[arch-general] Broke my system again - device-mapper & readline conflict - best way to fix?
Listmates, I have more fun in store for this weekend. After testing dmraid-1.0.0rc15, I had downgraded a number of packages that were installed from testing back to there normal versions but had left dmraid-1.0.0rc15 and device mapper from testing. Everything was working fine. Apparently some packages were moved from testing to extra or another normal repo because all of a sudden I began getting readline...so.6 error (from memory) on boot and was dumped into maintenance mode. In maintenance mode, I remounted to root filesystem rw and then mounted all the partitions and ran pacman -Sy readline (testing is disabled) After installing the latest readline, my box will not boot. I now gives a readline...so.5 error and never gets to maintenance mode. My question is "What is the best way to try and recover?" I have two options to work on the archlinux install: (1) the machine is dual boot with openSuSE so I can boot to suse and then mount the Archlinux filesystem (rw) under /mnt/arch. I'm not sure what I can do here unless there is a way for me to manually unpack some pacman packages and overwrite the problem files on the Arch install in this configuration; or (2) boot using the Arch install disk. Here is where I'm a little lost on the recovery procedure. I can manually assemble my raid array after booting to the install disk, but then what next? How would I go about reinstalling the various packages from either testing or extra when I have booted from the install disk? Thus my need for help. What say the experts? How best to go about fixing the device-mapper readline conflict? -- David C. Rankin, J.D.,P.E. Rankin Law Firm, PLLC 510 Ochiltree Street Nacogdoches, Texas 75961 Telephone: (936) 715-9333 Facsimile: (936) 715-9339 www.rankinlawfirm.com
On Fri, 2009-07-10 at 19:26 -0500, David C. Rankin wrote:
Listmates,
I have more fun in store for this weekend. After testing dmraid-1.0.0rc15, I had downgraded a number of packages that were installed from testing back to there normal versions but had left dmraid-1.0.0rc15 and device mapper from testing. Everything was working fine.
Apparently some packages were moved from testing to extra or another normal repo because all of a sudden I began getting readline...so.6 error (from memory) on boot and was dumped into maintenance mode. In maintenance mode, I remounted to root filesystem rw and then mounted all the partitions and ran pacman -Sy readline (testing is disabled)
After installing the latest readline, my box will not boot. I now gives a readline...so.5 error and never gets to maintenance mode. My question is "What is the best way to try and recover?"
I have two options to work on the archlinux install:
(1) the machine is dual boot with openSuSE so I can boot to suse and then mount the Archlinux filesystem (rw) under /mnt/arch. I'm not sure what I can do here unless there is a way for me to manually unpack some pacman packages and overwrite the problem files on the Arch install in this configuration; or
(2) boot using the Arch install disk. Here is where I'm a little lost on the recovery procedure. I can manually assemble my raid array after booting to the install disk, but then what next? How would I go about reinstalling the various packages from either testing or extra when I have booted from the install disk?
Thus my need for help. What say the experts? How best to go about fixing the device-mapper readline conflict?
I am using bash as my shell and I was bitten by this bed bug as well. This is how I fixed it.... Fetch bash from an archlinux repos put the bash package some place where you can get to it boot from arch install/live cd chroot to the broken system pacman -U bash...... reboot
On Friday 10 July 2009 08:41:36 pm Baho Utot wrote:
On Fri, 2009-07-10 at 19:26 -0500, David C. Rankin wrote:
Listmates,
I have more fun in store for this weekend. After testing dmraid-1.0.0rc15, I had downgraded a number of packages that were installed from testing back to there normal versions but had left dmraid-1.0.0rc15 and device mapper from testing. Everything was working fine.
Apparently some packages were moved from testing to extra or another normal repo because all of a sudden I began getting readline...so.6 error (from memory) on boot and was dumped into maintenance mode. In maintenance mode, I remounted to root filesystem rw and then mounted all the partitions and ran pacman -Sy readline (testing is disabled)
After installing the latest readline, my box will not boot. I now gives a readline...so.5 error and never gets to maintenance mode. My question is "What is the best way to try and recover?"
