Proxmox V4 bugs

[newacecorp]

There is a new feature available to do that for you. You could also use simple shell commands do migrate all running VMs via qm migrate. That's what I've done in the past.

Yes, that's a workaround that is usable. However, if I already manage the VMs in the web-interface, I expect to have the capability there.

 

[ffuentes]

Yes, that's a workaround that is usable. However, if I already manage the VMs in the web-interface, I expect to have the capability there.

I think there is.
Under the GUI go to the desire host and on the top right under "More" select "Migrate All VMs and Containers"

 

[newacecorp]

I think there is.
Under the GUI go to the desire host and on the top right under "More" select "Migrate All VMs and Containers"

And so there is! Good find. Thanks for the heads up.

 

[LnxBil]

As I mentioned previously, if the qemu process dies or the OS itself crashes, I want to have it restarted.

Did you ever encounter a qemu process crash? I haven't, but yes, in that case the VM is down should be started automatically.

For your requirement, that a crash of the OS itself is somehow detected is difficult. If the OS reboots itself, you have nothing to do, but if it gets stuck, that's more problematic. I don't know if other hypervisors do this (via some heartbeat over the agent or something), but you can - in my experience - only fix this with a watchdog timer. In Linux, you can simple add the watchdog device from KVM to your machine and monitor the OS within. That works very well, but for Windows, there are no 64-bit drivers available (at least to my knowledge).

 

[newacecorp]

Did you ever encounter a qemu process crash? I haven't, but yes, in that case the VM is down should be started automatically.

For your requirement, that a crash of the OS itself is somehow detected is difficult. If the OS reboots itself, you have nothing to do, but if it gets stuck, that's more problematic. I don't know if other hypervisors do this (via some heartbeat over the agent or something), but you can - in my experience - only fix this with a watchdog timer. In Linux, you can simple add the watchdog device from KVM to your machine and monitor the OS within. That works very well, but for Windows, there are no 64-bit drivers available (at least to my knowledge).

Actually, this is done with the QEMU-GUEST-AGENT in the VM so that it can send a heartbeat back to the QEMU process and reboot it if it stops responding.

 

[newacecorp]

Did you ever encounter a qemu process crash? I haven't, but yes, in that case the VM is down should be started automatically.

For your requirement, that a crash of the OS itself is somehow detected is difficult. If the OS reboots itself, you have nothing to do, but if it gets stuck, that's more problematic. I don't know if other hypervisors do this (via some heartbeat over the agent or something), but you can - in my experience - only fix this with a watchdog timer. In Linux, you can simple add the watchdog device from KVM to your machine and monitor the OS within. That works very well, but for Windows, there are no 64-bit drivers available (at least to my knowledge).

I run primarily CentOS/RH installations, so all I do is "yum install qemu-guest-agent". You have to check the Proxmox options for the VM to enable the Agent otherwise qemu-guest-agent will exit shortly after boot up of the VM.