Proposed EC2 OpenStack extension
'PAUSED' OS instance status is not mappable to any EC2 status correctly.
Adding a 'paused' status seams to the only viable way to handle this.
Normally you cannot reach this status just with the EC2 API, but all other
defined EC2 statuses are schematically incorrect for status mapping.
A suspend can be considered as a 'stopped' status.
'rebooting' can be considered as 'pending' status.
Using the 'paused' status is only valid when the VM paused because
of the user's explicit request, so it is not temporary pause because
of an internal operation sequence. If the nova just pauses the VM
just temporary it must report 'pending' (or 'running' ) status.
Change-Id: Id4500dd6e38e078d0b0b76fb0dd6a81c054ef646
diff --git a/tempest/testboto.py b/tempest/testboto.py
index 7031fe2..ab79f3c 100644
--- a/tempest/testboto.py
+++ b/tempest/testboto.py
@@ -202,8 +202,10 @@
s3_error_code.server = ServerError()
s3_error_code.client = ClientError()
valid_image_state = set(('available', 'pending', 'failed'))
+ #NOTE(afazekas): 'paused' is not valid status in EC2, but it does not have
+ # a good mapping, because it uses memory, but not really a running machine
valid_instance_state = set(('pending', 'running', 'shutting-down',
- 'terminated', 'stopping', 'stopped'))
+ 'terminated', 'stopping', 'stopped', 'paused'))
valid_volume_status = set(('creating', 'available', 'in-use',
'deleting', 'deleted', 'error'))
valid_snapshot_status = set(('pending', 'completed', 'error'))