Originally Posted by theWeird
I had the same problem on my ISP-installations inside an OpenVZ container.
It is a bug in an older openvz-kernel on the host node which has been fixed some month ago. But because openvz is developed on a redhat kernel the patches need really long time to be ported into the official debian kernel.
But the people of the Proxmox project work much faster on that point than the debian devs.
I switched the kernel used on the host node to Proxmox (which is based on debian) and now all problems I had with openVZ disappeared.
Just follow the steps
'Adapt your sources.list' and 'Install Proxmox VE Kernel'
on this wiki page to make it work like a charm.
If you already had VEs configurated:
Proxmox moves the directory /etc/vz/conf to /etc/vz/conf.org and creates a symlink from /etc/vz/conf to /etc/pve/openvz.
So after installing Proxmox VE it seems that all your VEs are gone.
Just copy the config files from conf.org back to conf and restart openvz
# /etc/init.d/vz restart
Afterwards everything is fine.
My original post in the german forum:
I just tried switching to the Proxmox kernel, but the system won't boot anymore...
1) What could have been gone wrong?
2) I have access to a rescue system where I can mount all partitions e.g. /dev/md1 etc. - how can I revert to the old kernel so it can boot up again?
//edit: OK I was able to boot with the old kernel, but still it would be cool to know what went wrong
Would really appreciate your help