Ask Your Question

(Essex) Starting instances that were running on a failed node

asked 2013-07-06 19:24:47 -0500

Sam W. gravatar image

Hi All,

I have an old Essex cluster that just had a compute node fail. Is there a way to start the instances that were running on the failed node on a different host?

I have tried manually editing the database and changing "host" to the machine I want it to be running on and then rebooting the VM. When I go that I get:

2013-07-06 13:59:14 TRACE nova.rpc.amqp InstanceNotFound: Instance instance-00000053 could not be found.

Which I believe means the xml files in /etc/libvirt/qemu/ can not be found...

Is there any good way to have openstack recreate that xml file on a new host and boot the VM?

edit retag flag offensive close merge delete

1 answer

Sort by ยป oldest newest most voted

answered 2013-08-10 20:48:59 -0500

fifieldt gravatar image

The XML file is created from a template. If you copy one from an existing instance, you should be able to change the values inside it (ensuring they're unique) to recreate something similar enough to the original file that the instance will boot.

I'd also recommend upgrading from Essex if you can - Folsom and Grizzly both introduced features that made this kind of thing easier.

edit flag offensive delete link more

Get to know Ask OpenStack

Resources for moderators

Question Tools


Asked: 2013-07-06 19:24:47 -0500

Seen: 237 times

Last updated: Aug 10 '13