Hello all,
On Feb 29 my Fedora 16 box updated itself (I'm not wild about the only options being completely manual update or completely automatic update - but that's another story). It must have updated the Kernel because the next day when I turned on the PC I had a kernel panic (details below) and the machine failed to boot.
I had a zero timeout on grub so this meant that I had to use a Mint liveCD to edit grub so that I could try booting with the previous kernel.
The fact that I am able to write this shows that the previous kernel on the very same machine boots without problem. This is the "previous" working kernel: $ uname -a Linux localhost.localdomain 3.2.6-3.fc16.i686.PAE #1 SMP Mon Feb 13 20:44:23 UTC 2012 i686 i686 i386 GNU/Linux
The kernel that is failing is 3.2.7-1.fc16.i686.PAE
Below I have tried to type as accurately as possible the kernel panic message that I get when I try to boot into the new kernel (note It's typed manually - NOT copied and pasted - so i might have made the odd typo).
What should I do?
I can think of no obvious hardware problem given the fact this the machine boots happily with the previous kernel...
Thanks in advance...
Mark
==========8<======================================================= Booting 'Fedora (3.2.7-1.fc16.i686 PAE)'
Loading Fedora (3.2.7-1.fc16.i686 PAE) Loading initial ramdisk ... [ 2.957551] Kernel panic - not syncing: VFS: Unable to mount root fs on unknown-block (0,0) [ 2.957599] Pid: 1, comm: swapper/o Not tainted 3.2.7-1.fc16.i686.PAE #1 [ 2.957639] Call Trace: [ 2.957682] [<c0933468>] ? printk +0x2d/0x2f ... another 8 lines similar to this ... [ 2.957599] [<c0943efe>] kernel_thread_helper +0x6/0x10 ==========8<=======================================================
On 03/02/2012 12:40 PM, Arthur Dent wrote:
Hello all,
On Feb 29 my Fedora 16 box updated itself (I'm not wild about the only options being completely manual update or completely automatic update - but that's another story). It must have updated the Kernel because the next day when I turned on the PC I had a kernel panic (details below) and the machine failed to boot.
I had a zero timeout on grub so this meant that I had to use a Mint liveCD to edit grub so that I could try booting with the previous kernel.
The fact that I am able to write this shows that the previous kernel on the very same machine boots without problem. This is the "previous" working kernel: $ uname -a Linux localhost.localdomain 3.2.6-3.fc16.i686.PAE #1 SMP Mon Feb 13 20:44:23 UTC 2012 i686 i686 i386 GNU/Linux
The kernel that is failing is 3.2.7-1.fc16.i686.PAE
Below I have tried to type as accurately as possible the kernel panic message that I get when I try to boot into the new kernel (note It's typed manually - NOT copied and pasted - so i might have made the odd typo).
What should I do?
I can think of no obvious hardware problem given the fact this the machine boots happily with the previous kernel...
Thanks in advance...
Mark
==========8<======================================================= Booting 'Fedora (3.2.7-1.fc16.i686 PAE)'
Loading Fedora (3.2.7-1.fc16.i686 PAE) Loading initial ramdisk ... [ 2.957551] Kernel panic - not syncing: VFS: Unable to mount root fs on unknown-block (0,0) [ 2.957599] Pid: 1, comm: swapper/o Not tainted 3.2.7-1.fc16.i686.PAE #1 [ 2.957639] Call Trace: [ 2.957682] [<c0933468>] ? printk +0x2d/0x2f ... another 8 lines similar to this ... [ 2.957599] [<c0943efe>] kernel_thread_helper +0x6/0x10 ==========8<=======================================================
Looks like either dracut didn't put the driver you need for your hard drive in the ramdisk image or the grub configuration is telling it to use the wrong root partition. Check the grub config first and make sure the entry for the faulty kernel has the same root partition as the working kernel. ---------------------------------------------------------------------- - Rick Stevens, Systems Engineer, AllDigital ricks@alldigital.com - - AIM/Skype: therps2 ICQ: 22643734 Yahoo: origrps2 - - - - "I'd explain it to you, but your brain might explode." - ----------------------------------------------------------------------
On Fri, 2012-03-02 at 13:33 -0800, Rick Stevens wrote:
On 03/02/2012 12:40 PM, Arthur Dent wrote:
Hello all,
On Feb 29 my Fedora 16 box updated itself (I'm not wild about the only options being completely manual update or completely automatic update - but that's another story). It must have updated the Kernel because the next day when I turned on the PC I had a kernel panic (details below) and the machine failed to boot.
