LinuxQuestions.org

LinuxQuestions.org (/questions/)
-   Linux - Software (https://www.linuxquestions.org/questions/linux-software-2/)
-   -   Kernel Upgrade issue (https://www.linuxquestions.org/questions/linux-software-2/kernel-upgrade-issue-763838/)

brentd 10-22-2009 06:43 PM

Kernel Upgrade issue
 
I have a system that is running Centos Linux 5.2 on a motherboard that uses an nvidia SATA controller. I am currently running kernel version 2.6.23.8. I've been trying to upgrade to 2.6.31.4 and am having problems booting.

I am getting a Kernel Panic that it can't find /dev/root. My root partition should be on /dev/VolGroup00/LogVol00.

I have double-checked the initrd files to make sure that all of the correct drivers are loading, and when I boot without the "quiet" kernel parameter I can see that sata_nv is being loaded and is seeing my hard drive, then I get a message that says something about "Module Initialization" then "Switching root" and then I get the kernel panic.

I've double-checked my kernel configs to make sure that LVM support is compiled into the new kernel, but I just can't seem to get this thing to boot. Any ideas?

lazlow 10-22-2009 09:08 PM

1st. Centos only supports the most current . version of each release. Currently (today) that is changing from 5.3 to 5.4.

2nd. if you are running a .23 kernel on Centos5.X you are not running a Centos kernel. Centos5.X runs a .18 based kernel(exclusively). Many people make the mistake that there will be more hardware support and security updates in a .23 kernel than in a current Centos5.X .18 kernel, this is a mistake. The current Centos 5.X kernel is 2.6.18-164. It has roughly the same hardware support as a .30 kernel(or newer). The reason Centos/RHEL stick with one kernel version(.18 in this case) for the entire life cycle of the release is to maximize compatibility.

Bring your system up to date (5.4/ -164 kernel) and your hardware will be supported and still maintain compatibility.

brentd 10-26-2009 10:48 AM

Quote:

Originally Posted by lazlow (Post 3729286)
1st. Centos only supports the most current . version of each release. Currently (today) that is changing from 5.3 to 5.4.

2nd. if you are running a .23 kernel on Centos5.X you are not running a Centos kernel. Centos5.X runs a .18 based kernel(exclusively). Many people make the mistake that there will be more hardware support and security updates in a .23 kernel than in a current Centos5.X .18 kernel, this is a mistake. The current Centos 5.X kernel is 2.6.18-164. It has roughly the same hardware support as a .30 kernel(or newer). The reason Centos/RHEL stick with one kernel version(.18 in this case) for the entire life cycle of the release is to maximize compatibility.

Bring your system up to date (5.4/ -164 kernel) and your hardware will be supported and still maintain compatibility.

I know I am not running a centos kernel. There is specific functionality that was added to the .28 source that I need (specifically the OSLEC code) and yes I know I could probably compile oslec without updating the kernel.

I'm having this same problem on a server where I need to update the kernel due to a bug in the raid card driver that causes a storm of error messages in the logfile every time I activate the the raid manager software.

While I appreciate your advice, I rellay need a solution to the problem at hand.

lazlow 10-26-2009 11:32 AM

Most of the hardware support in the .28 kernels is in the -164 Centos/RHEL kernel(backports). You would have to check on your specific code, but if it is not there odds are very high it is not there because it causes issues(ie it breaks something).

Edit: These guys look to be running the -128 kernel(post is from June of 08) and OSLEC.


All times are GMT -5. The time now is 08:09 AM.