[elrepo] Kernel panic using recent kernels

[Miguel Alvarez]

On Mon, Feb 28, 2011 at 4:31 PM, Alan Bartlett <ajb at elrepo.org> wrote:
>
> On 28 February 2011 21:11, Miguel Alvarez <miguellvrz9 at gmail.com> wrote:
> > On Mon, Feb 28, 2011 at 12:06 PM, Alan Bartlett <ajb at elrepo.org> wrote:
> >> On 28 February 2011 18:26, Miguel Alvarez <miguellvrz9 at gmail.com> wrote:
>
> >> > dm-region_hash.ko
> >> > dm-mem-cache.ko
> >> > dm-message.ko
> >> > dm-raid45.ko
>
> >> Please boot up the current distro kernel (-194.32.1.el5, if CentOS, or
> >> -238.1.1.el5, if RHEL) and go to the /etc/sysconfig/mkinitrd/
> >> directory. Do you have a file name "dmraid"? If not, create one with
> >> one line contents:
> >>
> >> DMRAID=no
> >>
> >> The file should be UID/GID == root, with mode == 755.
> >>
> >> Now go to the /boot directory and re-make the initrd for the
> >> 2.6.37-1.el5.elrepo kernel. If I've remembered the details correctly,
> >> that should solve the problem.
> >>
> >> It might also be worthwhile adding:
> >>
> >> nodmraid
> >>
> >> to the kernel boot parameters. (Via the line in your /etc/grub.conf file.)
>
> > Thanks for the reply.  I created the file, ensured perms/ownership,
> > re-created the ramdisk and added the 'nodmraid' boot option.  Unfortunately,
> > I'm still panicing.    It looks like it's an issue detecting lvm volumes
> > (see attached screenshots).
> >
> > When 2.6.18 boots, it scans for logical volumes and finds the "VolGroup00"
> > VG.  When 2.6.37 boots, it scans but is unable to find anything and
> > naturally panic shortly thereafter.  Any ideas?
>
> Miguel,
>
> I've looked again at your initial message and I am somewhat surprised
> that the (2.6.37) system panics over the dm-region_hash.ko versus
> dm-region-hash.ko naming . . . I see a brief note is present on the
> kernel-ml page [1].
>
> Putting that to one side, as the boot process now proceeds beyond that
> point, I agree that the latest panic is down to the kernel's inability
> to find "VolGroup00". That has me completely puzzled.
>
> As I mentioned earlier, I currently don't have access to my EL5 system
> which runs the 2.6.35-11.el5.elrepo kernel. What I can tell you is,
> previously, that system used to use LVM to join together two small
> disks to make one larger logical device. The 2.6.37-1.el5.elrepo
> kernel was tested on it before those two disks were removed and
> replaced by one device. When the system was rebuilt, it was decided
> not to use LVM. So although I could load the current
> 2.6.37-2.el5.elrepo kernel and recheck the boot process, I doubt we
> would learn anything of significance as it does not use LVM.
>
> What might help is sight of your /etc/fstab file and the output
> produced by /sbin/fdisk -l
>
> [quote]
> HP DL-380 G7 (also G6 with same results)
> [/quote]
>
> I'm not familiar with your hardware, above, so I wonder if it could be
> a configuration quirk? Could it be the way that the integrated P410i
> smart array controller is configured?
>
> Perhaps other subscribers to this m/l would like to comment? If we
> fail to find the solution it might be worth your while taking this
> query to the CentOS fora [2], as I know that some of the regular
> people there are familiar with HP ProLiant server hardware.

Hi Alan,

Is there an archive of the older kernel-ml RPMs available?  I could
have sworn an older version of kernel-ml worked on my G6 but of course
I can't remember which version it was.

But I agree, very strange indeed but I can verify I'm seeing the same
behavior on both my G6 and G7.  There's nothing special about the fs
layout -- I always just accept the recommended default configs during
the install process:

---/etc/fstab---
/dev/VolGroup00/LogVol00 /                       ext3    defaults        1 1
LABEL=/boot             /boot                   ext3    defaults        1 2
tmpfs                   /dev/shm                tmpfs   defaults        0 0
devpts                  /dev/pts                devpts  gid=5,mode=620  0 0
sysfs                   /sys                    sysfs   defaults        0 0
proc                    /proc                   proc    defaults        0 0
/dev/VolGroup00/LogVol01 swap                    swap    defaults        0 0

---fdisk -l ---
Disk /dev/cciss/c0d0: 146.7 GB, 146778685440 bytes
255 heads, 63 sectors/track, 17844 cylinders
Units = cylinders of 16065 * 512 = 8225280 bytes

           Device Boot      Start         End      Blocks   Id  System
/dev/cciss/c0d0p1   *           1          13      104391   83  Linux
/dev/cciss/c0d0p2              14       17844   143227507+  8e  Linux LVM

I'll keep working on things on my end, but if you could point me to
the archive (presuming there is one), that would be greatly
appreciated.

Thanks again