I have two options to work on the archlinux install:
(1) the machine is dual boot with openSuSE so I can boot to suse and then mount the Archlinux filesystem (rw) under /mnt/arch. I'm not sure what I can do here unless there is a way for me to manually unpack some pacman packages and overwrite the problem files on the Arch install in this configuration; or
(2) boot using the Arch install disk. Here is where I'm a little lost on the recovery procedure. I can manually assemble my raid array after booting to the install disk, but then what next? How would I go about reinstalling the various packages from either testing or extra when I have booted from the install disk?
Thus my need for help. What say the experts? How best to go about fixing the device-mapper readline conflict?
I am using bash as my shell and I was bitten by this bed bug as well.
This is how I fixed it....
Fetch bash from an archlinux repos
put the bash package some place where you can get to it
boot from arch install/live cd
chroot to the broken system
pacman -U bash......
reboot
Baho, thanks. Here is my plan. My laptop is also running arch, testing, kde4, so I did a system update and I have rsync'ed the pacman/pkg files to my openSuSE server where the arch filesttem Arch server where the dmraid issue is occurring. I'll boot from the CD, chroot my filesystem and then just pacman -Su to install all of the updated packages and then it should be fixed (and if it's not, I'm too tired to jack with it for the rest of the night, so it will be tomorrow before I know whether I will have the dev-mapper readline issues solved -- David C. Rankin, J.D.,P.E. Rankin Law Firm, PLLC 510 Ochiltree Street Nacogdoches, Texas 75961 Telephone: (936) 715-9333 Facsimile: (936) 715-9339 www.rankinlawfirm.com
On Friday 10 July 2009 08:41:36 pm Baho Utot wrote:
On Fri, 2009-07-10 at 19:26 -0500, David C. Rankin wrote:
Listmates,
I have more fun in store for this weekend. After testing dmraid-1.0.0rc15, I had downgraded a number of packages that were installed from testing back to there normal versions but had left dmraid-1.0.0rc15 and device mapper from testing. Everything was working fine.
Apparently some packages were moved from testing to extra or another normal repo because all of a sudden I began getting readline...so.6 error (from memory) on boot and was dumped into maintenance mode. In maintenance mode, I remounted to root filesystem rw and then mounted all the partitions and ran pacman -Sy readline (testing is disabled)
After installing the latest readline, my box will not boot. I now gives a readline...so.5 error and never gets to maintenance mode. My question is "What is the best way to try and recover?"
I have two options to work on the archlinux install:
(1) the machine is dual boot with openSuSE so I can boot to suse and then mount the Archlinux filesystem (rw) under /mnt/arch. I'm not sure what I can do here unless there is a way for me to manually unpack some pacman packages and overwrite the problem files on the Arch install in this configuration; or
(2) boot using the Arch install disk. Here is where I'm a little lost on the recovery procedure. I can manually assemble my raid array after booting to the install disk, but then what next? How would I go about reinstalling the various packages from either testing or extra when I have booted from the install disk?
Thus my need for help. What say the experts? How best to go about fixing the device-mapper readline conflict?
I am using bash as my shell and I was bitten by this bed bug as well.
This is how I fixed it....
Fetch bash from an archlinux repos
put the bash package some place where you can get to it
boot from arch install/live cd
chroot to the broken system
pacman -U bash......
reboot
Baho, all: Err.. My system is still broken and I can't figure out how to get around the readline error. I should have thought about this before going through the trouble of trying to set up the chroot to install bash, but the error I get when I try to execute chroot is thee *same* error I was getting when I tried to boot the system anyway. To try and recover, I booted from the install CD and then logged in as root. Here is how I set up to create the chroot environment and the error I received: modprobe dm_mod modprobe sata_sil (or whatever chipset driver you need) dmraid -ay mkdir -p /mnt/rscu mount /dev/mapper/nvidia_ecaejfdi5 /mnt/rscu mount /dev/mapper/nvidia_ecaejfdi6 /mnt/rscu/boot mount /dev/mapper/nvidia_ecaejfdi7 /mnt/rscu/home mount /dev/mapper/nvidia_ecaejfdi9 /mnt/rscu/var mount /dev/mapper/nvidia_ecaejfdi10 /mnt/rscu/srv cd /mnt/rscu mount -o bind /dev dev mount -o bind /proc proc mount -o bind /sys sys chroot . The error message: /bin/bash: error while loading shared libraries: libreadline.so.5: cannot open shared object file: No such file or directory How do I get around the libreadline error? -- David C. Rankin, J.D.,P.E. Rankin Law Firm, PLLC 510 Ochiltree Street Nacogdoches, Texas 75961 Telephone: (936) 715-9333 Facsimile: (936) 715-9339 www.rankinlawfirm.com
Download a copy of the new bash package by hand, and instead of the chroot command, do "pacman -r /mnt/rscu -U /path/to/bash_package". On Sat, Jul 11, 2009 at 3:20 PM, David C. Rankin < drankinatty@suddenlinkmail.com> wrote:
On Fri, 2009-07-10 at 19:26 -0500, David C. Rankin wrote:
Listmates,
I have more fun in store for this weekend. After testing dmraid-1.0.0rc15, I had downgraded a number of packages that were installed from testing back to there normal versions but had left dmraid-1.0.0rc15 and device mapper from testing. Everything was working fine.