Looks like either dracut didn't put the driver you need for your hard drive in the ramdisk image or the grub configuration is telling it to use the wrong root partition. Check the grub config first and make sure the entry for the faulty kernel has the same root partition as the working kernel.
Ah yes! Thanks for pointing me in the right direction.
Looking at my /etc/grub2.cfg I noticed that the stanza for the most recent kernel was the same (except for kernel version) in every respect as its predecessor - EXCEPT that it lacked the line: initrd /boot/initramfs-3.2.7-1.fc16.i686.PAE.img
Looking in /boot I see that the following files are present: config-3.2.7-1.fc16.i686.PAE System.map-3.2.7-1.fc16.i686.PAE vmlinuz-3.2.7-1.fc16.i686.PAE
but NO initramfs-3.2.7-1.fc16.i686.PAE.img !
I have done a yum reinstall kernel-3.2.7-1.fc16.i686.PAE and I am now speaking to you from that new kernel!
It has however pointed up to me the fact that the last yum update did not complete properly and there are several (about 24) other errors. I will make that the subject of another post.
Thank you once again.
Mark
On 03/02/2012 03:35 PM, Arthur Dent wrote:
On Fri, 2012-03-02 at 13:33 -0800, Rick Stevens wrote:
On 03/02/2012 12:40 PM, Arthur Dent wrote:
Hello all,
On Feb 29 my Fedora 16 box updated itself (I'm not wild about the only options being completely manual update or completely automatic update - but that's another story). It must have updated the Kernel because the next day when I turned on the PC I had a kernel panic (details below) and the machine failed to boot.
Looks like either dracut didn't put the driver you need for your hard drive in the ramdisk image or the grub configuration is telling it to use the wrong root partition. Check the grub config first and make sure the entry for the faulty kernel has the same root partition as the working kernel.
Ah yes! Thanks for pointing me in the right direction.
Looking at my /etc/grub2.cfg I noticed that the stanza for the most recent kernel was the same (except for kernel version) in every respect as its predecessor - EXCEPT that it lacked the line: initrd /boot/initramfs-3.2.7-1.fc16.i686.PAE.img
Looking in /boot I see that the following files are present: config-3.2.7-1.fc16.i686.PAE System.map-3.2.7-1.fc16.i686.PAE vmlinuz-3.2.7-1.fc16.i686.PAE
but NO initramfs-3.2.7-1.fc16.i686.PAE.img !
I have done a yum reinstall kernel-3.2.7-1.fc16.i686.PAE and I am now speaking to you from that new kernel!
Great! Glad to hear it and very happy to have helped.
It has however pointed up to me the fact that the last yum update did not complete properly and there are several (about 24) other errors. I will make that the subject of another post.
Thank you once again.
Yes, one should always keep an eye on updates. I never permit automatic updates...I always want to watch it when it happens to catch glitches like that. Also, keep an eye on the /var/log/yum.log file--just in case! ---------------------------------------------------------------------- - Rick Stevens, Systems Engineer, AllDigital ricks@alldigital.com - - AIM/Skype: therps2 ICQ: 22643734 Yahoo: origrps2 - - - - Huked on foniks reely wurked for me! - ----------------------------------------------------------------------
Arthur Dent writes:
I had a zero timeout on grub so this meant that I had to use a Mint
A five second timeout makes very little difference, and will give you an option to recover from these kinds of situations.
Booting 'Fedora (3.2.7-1.fc16.i686 PAE)'
Loading Fedora (3.2.7-1.fc16.i686 PAE) Loading initial ramdisk ... [ 2.957551] Kernel panic - not syncing: VFS: Unable to mount root fs on unknown-block (0,0)
Looks like a busted initramfs to me. Something might've gotten broken during the automatic update.
First-ly, carefully compare the command lines in grub.cfg for both this kernel, and the one you can boot, to see if there are any differences.
Second-ly, if they're identical, rpm -e the bad one, then run yum update again to install it, and see what happens.
-
Arthur Dent -
Rick Stevens -
Sam Varshavchik