Open Stack

I found useful to share a status on what's going on in CI now.

1) We fixed ovb-updates job: https://review.openstack.org/#/c/416706/
It should be stable again now. Please don't ignore it anymore (it was
for a few days until last night).

2) multinode & ovb-nonha are green and pretty stable.

3) ovb-ha is having (afik) 3 problems:

- pingtest not able to finish, floating-ip is not reachable.
It sounds like a problem in neutron, see the bug report.
John Schwarz (jschwarz) from Neutron team is actively working on it.
https://bugs.launchpad.net/tripleo/+bug/1654032
Note, the rate of failure was pretty high before we temp-reverted the
"broken" Neutron patch in TripleO CI:
http://logstash.openstack.org/#dashboard/file/logstash.json?query=build_name%3A%20*tripleo-ci*%20AND%20build_status%3A%20FAILURE%20AND%20message%3A%20%5C%22From%2010.0.0.1%20icmp_seq%3D1%20Destination%20Host%20Unreachable%5C%22
Since the Neutron patch has been reverted, we shouldn't hit this bug
anymore in TripleO CI, which increase our change to have successful
ovb-ha CI jobs correctly passing.

- pingtest not able to finish: 504 Gateway Time-out when creating a server.
This one has no launchpad yet. I'll create it today and investigate.
Note, the rate of failure is not that high:
http://logstash.openstack.org/#dashboard/file/logstash.json?query=build_name%3A%20*tripleo-ci*%20AND%20build_status%3A%20FAILURE%20AND%20message%3A%20%5C%22504%20Gateway%20Time-out%5C%22

- pingtest not able to finish: failing to upload an image in Glance
(with Swift backend)
https://bugs.launchpad.net/tripleo/+bug/1646750
Right now, the rate of failure is too low to be able to reproduce it
easily and have more debug. I'm working on it today, so we can ask
Swift or Glance folks to have a look when they can.


As a conclusion, CI should be pretty stable today:
http://tripleo.org/cistatus.html
Except for the ovb-ha job which might fail sometimes. Please be
careful when merging a patch that is not passing ovb-ha. Please look
at the logs and see if it hits one of the bugs mentioned here. We
don't want to inject more regressions.

Also, please do not try CI promotions until we have solved all blockers.

Thanks,
-- 
Emilien Macchi