Request #2328
Fail over VM after host failure (non persistent image)
| Status: | Closed | Start date: | 09/18/2013 | |
|---|---|---|---|---|
| Priority: | Normal | Due date: | ||
| Assignee: | - | % Done: | 0% | |
| Category: | - | |||
| Target version: | - | |||
| Pull request: |
Description
According to documentation (http://opennebula.org/documentation:rel4.2:ftguide), when there is a host failure, the only way to restart a VM on another host is to recreate it.
This is OK when using persistent images but in our case, we are using non persistent images on a shared filesystem (glusterFS)
We don't have found yet any way to enforce the migration of the VM to another host when its status is UNKNOWN
Is there a command to failover a VM created from a non persistent image when its status is UNKNOWN ? Otherwise, it will be nice to add a new option to the boot command to specify on which host the VM must be restarted.
History
#1
Updated by Ruben S. Montero almost 8 years ago
- Status changed from Pending to Closed
We are implementing in #2135 the option to shutdown, shutdown --hard, poweroff and undeploy a VM in UNKNOWN state.
In your case you may undeploy the VM in unknown and deploy it again in other host.
Closing this one, thanks for the feedback.