Hi All, I have been strugalling for the past few days with out any answer!!!!!!!!

I am using CentOS release 6.8 (Final) with KVM and this is a compute in a cloudestack installation.

Now the issue is, I hade to reset the machine due to RAM issue and when finally the node came up, I am missing one VM. Through cloudstack I can see the VM but when I do "virsh list --all" I only see 2 where there should be 3. The disks attached to this VM is visible though ls.

Can some one help me to get this reattached (if that is the term to use ) KVM or to diagnose what is going on

Thanks and Regards

Try
This just shows the storage, but its a start

Hi Chrism01, thank you for the reply, when I do below command I get an error. please see below

[root@kvm01 primary]# virsh vol-list --details --all
error: command 'vol-list' doesn't support option --all


Thanks and Bet Regards

Sorry about that

I don't pretend to know cloudstack, but if it is any similar to OpenStack, it could well be that the VM persists in the cloudstack database but has disappeared from the VM host. Perhaps cloudstack has taken a snapshot of it and removed it?

Does cloudstack allow you to see the status of the VM? What is it?
You could try some actions on the VM via cloudstack. For example, start or restart or stop or suspend (or whatever similar actions exist in cloudstack). What happens when you do that?

Hi Bernbdaush thanks for the update

The status on the UI says it is stopped but when I try to start it says the action is completed but the status remains the same. If I to see from the compute node there is no entry for this VM or it is running. I can see the vm disks where were attached to this lost instance, that's about it.

It feels the same but setting cloudstack up is much, much easer then openstack

By the way this is just a POC environment with once control node and a one worker node

Thanks and Regards

I need to guess, but it's likely that the VM doesn't exist anymore, and what you see in cloudstack is its ghost. It disappeared but cloudstack didn't notice.
I suggest you find a way to remove it from your cloud.

Hi Berndbausch, I can remove this from the db (mysql), but it would be nice to find out what happened since its rather scary the situation. Just reboot can cause such a havoc, think it was on a production .

But I can see the v-disks from the OS side, do you think this has something to do with KVM it self since verish list --all does not show this vm but rest are fine and working. Is there any chance that I could import this disk to a KVM instance with out the aid of cloudstack.

Thanks and Best regards

Yes you can. You create a new VM and specify the existing disk as root disk (and add any other existing disks). The easiest way is via the GUI, virt-manager. If the compute node has desktop software installed, you may have virt-manager installed already. If you can't add desktop software to the compute node, you should be able to run virt-manager somewhere else and remote-control the compute node via a libvirt connection. From memory I don't know how this is done.

If GUI is not an option, I am aware of two ways to create a VM via command line. One is virt-install, providing all the parameters like network, console, RAM, and storage, as command line options.
The other is copying the XML that describes another VM and adapt it. Roughly:
libvirt.org has useful documentation for virsh and the XML file.

Hi Bernbdaush thank you for the reply and I had to recreate the VM since this was to be a poc but I do have a backup of the issued VM, let me try and import the vmdisk at he vm create time

Thanks you and best regards