Tripleo-quickstart jobs are timing out in both run and collect logs

[Arx Cruz]

This happened today in tripleo-ci-centos-7-scenario001-multinode-oooq-container job.

The job timed out, and then started the collection of the logs, that have 10 minutes reserved to run, however, the collect logs was killed by global timeout set by devstack-gate script, leaving us without any logs from undercloud [1].

Furter investigation, showed that the collect-logs playbook stopped in:

TASK [collect-logs : Call get_host_info script] ********************************
Thursday 04 January 2018 02:56:23 +0000 (0:00:12.184) 0:06:55.098 ******
changed: [subnode-2]

As can be seen in [2]

Due that, we are not able to identify the root cause of the timeout in the job, since we don't have all the logs available to investigate.

1 - http://logs.openstack.org/07/472607/168/check/tripleo-ci-centos-7-scenario001-multinode-oooq-container/35d6683/

2 - http://logs.openstack.org/07/472607/168/check/tripleo-ci-centos-7-scenario001-multinode-oooq-container/35d6683/logs/quickstart_collect_logs.txt.gz

[wes hayutin (weshayutin)]

Looking to see how often this occurs

[Gabriele Cerami (gcerami)]

We've not seen any other occurrence of this error since the last. Changing priority to low

[Gabriele Cerami (gcerami)]

We didn't see it anymore at least in multinode, and anyway we increased the amount of time dedicated to the collection in ovb in the recent patch https://review.openstack.org/556884.
Marking as invalid