[HEAT] Sahara fails to rollback if stack in 'Failed' state

[Andrew Lazarev]

Step to repro:
1. Create cluster
2. Create nodegroup with volume exceeding quota
3. Scale cluster adding nodegroup created in step 2

Expected result:
Sahara fails to scale and rollback cluster

Observed result:
Cluster in Error state

Probably problem related to <"disable_rollback": true> parameter in heat request.

[OpenStack Infra (hudson-openstack)]

Fix proposed to branch: master
Review: https://review.openstack.org/105321

[OpenStack Infra (hudson-openstack)]

Reviewed: https://review.openstack.org/105321
Committed: https://git.openstack.org/cgit/openstack/sahara/commit/?id=11d283011749406d15751a55e34969294dd8060c
Submitter: Jenkins
Branch: master

commit 11d283011749406d15751a55e34969294dd8060c
Author: Andrew Lazarev <email address hidden>
Date: Mon Jul 7 18:35:49 2014 -0700

    Fixed cluster rollback on scaling with heat engine

    We can't allow stack to move into Error state if we want to use it futher.
    disable_rollback parameter controlls such behavior. It is Ok to have errored
    stack for cluster create and rollback since we will not be using it anyway. But
    for scaling we need to keep it Active since we can just rollback in case of
    error.

    Change-Id: Iab1a58f4924ca6a7689060f71217956a72316517
    Closes-Bug: #1337006