Timeout trying to delete overcloud stack in CI

Seen here:

http://logs.openstack.org/13/260413/6/check-tripleo/gate-tripleo-ci-f22-ha/fe2b77a/console.html

2016-01-14 14:23:58.898 | #################
2016-01-14 14:23:58.898 | tripleo.sh -- Overcloud delete
2016-01-14 14:23:58.898 | #################
2016-01-14 14:24:02.016 | +--------------------------------------+------------+-----------------+---------------------+--------------+
2016-01-14 14:24:02.016 | | id | stack_name | stack_status | creation_time | updated_time |
2016-01-14 14:24:02.017 | +--------------------------------------+------------+-----------------+---------------------+--------------+
2016-01-14 14:24:02.017 | | 9a8a8a20-da8e-4de7-b9af-71e48e567aa5 | overcloud | CREATE_COMPLETE | 2016-01-14T13:17:29 | None |
2016-01-14 14:24:02.017 | +--------------------------------------+------------+-----------------+---------------------+--------------+
2016-01-14 14:29:19.168 | #################
2016-01-14 14:29:19.168 | tripleo.sh -- Overcloud 9a8a8a20-da8e-4de7-b9af-71e48e567aa5 delete failed or timed out:
2016-01-14 14:29:19.168 | #################
2016-01-14 14:29:19.168 | /tmp/tripleo.sh: line 447: Timing: command not found

The "Timing: command not found" is odd, it's not coming from tripleo.sh, but we can't see where it actually came from (we just do a heat stack-show after the delete timeout in tripleo.sh).

Looking later in the logs we see:

2016-01-14 14:29:36.653 | | stack_status | DELETE_IN_PROGRESS

And the events show:

2016-01-14 14:31:44.207 | | Controller | c54bc713-f12f-41a0-a0dc-00e6ed606c00 | state changed | DELETE_IN_PROGRESS | 2016-01-14T14:28:01 |

which is the last event, relating to this resource, which I think is an OS::Nova::Server resource (we don't list resources recursively in the CI error path, so it's hard to be sure, we should fix that).

Looking at the heat logs:
2016-01-14 14:24:01.861 16286 INFO heat.engine.stack [-] Stack DELETE IN_PROGRESS (overcloud): Stack DELETE started
...
2016-01-14 14:35:00.891 16286 INFO heat.engine.stack [-] Stack DELETE COMPLETE (overcloud): Stack DELETE completed successfully

It appears the delete did actually work, but it took too long as the job timed out.

11 minutes seems excessive - my local 2 node stacks delete in about 40 seconds, so we may have an issue to diagnose specific to the HA job/stack here.