0

I'm trying to import an existing CentOS VM from a VMware environment to run on Digitalocean vis their custom images option. I'm able to successfully run the VMDK in both VMware Workstation and VirtualBox. I've tried using the VMDK itself, and cloning the VMDK from VirtualBox to a VDI format, and uploading these to DigitalOcean and restoring. The restoration succeeds, but during boot, the below error occurs.:

root (hd0,0)
 Filesystem type is ext2fs, partition type 0x83
kernel /vmlinuz-2.6.18.164.11.1.e15 ro root=/dev/VolGroup00/LogVol00 rhgb quiet
   [Linux-bzImage, setup=0x1e00, size=0x1d6b1c]
initrd /initrd-2.6.18-164.11.1.e15.img
   [Linux-initrd @ 0x37cab000, 0x344acb bytes]

Memory for crash kernel (0x0 to 0x0) notwithin permissible range
WARNING calibrate_APIC_clock: the APIC timer calibration may be wrong.
PCI: PIIX3: Enabling Passive Release on 000:00:01.0
Red Hat nash version 5.1.19.6 starting
  Reading all physical volumes.  This may take a while...
  Volume group "VolGroup00" not found
Unable to access resume device (/dev/VolGroup/LogVol01)
mount: could not find filesystem '/dev/root'
setuproot: moving /dev failed: No such file or directory
setuproot: error mounting /proc: No suck file or directory
setuproot: error mounting /sys: No suck file or directory
switchroot: mount failed: No such file or directory
Kernel panic - not syncing: Attempt to kill init!

Kernel alive
kernel direct mapping tables up to 120000000 @ 10000-16000

The only changes the VM went through between working on a desktop hypervisor and booting in DigitalOcean was the compressing of the VMDK/VDI file to gzip to make the upload more efficient. Would this be the cause, or is it more likely that the infrastructure differences between a desktop hypervisor and DigitalOcean's infrastructure was enough to shake-up the volume names?

Any suggestions would be much appreciated.

simonlehmann
  • 350
  • 1
  • 4
  • 15
  • CentOS 5 is past end of life. The amount of help you can get is therefore pretty limited. Whatever you're running should have been moved off it several years ago. That said, this looks like a driver issue. DO requires virtio drivers in the guest, and your initrd is unlikely to have it since (1) it's so old and (2) you didn't create the VM on KVM. – Michael Hampton Feb 25 '19 at 04:33
  • Thanks for your response. I contacted DigitalOcean support and they informed the image wouldn’t be supported. I was able to roll it out in AWS EC2. Thanks again. – simonlehmann Feb 26 '19 at 05:03
  • Not surprising. I was one of the earliest DO users, and I don't recall that they ever supported CentOS 5. – Michael Hampton Feb 26 '19 at 05:07
  • It wasn't so much CentOS 5 that they said wouldn't be supported (though that on its own makes sense too), it was more that the filesystem was partitioned with ext2 logical volumes rather than ext3 or ext4 direct partitioning. The AWS EC2 VM import worked very smoothly and there was good documentation on achieving an import. Thanks for your feedback. – simonlehmann Feb 26 '19 at 07:35

0 Answers0