Successful P2V Conversion
I use the VMware vCenter Converter Standalone frequently to migrate physical machines to my VMware cloud. One error that I get frequently after a successful conversion of CentOS 5 system is:
Kernel Panic Using VMware vCenter Converter
Kernel Panic after using VMware vCenter Converter Standalone
Without fail, this error has always occurred when I have attempted to convert a CentOS version 5 physical box using the VMware vCenter Converter Standalone that has been using logical volumes. To get the destination virtual machine to work, you will have to configure the VMware converter to migrate these logical volumes to physical disk devices ( A bit ironic no? Going P2V and changing the disk layout from V2P!).
Determining if you are using logical Volumes
Most linux systems are by default using logical volumes. You can tell by simply doing a quick df on the system:
| df -h //dev/mapper/VolGroup00-LogVol00 29G 14G 13G 52% / |
In the example above you see the device that the “/” filesystem is mounted on is a logical device.
Problem Resolution
The key to resolving this problem is to go to the “Destination Layout” tab of the “Data to Convert” settings. You will have to click the blue “Advanced” link to get to this specific view that you see below:
- Changing the disk partitions from logical volumes to physical disk references in the VMware Converter.
Define the VMware Disk Layout
Be certain to change the Type to “Thin” and then use the “To basic” button. With these settings in place, your logical volumes will be converted to what the Operating System will consider logical partitions although they will truly be VMware logical volumes.
Rerun the P2V Using the VMware vCenter Converter Standalone
I then re-run the conversion using the VMware vCenter Converter Standalone again. This has always resolved the above problem for me.
Other VMware vCenter Converter Issues
I have also found issues from time to time when the version of the physical CentOS 5 box had not been updated in quite some time. The last time I had such a problem, my CentOS version was 5.7:
| # cat /etc/redhat-release CentOS release 5.7 (Final) |
Update CentOS Prior to Conversion
It is always a good practice to update your CentOS prior to conversion:
| # yum update Loaded plugins: downloadonly, fastestmirror Loading mirror speeds from cached hostfile * addons: mirror.anl.gov * base: mirrors.tummy.com * epel: mirrors.tummy.com * extras: mirrors.usinternet.com * updates: centos.mirror.ndchost.com Setting up Update Process……… |
Let the update run to completion. It should get you to the latest, greatest version of CentOS version 5:
| # cat /etc/redhat-release CentOS release 5.9 (Final) |
Latest posts by Jeff Staten (see all)
- Configure Your HP Procurve Switch with SNTP - May 5, 2015
- Configuring HP Procurve 2920 Switches - May 1, 2015
- Troubleshooting Sendmail - November 28, 2014
Hi Jeff,
I have the same issue when try to convert Physical machine with CentOS release 5.5 installed on it, to VM for ESXi 5.1.
While power up new VM I see fails in mounting and kernel panic, the exact same as in this article.
df -h command on my CentOS machine provides following info:
/dev/mapper/VolGroup00-LogVol00 221G 27G 183G 13% /
/dev/sda1 99M 12M 82M 13% /boot
tmpfs 2.0G 0 2.0G 0% /dev/shm
Tried to change the Type to “Thin” and convert again. the same fault result with kernel panic.
What can I try to resolve the issue?
Best Regards,
Oleg Olshanechky
Oleg,
The important part of the solution is not to use thin partitions, it is to convert from LVM to basic using the “To Basic” button. The VMWare converter doesn’t work well with CentOS LVM. I have also found that by updating the latest releases of 5 it can help too.
I had the exact same error/kernel panic and this was the exact fix. Thanks for putting this out there, Jeff!
It worked for me!! Thanks so much!
What is working 100% all time :
boot on RH CD : linux rescue
then :
chroot /mnt/sysimage
next rebuild swap partition :
mkswap /dev/sysVG/swap
and rebuild initrd image :
cd /boot
mv initrd*.img initrd…..img.bck
/sbin/mkinitrd -v /boot/initrd…..img `uname -r`
exit
reboot
and that’s ok !