VM boot failure with kvm on Centos 7

Hey everyone!

I decided to come and seek help here as I’ve struggled with this for 2 days.

I use Centos 7.
When I try to create a VM it will fail the boot, because the kvm deployment seems to fail for some mysterious reasons.

Now the issues is that /usr/libexec/qemu-kvm exists and is totally fine

[oneadmin@localhost ]$ /usr/libexec/qemu-kvm --version
QEMU emulator version 1.5.3 (qemu-kvm-1.5.3-141.el7_4.6), Copyright © 2003-2008 Fabrice Bellard

I have copied the output of the logs down below.

I am also able to log into the host VM’s using oneadmin without a password and scp commands also work fine.

For that reason I have ran out of complete ideas, since the kvm also exists and seems to be working as stated above.

I would really appreciate help here! Could this be related to Centos 7 by any chance?


Versions of the related components and OS (frontend, hypervisors, VMs):

Steps to reproduce:

Current results:

When I try to create a Virtual Machine, the boot will fail. Looking at the opennebula logs, here’s what I get:

Fri Feb 2 04:55:46 2018 [Z0][VMM][D]: Message received: LOG I 47 Successfully execute transfer manager driver operation: tm_context.

Fri Feb 2 04:55:46 2018 [Z0][VMM][D]: Message received: LOG I 47 Successfully execute network driver operation: pre.

Fri Feb 2 04:55:47 2018 [Z0][VMM][D]: Message received: LOG I 47 Command execution fail: cat << EOT | /var/tmp/one/vmm/kvm/deploy ‘/var/lib/one//datastores/105/47/deployment.0’ ‘192.168.10.224’ 47 192.168.10.224

Fri Feb 2 04:55:47 2018 [Z0][VMM][D]: Message received: LOG I 47 error: Failed to create domain from /var/lib/one//datastores/105/47/deployment.0

Fri Feb 2 04:55:47 2018 [Z0][VMM][D]: Message received: LOG I 47 error: Cannot check QEMU binary /usr/libexec/qemu-kvm: No such file or directory

Fri Feb 2 04:55:47 2018 [Z0][VMM][D]: Message received: LOG E 47 Could not create domain from /var/lib/one//datastores/105/47/deployment.0

Fri Feb 2 04:55:47 2018 [Z0][VMM][D]: Message received: LOG I 47 ExitCode: 255

Fri Feb 2 04:55:47 2018 [Z0][VMM][D]: Message received: LOG I 47 Failed to execute virtualization driver operation: deploy.

Fri Feb 2 04:55:47 2018 [Z0][VMM][D]: Message received: DEPLOY FAILURE 47 Could not create domain from /var/lib/one//datastores/105/47/deployment.0

Is this a shared NFS datastore setup? Does the opennebula worker have access to the datastore?

On the worker host:

sudo virsh -c qemu:///system version --daemon

This thread may have some debugging tips https://bugs.launchpad.net/nova/+bug/1668164

Are you sure qemu kvm is installed on your node ??

Have solved the issue!

It was with a nfs datastore not being properly connected.

It was just difficult to figure it out.

Hi @margus, Please how did you solve this issue. Please explain. Thanks.