virtualbox.org

I am experiencing the 'kernel panic CentOS 6.2 guest - init_tsc_clocksource' as identified in a previous thread - last updated 4, May 2012 (sorry too new to post urls)

Bugtracker ticket #10427 identifies a fix r41256 - Changeset 41256 in vbox

The Guest is Centos 6.2 built on VirtualBox 3.2.10 and updated the system via yum. After a power outage I attempted to restart the Centos 6.2 guest and experienced the kernel panic. After some searching, I found the (not recommended) work around to change the chipset. I decided to update VirtualBox to the current release - 4.1.18 r78361 as what I was running did not support the chipset change. Thought it was worth a try. Then I found the resolution, r41256, and that page references VurtualBox 4.1.12.

I am wondering if the fix was applied to 4.1.12 or if 4.1.12 was the release used when the fix was found... If I am running 4.1.18, would the fix have made it into this release...

More importantly, I'm lost. I have no idea what to do next to resolve the issue. Hard to admit, but yep, I don't know what I am doing. I hate it when I get stupid, but sometimes it happens.

The host is Windows XP, 32 bit. VirtualBox is at 4.1.18 r78361. The guest is Centos 6.2, also 32 bit, and reports the kernel to be 2.6.32-220.23.1.el6.i686. Guest additions were not installed yet.

Can someone give me a nudge in the right direction?

Thank you.

Try adding divider=100 to the boot line.
Also Post the guests log file (as an attachment)
See Online Users Manual for the location of the file.

Thank you for your response!

Log file is attached.

I booted from DVD 1 into rescue mode and edited grub.conf to add divider=100 the end of the first kernel entry.

Kernel panic still occurred.

Attached log file is from the failed boot attempt.

Thank you.

Well You could have done this from the grub boot screen to test with instead of make it permanent. Now I don't know what you actually edited. Can you post the contents of the grub file?

I also see you are running on very little memory and the amount you are assigning to the guest is putting you at the border of running out. Set the Cent guest to 512 and see what happens.

Sorry, forgot I could modify boot parms on the fly, haven't done it for quite a while. Yes, it would have been easier.

The grub.conf I used is attached.

The file was edited with vi - copied the first kernel section with yy4, cursored down to the bottom line and pasted the copy with a p. I then modified the first section to add 'divider=100' to title and the end of the kernel line, saved the file, exited the rescue mode and attempted to boot. As an aside, I have a copy of the original grub.conf as a backup from when I made the changes to allow for a more verbose boot message list.

Your recommendation regarding the memory has me confused. The memory allocated to this machine should be 750 MB.
I thought it would be sufficient to get the machine up and running. I changed the memory to 512 MB per your recommendation and attempted to boot. The vbox log file is also attached.

I did see a post somewhere that indicated there is / was a problem attempting to boot with the Centos DVD mounted, even if it was not the boot device. All boot attempts are made with no media mounted in the cd/dvd drives.

Thank you.

I told you wrong the other day. it should be divider=10
See if that helps.

I edited grub.conf to change divider=100 to divider=10. I also changed the memory allocated to this machine back to 750 MB. The system experienced a Kernel panic on boot. Vbox log is attached.

Thank you.

The log files are pointless in this case since it is a guest kernel panic and will not show in the log file.

It looks like your best option is to restore from backup. Pulling the cord (so to say) must have corrupted the guest. You can try to recover this using the guests recovery methods but that is not something that we have control over and would fall outside the scope of this forum.

If you do not have a backup and you have data on the guest that you need, you should be able to install a new guest and attach the broken one as a second drive to be able to retrieve the data.

I was in the process of building this machine when the outage occurred. While it is a bit of a setback, 'it's only a flesh wound'. The OS had been built and updated, and I was in the process of downloading apps (nagios and such). Fortunately, I had only started to build and install the apps and it was not fully configured yet.

On the bright side, I learned and was reminded of a few things, and got my Vbox updated...

Thank you again for your assistance!

No joy here in Mudville... Recreated the machine, copies vdi file as a backup, and rebooted several times. all went well. Ran yum update, updates completed, shutdown. Attempted to boot and process failed, same error. Retried with divider=10 appended, this time added during via grub and process failed again.

log attached for boot process without divider=10.

Set the amount of RAM to the guest at no more than 512MB. As I said before you are running very low on available memory. Other than that I can not tell you what is causing the kernel panic. It is not showing in VirtualBox because it is a guest OS error.