pxc grow hangs if shrink was called before

Noticed that if you call cluster-shrink before you call cluster-grow there was an error and causes the cluster to hang up in a GROWING_CLUSTER state.

The issue is when looking for an existing instance in the cluster we are not ignoring the deleted instances.

2016-02-06 16:52:45.976 DEBUG oslo_messaging._drivers.amqpdriver [-] received message msg_id: None reply to None from (pid=22260) __call__ /usr/local/lib/python2.7/dist-packages/oslo_messaging/_drivers/amqpdriver.py:197
2016-02-06 16:52:45.992 DEBUG trove.instance.models [-] Instance 571a0fa1-fd0b-4af6-845e-7da3795e1eaf service status is new. from (pid=22260) load_instance /opt/stack/trove/trove/instance/models.py:488
2016-02-06 16:52:46.001 INFO trove.taskmanager.models [-] Creating instance 571a0fa1-fd0b-4af6-845e-7da3795e1eaf.
2016-02-06 16:52:46.146 DEBUG oslo_messaging._drivers.amqpdriver [-] received message msg_id: None reply to None from (pid=22260) __call__ /usr/local/lib/python2.7/dist-packages/oslo_messaging/_drivers/amqpdriver.py:197
2016-02-06 16:52:46.162 DEBUG trove.common.strategies.cluster.strategy [-] Loading class trove.common.strategies.cluster.experimental.pxc.taskmanager.PXCTaskManagerStrategy from (pid=22260) load_taskmanager_strategy /opt/stack/trove/trove/common/strategies/cluster/strategy.py:37
2016-02-06 16:52:46.175 DEBUG trove.common.strategies.cluster.experimental.pxc.taskmanager [-] Begin pxc grow_cluster for id: 7f4238ce-6147-484e-85c2-568b9c724db6. from (pid=22260) grow_cluster /opt/stack/trove/trove/common/strategies/cluster/experimental/pxc/taskmanager.py:155
2016-02-06 16:52:46.186 ERROR trove.common.strategies.cluster.experimental.pxc.taskmanager [-] Error growing cluster 7f4238ce-6147-484e-85c2-568b9c724db6.
2016-02-06 16:52:46.186 TRACE trove.common.strategies.cluster.experimental.pxc.taskmanager Traceback (most recent call last):
2016-02-06 16:52:46.186 TRACE trove.common.strategies.cluster.experimental.pxc.taskmanager File "/opt/stack/trove/trove/common/strategies/cluster/experimental/pxc/taskmanager.py", line 225, in grow_cluster
2016-02-06 16:52:46.186 TRACE trove.common.strategies.cluster.experimental.pxc.taskmanager _grow_cluster()
2016-02-06 16:52:46.186 TRACE trove.common.strategies.cluster.experimental.pxc.taskmanager File "/opt/stack/trove/trove/common/strategies/cluster/experimental/pxc/taskmanager.py", line 162, in _grow_cluster
2016-02-06 16:52:46.186 TRACE trove.common.strategies.cluster.experimental.pxc.taskmanager if db_inst.id not in new_instance_ids]
2016-02-06 16:52:46.186 TRACE trove.common.strategies.cluster.experimental.pxc.taskmanager File "/opt/stack/trove/trove/instance/models.py", line 647, in load
2016-02-06 16:52:46.186 TRACE trove.common.strategies.cluster.experimental.pxc.taskmanager return load_instance(cls, context, id, needs_server=True)
2016-02-06 16:52:46.186 TRACE trove.common.strategies.cluster.experimental.pxc.taskmanager File "/opt/stack/trove/trove/instance/models.py", line 466, in load_instance
2016-02-06 16:52:46.186 TRACE trove.common.strategies.cluster.experimental.pxc.taskmanager db_info = get_db_info(context, id, include_deleted=include_deleted)
2016-02-06 16:52:46.186 TRACE trove.common.strategies.cluster.experimental.pxc.taskmanager File "/opt/stack/trove/trove/instance/models.py", line 449, in get_db_info
2016-02-06 16:52:46.186 TRACE trove.common.strategies.cluster.experimental.pxc.taskmanager raise exception.NotFound(uuid=id)
2016-02-06 16:52:46.186 TRACE trove.common.strategies.cluster.experimental.pxc.taskmanager NotFound: Resource 580097ca-ec56-4fb5-aaaa-644fff79349b cannot be found.
2016-02-06 16:52:46.186 TRACE trove.common.strategies.cluster.experimental.pxc.taskmanager