Apparently some packages were moved from testing to extra or another normal repo because all of a sudden I began getting readline...so.6 error (from memory) on boot and was dumped into maintenance mode. In
After installing the latest readline, my box will not boot. I now
gives a readline...so.5 error and never gets to maintenance mode. My question is "What is the best way to try and recover?"
I have two options to work on the archlinux install:
(1) the machine is dual boot with openSuSE so I can boot to suse
and then mount the Archlinux filesystem (rw) under /mnt/arch. I'm not sure what I can do here unless there is a way for me to manually unpack some
On Friday 10 July 2009 08:41:36 pm Baho Utot wrote: maintenance mode, I remounted to root filesystem rw and then mounted all the partitions and ran pacman -Sy readline (testing is disabled) pacman packages and overwrite the problem files on the Arch install in this configuration; or
(2) boot using the Arch install disk. Here is where I'm a little
lost on the recovery procedure. I can manually assemble my raid array after booting to the install disk, but then what next? How would I go about reinstalling the various packages from either testing or extra when I have booted from the install disk?
Thus my need for help. What say the experts? How best to go about
fixing the device-mapper readline conflict?
I am using bash as my shell and I was bitten by this bed bug as well.
This is how I fixed it....
Fetch bash from an archlinux repos
put the bash package some place where you can get to it
boot from arch install/live cd
chroot to the broken system
pacman -U bash......
reboot
Baho, all:
Err.. My system is still broken and I can't figure out how to get around the readline error. I should have thought about this before going through the trouble of trying to set up the chroot to install bash, but the error I get when I try to execute chroot is thee *same* error I was getting when I tried to boot the system anyway. To try and recover, I booted from the install CD and then logged in as root. Here is how I set up to create the chroot environment and the error I received:
modprobe dm_mod modprobe sata_sil (or whatever chipset driver you need) dmraid -ay mkdir -p /mnt/rscu mount /dev/mapper/nvidia_ecaejfdi5 /mnt/rscu mount /dev/mapper/nvidia_ecaejfdi6 /mnt/rscu/boot mount /dev/mapper/nvidia_ecaejfdi7 /mnt/rscu/home mount /dev/mapper/nvidia_ecaejfdi9 /mnt/rscu/var mount /dev/mapper/nvidia_ecaejfdi10 /mnt/rscu/srv cd /mnt/rscu mount -o bind /dev dev mount -o bind /proc proc mount -o bind /sys sys chroot .
The error message:
/bin/bash: error while loading shared libraries: libreadline.so.5: cannot open shared object file: No such file or directory
How do I get around the libreadline error?
-- David C. Rankin, J.D.,P.E. Rankin Law Firm, PLLC 510 Ochiltree Street Nacogdoches, Texas 75961 Telephone: (936) 715-9333 Facsimile: (936) 715-9339 www.rankinlawfirm.com
On Saturday 11 July 2009 03:15:31 pm Ray Kohler wrote:
Download a copy of the new bash package by hand, and instead of the chroot command, do "pacman -r /mnt/rscu -U /path/to/bash_package".
