Deployment fails because RabbitMQ is down on one of the nodes

[Dmitry Mescheryakov]

Version: 10.0

Steps to reproduce:
1. Deploy an environment consisting of 3 controllers.

With rather high probability deployment fails because RabbitMQ fails to start on one of the nodes. Example fail: https://ci.fuel-infra.org/job/master.puppet-openstack.fuel-library.pkgs.ubuntu.review_in_fuel_library/2522/

Here unpack the snapshot and see see fuel/var/log/remote/node-3.test.domain.local/lrmd.log Here the problem is that the node was demoted, but not stopped. During demote rabbit app was shut down. But that is not discovered during subsequent monitor calls because we deleted monitor with CHECK level 30. So overall, it is a problem of outdated OCF script in our rabbitmq-server package for 10.0.

Failure of RabbitMQ to start on one node leads to deployment failure: http://paste.openstack.org/show/570269/

[Dmitry Mescheryakov (dmitrymex)]

Should be fixed by https://review.fuel-infra.org/#/c/26188. Right now we need to wait for CI team to enable that new package in the CI.

[Andres Toomsalu (andres-active)]

Im having same problem with 9.1 as well.