the partition table that is screwed up. Is there a way to fix it without reinstalling the machine?

Regards

André

in case it's partition table problem - testdisk is the best tool for fixing partition tables I know (but that should be done only after extensive reading of documentation and backing up all important data).

Hmm... I really don't get it. I have had no problems with my RAID, dmraid and fsck when running 1.0.0.rc14. No errors or anything at any check I've done at boot.

Regards

André

well, if tpowa's suggestions didn't work for you then I don't know why the upgrade made it fail on your system

What about this error I mentioned earlier? (comes right after grub)

Error: dos: partition address past end of RAID device

Regards

André

I have no idea what causes it

I just tried to run

e2fsck -b 8193 <device>

on the partition that fails, e.g.

e2fsck -b /dev/mapper/isw_baaggideei_Volume0p3

and I get the error that it can't find it. Have I changed to the wrong naming scheme?

Regards

André

I just removed splashy and initscripts-splashy and everything else bound to it and installed initcripts and latest dmraid again and if test the dmraid with the new naming scheme I get

[root@andre ~]# dmraid -tay
isw_baaggideei_Volume0: 0 1953536000 striped 2 256 /dev/sda 0 /dev/sdb 0
ERROR: dos: partition address past end of RAID device
isw_baaggideei_Volume0p1: 0 199993122 linear /dev/mapper/isw_baaggideei_Volume0 63
isw_baaggideei_Volume0p2: 0 8000370 linear /dev/mapper/isw_baaggideei_Volume0 199993185

There is no isw_baaggideei_Volume0p3 which is my /home directory it seems to be named isw_baaggideei_Volume0.

I misinterpreted the above this what dmraid says using the old version

isw_baaggideei_Volume0: 0 1953536512 striped 2 256 /dev/sda 0 /dev/sdb 0
isw_baaggideei_Volume01: 0 199993122 linear /dev/mapper/isw_baaggideei_Volume0 63
isw_baaggideei_Volume02: 0 8000370 linear /dev/mapper/isw_baaggideei_Volume0 199993185
isw_baaggideei_Volume03: 0 1745542575 linear /dev/mapper/isw_baaggideei_Volume0 207993555

clearly dmraid-1.0.0.rc15 is haivng an issue with my last partition, while dmraid-1.0.0.rc14 doesn't have an issue.

Regards

André

hi all,

first of all, I was using dmraid about 6 months ago, but I am not using it anymore (I am using softraid now, at the cost of windows not being on raid anymore), so I speak about what I remember about dmraid.

- It seems to me that it is the second time that dmraid is tried to upgrade from rc14 to rc15 release, and the first try ended in a downgrade. (see http://bbs.archlinux.org/viewtopic.php?id=60794)

- "Error: dos: partition address past end of RAID device" :
This looks like the "attempt to access beyond end of device" from this topic on the forum : http://bbs.archlinux.org/viewtopic.php?id=62502
I have explained this on the forum, but it was only a warning that could be ignored, not an error, so I am not sure this is the same thing.

- "I updated dmraid almost a week ago and I didn't need to do those changes to the partitions until now when the kernel update came out and I've restating my machine several times since then."
I had exactly the same at the first rc15 update, your story confirms what I had thought about that at the end : the problem is that, as you are using the dmraid hook in the initcpio image, the dmraid module is always loaded from the initcpio image, never from the system. So the naming scheme problem arise in 3 steps :
- before the dmraid update :
the dmraid module in the system and in the initcpio image are the rc14, which expect the old naming scheme, no problem.
- after the dmraid update, but before any kernel update :
the dmraid module in the system is the rc15, but the one in the initcpio image is still the rc14, since the initcpio image has not been rebuilt.
the dmraid module being loaded from the initcpio image, the rc14 version is loaded, still expecting the old naming scheme, that's why your system was working fine.
- after the first kernel update :
the initcpio image was rebuilt by the kernel update, so the last dmraid module version was included in it.
at the first reboot, the rc15 version is loaded from the initcpio image, expecting the new naming scheme.

That's why your system hangs just after the kernel update, but I don't know why your system still hangs after updating and reconfiguring everything.
I think you should try to update to rc15 as cleanly as possible :
- downgrade to a fully functional system, with rc14 dmraid and old naming scheme in menu.lst and fstab (rebuild the initcpio image after that).
- reboot and check that everything works fine.
- update dmraid to rc15 (do not reboot here).
- change menu.lst and fstab to the new naming scheme (before rebuilding the initcpio images, so that fstab in the initcpio image is up to date).
- rebuild the initcpio images.
- reboot.

If this does not work, I can't help more since I am not using dmraid anymore.

Thanks for your suggestions but I already tried that. My problem is that rc15 works but for some reason it can't see all my partitions, e.g.

[root@andre ~]# dmraid -tay
isw_baaggideei_Volume0: 0 1953536000 striped 2 256 /dev/sda 0 /dev/sdb 0
ERROR: dos: partition address past end of RAID device
isw_baaggideei_Volume0p1: 0 199993122 linear /dev/mapper/isw_baaggideei_Volume0 63
isw_baaggideei_Volume0p2: 0 8000370 linear /dev/mapper/isw_baaggideei_Volume0 199993185

isw_baaggideei_Volume0p3 is missing for some reason....

http://bugs.debian.org/cgi-bin/bugreport.cgi?bug=523660
seems to be a bit of same kind, could it be that your array has multiple superblocks?
Have you changed your controller type?

No, I haven't still using sata_sil and I haven't touched my RAIS setup in the bios at all since I set it the first time and did the initial i9nstallation of Arch.

Regards

André

I read through the debian bug report and it seems like the way this guy got his RAID working with 1.0.0.rc15 was by recreating his RAID array. I'm not quite sure what he means by that but I assume it is by resetting it in bios and rebuilding it there.

Regards

André

I don'T know but perhaps during first creation of array something went wrong, which is now triggered by dmraid 1.0.15.
What you can still do. just install old dmraid package and set it to ignorepkg in etc/pacman.conf.

Hmm, this sucks a bit but I don't want to reinstall either ad do all the RAID stuff again. Plus I don't have a place to put 300 GB of data to do backup, at the moment. When will the next live CD come out where dmraid 1.0.0.rc15 is included?

Regards

André

archboot isos here contain it:
http://bbs.archlinux.org/viewtopic.php?id=77017

new official rc isos i think should include it too:
http://bbs.archlinux.org/viewtopic.php?id=77680

@André: new official ISOs were released. Can you provide any update on this bug?

Yeah, I saw that but I don't have the possibility to back up my data at the moment and do a reinstall. So I have no update on this bug. I could try to see if I could perform a rescue session to see if I can see all my partitions with 2009.08.

Regards

André

Any status/notices of this issue? and/or is still "critical"?

I'm gonna reswinstall my machine next week (moving to 64bit). I'll report back how my raid is running with the latest dmraid then. You can downgrade the priority if you want to.

Regards

André

Thanks!

This is fixed now. I'm doing fresh installation of arch and my raid was messed up in the bios so I had to set it up again. So close bug this please!