c2ad4b28da
When only running volume migrations, a zone migration strategy audit without setting compute_nodes should work. Before this change, an audit with defined storage_pools, no compute_nodes parameters, and with_attached_volume is set to True would trigger the migration of the instances attached to the volumes being migrated. This patch decouples instance and volume migrations unless the user explicitely asks for both. When migrating attached volumes, the zone migration strategy will check for which instances should be migrated according to the audit parameters, and if the instance the volume is attached to can be migrated, it will be just after the volume. On the other hand, when the attached instances should not be migrated according to user input, only the volumes will be migrated. In an audit that migrates instnaces but not volumes, the with_attached_volume parameter will continue doing nothing. Closes-Bug: 2111429 Change-Id: If641af77ba368946398f9860c537a639d1053f69 Signed-off-by: jgilaber <jgilaber@redhat.com>
12 lines
492 B
YAML
12 lines
492 B
YAML
---
|
|
fixes:
|
|
- |
|
|
The zone migration strategy no longer fails when when an audit is created
|
|
with defined storage_pools, compute_nodes is not provided, and
|
|
with_attached_volume is set to True.
|
|
The strategy now creates the required volume migrations, but no instance
|
|
migrations. Both volumes and instances will only be migrated if the audit parameters
|
|
have both compute_nodes and storage_pools.
|
|
|
|
See: https://bugs.launchpad.net/watcher/+bug/2111429 for more details.
|