HELL YES! Thanks Ray! All it took was: pacman -r /mnt/rscu -U /mnt/rscu/var/cache/pacman/pkg/bash-4.0.024-1-x86_64.pkg.tar.gz /mnt/rscu/var/cache/pacman/pkg/readline-6.0.00-1-x86_64.pkg.tar.gz chroot . pacman -U /var/cache/pacman/pkg/device-mapper-1.02.33-1-x86_64.pkg.tar.gz /var/cache/pacman/pkg/dmraid-1.0.0.rc15-6-x86_64.pkg.tar.gz reboot Once you got me over my mental block, recover took all of 30 seconds :p -- David C. Rankin, J.D.,P.E. Rankin Law Firm, PLLC 510 Ochiltree Street Nacogdoches, Texas 75961 Telephone: (936) 715-9333 Facsimile: (936) 715-9339 www.rankinlawfirm.com
On Sat, 2009-07-11 at 14:20 -0500, David C. Rankin wrote:
On Friday 10 July 2009 08:41:36 pm Baho Utot wrote:
On Fri, 2009-07-10 at 19:26 -0500, David C. Rankin wrote:
Listmates,
I have more fun in store for this weekend. After testing dmraid-1.0.0rc15, I had downgraded a number of packages that were installed from testing back to there normal versions but had left dmraid-1.0.0rc15 and device mapper from testing. Everything was working fine.
Apparently some packages were moved from testing to extra or another normal repo because all of a sudden I began getting readline...so.6 error (from memory) on boot and was dumped into maintenance mode. In maintenance mode, I remounted to root filesystem rw and then mounted all the partitions and ran pacman -Sy readline (testing is disabled)
After installing the latest readline, my box will not boot. I now gives a readline...so.5 error and never gets to maintenance mode. My question is "What is the best way to try and recover?"
I have two options to work on the archlinux install:
(1) the machine is dual boot with openSuSE so I can boot to suse and then mount the Archlinux filesystem (rw) under /mnt/arch. I'm not sure what I can do here unless there is a way for me to manually unpack some pacman packages and overwrite the problem files on the Arch install in this configuration; or
(2) boot using the Arch install disk. Here is where I'm a little lost on the recovery procedure. I can manually assemble my raid array after booting to the install disk, but then what next? How would I go about reinstalling the various packages from either testing or extra when I have booted from the install disk?
Thus my need for help. What say the experts? How best to go about fixing the device-mapper readline conflict?
I am using bash as my shell and I was bitten by this bed bug as well.
This is how I fixed it....
Fetch bash from an archlinux repos
put the bash package some place where you can get to it
boot from arch install/live cd
chroot to the broken system
pacman -U bash......
reboot
Baho, all:
Err.. My system is still broken and I can't figure out how to get around the readline error. I should have thought about this before going through the trouble of trying to set up the chroot to install bash, but the error I get when I try to execute chroot is thee *same* error I was getting when I tried to boot the system anyway. To try and recover, I booted from the install CD and then logged in as root. Here is how I set up to create the chroot environment and the error I received:
modprobe dm_mod modprobe sata_sil (or whatever chipset driver you need) dmraid -ay mkdir -p /mnt/rscu mount /dev/mapper/nvidia_ecaejfdi5 /mnt/rscu mount /dev/mapper/nvidia_ecaejfdi6 /mnt/rscu/boot mount /dev/mapper/nvidia_ecaejfdi7 /mnt/rscu/home mount /dev/mapper/nvidia_ecaejfdi9 /mnt/rscu/var mount /dev/mapper/nvidia_ecaejfdi10 /mnt/rscu/srv cd /mnt/rscu mount -o bind /dev dev mount -o bind /proc proc mount -o bind /sys sys chroot .
The error message:
/bin/bash: error while loading shared libraries: libreadline.so.5: cannot open shared object file: No such file or directory
How do I get around the libreadline error?
Ok We try again: We shall try to restore manually linreadline.so.5 so the system will boot. boot the live/install cd do a whereis libreadline.so.5 so we know we have the right one. If found mount the broken drive and copy to /lib like this: for example mount the broken system to /mnt then copy libreadline.so.5 to /mnt/lib as root or use sudo mount /dev/sdx /mnt copy /lib/libreadline.so.5 /lib/libreadline.so /mnt/lib reboot to see if that works What this does is to "install" libreadline.so.5 to the broken system If after booting and no errors on boot, then you can update to the "proper readline and bash by pacman -Syy; pacman -S readline bash As far as I know that will reinstall bash/readline to the current ones and you shouldn't have any more fun with those ;)
On Saturday 11 July 2009 06:43:51 pm Baho Utot wrote:
On Sat, 2009-07-11 at 14:20 -0500, David C. Rankin wrote:
On Friday 10 July 2009 08:41:36 pm Baho Utot wrote:
On Fri, 2009-07-10 at 19:26 -0500, David C. Rankin wrote:
Listmates,
I have more fun in store for this weekend. After testing dmraid-1.0.0rc15, I had downgraded a number of packages that were installed from testing back to there normal versions but had left dmraid-1.0.0rc15 and device mapper from testing. Everything was working fine.
