Open Stack

All;

What is the most effective means of having the OpenStack cluster restart guests when a hypervisor host fails?  We're running OpenStack Victoria, installed manually through packages.

My apologies, but my Google foo fails me on this issue; I don't know how to ask it the question.

I recognize that OpenStack covers a great many different deployment scenarios, and in many of these this isn't feasible.  In our case, images, volumes, and ephemeral storage are all on our Ceph cluster, so all storage is always available to all hypervisor hosts.

I also recognize that resource restrictions mean that even in an environment such as mine, not all failed guests may be able to be restarted on new hosts.  I'm ok with a dumb best effort, at least for now.

Is there something already present in OpenStack which would allow this?

Thank you,

Dominic L. Hilsbos, MBA 
Vice President - Information Technology 
Perform Air International Inc.
DHilsbos at PerformAir.com 
www.PerformAir.com