OpenStack instance fail to boot after unexpected compute node shutdown

Thomas Goirand zigo at
Thu Apr 25 21:58:42 UTC 2019

On 4/25/19 6:43 PM, Md. Farhad Hasan Khan wrote:
> Hi,
> We are running Openstack Rocky version. Our instances are getting volume
> from CEPH (mimic version).
> Suddenly one of our compute node got shutdown due to RAM issue. So all
> instances running on that compute node (linux/windows) got effected.
> After starting the compute node all services up & running. But instances
> on that compute node fail to boot. Instances are getting volume from
> CEPH storage& CEPH is in Healthy state. Please see below log.
> Requesting if anyone can help me to solve.


Make sure you have set in nova.conf:

on each of your compute nodes.

Also, it's wise to have instance_usage_audit=True

I hope this helps,

Thomas Goirand (zigo)

More information about the openstack-discuss mailing list