Hi all,
I have recently installed Proxmox v3.4 on a couple of DELL systems in my office and on another two into the Paris OVH's datacenter.
All the installations have been updated (no commercial licensing) just yesterday with the latest community kernel.
The issue is that if I try to shutdown or reboot any node (from bash or also from the web admin interface) the various VMs, that are all Linux (CentOS, Debian, OpenSUSE, etc.), begin to be destroyed and the system stops in few seconds. In other words, Proxmox doesn't perform a graceful shutdown of the various guests.
This is terrible! In particular if you have guests with running databases, you will corrupt them all at 100%!
N.B.: This happen on all PVE servers also before that I have updated all of them.
The system seem to feel correctly the shutdown ("Stop all VMs and Containers" is showed inside the logs), but then it doesn't show the shutdown of etch single VMs ("VMXXX - Shutdown" messages didn't show/exist), because it didn't do it!
Have you any advice? Do you know if this is a bug?
Thanks,
Andy
I have recently installed Proxmox v3.4 on a couple of DELL systems in my office and on another two into the Paris OVH's datacenter.
All the installations have been updated (no commercial licensing) just yesterday with the latest community kernel.
The issue is that if I try to shutdown or reboot any node (from bash or also from the web admin interface) the various VMs, that are all Linux (CentOS, Debian, OpenSUSE, etc.), begin to be destroyed and the system stops in few seconds. In other words, Proxmox doesn't perform a graceful shutdown of the various guests.
This is terrible! In particular if you have guests with running databases, you will corrupt them all at 100%!
N.B.: This happen on all PVE servers also before that I have updated all of them.
The system seem to feel correctly the shutdown ("Stop all VMs and Containers" is showed inside the logs), but then it doesn't show the shutdown of etch single VMs ("VMXXX - Shutdown" messages didn't show/exist), because it didn't do it!
Have you any advice? Do you know if this is a bug?
Thanks,
Andy