Live-migration leaves instance in MIGRATING state if scheduler fails to find a host
Bug #1166771 reported by
Hans Lindgren
This bug affects 1 person
Affects | Status | Importance | Assigned to | Milestone | |
---|---|---|---|---|---|
OpenStack Compute (nova) |
Fix Released
|
High
|
Hans Lindgren | ||
Grizzly |
Fix Released
|
High
|
Unassigned |
Bug Description
This can be easily reproduced on a single devstack install:
1. Boot an instance and wait for it to become active
2. Issue live-migration command without setting destination host to trigger scheduler host selection
Since there is just one compute node in this setup (the source node), scheduling will return a NoValidHost exception.
Expected behavior is for the instance task state to be reset to None, while the actual behavior is that it is left as MIGRATING.
The solution is to add this exception to the list of exceptions for which the instance state is reset on a failure.
tags: | added: grizzly-backport-potential |
tags: | removed: grizzly-backport-potential |
Changed in nova: | |
milestone: | none → havana-1 |
status: | Fix Committed → Fix Released |
Changed in nova: | |
milestone: | havana-1 → 2013.2 |
To post a comment you must log in.
Fix proposed to branch: master /review. openstack. org/26475
Review: https:/