Apparently some packages were moved from testing to extra or another normal repo because all of a sudden I began getting readline...so.6 error (from memory) on boot and was dumped into maintenance mode. In maintenance mode, I remounted to root filesystem rw and then mounted all the partitions and ran pacman -Sy readline (testing is disabled)
After installing the latest readline, my box will not boot. I now gives a readline...so.5 error and never gets to maintenance mode. My question is "What is the best way to try and recover?"
I have two options to work on the archlinux install:
(1) the machine is dual boot with openSuSE so I can boot to suse and then mount the Archlinux filesystem (rw) under /mnt/arch. I'm not sure what I can do here unless there is a way for me to manually unpack some pacman packages and overwrite the problem files on the Arch install in this configuration; or
(2) boot using the Arch install disk. Here is where I'm a little lost on the recovery procedure. I can manually assemble my raid array after booting to the install disk, but then what next? How would I go about reinstalling the various packages from either testing or extra when I have booted from the install disk?
Thus my need for help. What say the experts? How best to go about fixing the device-mapper readline conflict?
I am using bash as my shell and I was bitten by this bed bug as well.
This is how I fixed it....
Fetch bash from an archlinux repos
put the bash package some place where you can get to it
boot from arch install/live cd
chroot to the broken system
pacman -U bash......
reboot
Baho, all:
Err.. My system is still broken and I can't figure out how to get around the readline error. I should have thought about this before going through the trouble of trying to set up the chroot to install bash, but the error I get when I try to execute chroot is thee *same* error I was getting when I tried to boot the system anyway. To try and recover, I booted from the install CD and then logged in as root. Here is how I set up to create the chroot environment and the error I received:
modprobe dm_mod modprobe sata_sil (or whatever chipset driver you need) dmraid -ay mkdir -p /mnt/rscu mount /dev/mapper/nvidia_ecaejfdi5 /mnt/rscu mount /dev/mapper/nvidia_ecaejfdi6 /mnt/rscu/boot mount /dev/mapper/nvidia_ecaejfdi7 /mnt/rscu/home mount /dev/mapper/nvidia_ecaejfdi9 /mnt/rscu/var mount /dev/mapper/nvidia_ecaejfdi10 /mnt/rscu/srv cd /mnt/rscu mount -o bind /dev dev mount -o bind /proc proc mount -o bind /sys sys chroot .
The error message:
/bin/bash: error while loading shared libraries: libreadline.so.5: cannot open shared object file: No such file or directory
How do I get around the libreadline error?
Ok We try again:
We shall try to restore manually linreadline.so.5 so the system will boot.
boot the live/install cd
do a whereis libreadline.so.5 so we know we have the right one.
If found mount the broken drive and copy to /lib like this:
for example mount the broken system to /mnt then copy libreadline.so.5 to /mnt/lib
as root or use sudo
mount /dev/sdx /mnt copy /lib/libreadline.so.5 /lib/libreadline.so /mnt/lib
reboot to see if that works
What this does is to "install" libreadline.so.5 to the broken system
If after booting and no errors on boot, then you can update to the "proper readline and bash by pacman -Syy; pacman -S readline bash
As far as I know that will reinstall bash/readline to the current ones and you shouldn't have any more fun with those ;)
Thanks Baho, The problem I was having was not knowing how to do the manual install outside the chroot. I aleady had all of the current packages on my Arch system waiting to be installed, but I didn't know how to use the -r option to pacman. I do now ;-) From openSuSE, I had mounted the arch partitions and copied /var/cache/pacman/pkg over from my laptop after my latest update. So the '-r' option was the last missing piece of the puzzle. (I know, another "forrest for the trees issue") -- David C. Rankin, J.D.,P.E. Rankin Law Firm, PLLC 510 Ochiltree Street Nacogdoches, Texas 75961 Telephone: (936) 715-9333 Facsimile: (936) 715-9339 www.rankinlawfirm.com
participants (3)
-
Baho Utot
-
David C. Rankin
-
Ray Kohler