Mark regular non-containerized services with FIXME
to be switched, once it is containerized
Do not mark yet an external/backend/plugin/host-config
related puppet services templates with that FIXME
Mark puppet/services/ceph- related templates as TODO
switch it to containerized ceph-ansible eventually, maybe.
Change-Id: Ib9fbad05eeb57dc641499fbf411cb5870da7a8e9
Signed-off-by: Bogdan Dobrelya <bdobreli@redhat.com>
This commit introduces oslo.messaging services in place of a single
rabbitmq server. This will enable the separation of rpc and
notifications for the continued use of a single backend (e.g.
rabbitmq server) or a dual backend for the messaging communications.
This patch:
* add oslo_messaging_rpc and oslo_messaging_notify services
* add puppet services for rpc and notification
(rabbitmq and qdrouterd servers)
* add docker services to deploy rpc (rabbitmq or qdrouterd)
and notify (rabbitmq or shared)
* retains rabbit parameters for core services
* update resource registries, service_net_map, roles, etc.
* update ci environment container scenarios
* add environment generator for messaging
* add release note
Depends-On: Ic2c1a58526febefc1703da5fec12ff68dcc0efa0
Depends-On: I154e2fe6f66b296b9b643627d57696e5178e1815
Depends-On: I03e99d35ed043cf11bea9b7462058bd80f4d99da
Needed-By: Ie181a92731e254b7f613ad25fee6cc37e985c315
Change-Id: I934561612d26befd88a9053262836b47bdf4efb0
This change sets the ControllerPreConfig and ControllerPostConfig
resources to OS::Heat::None in order to avoid setting the cluster
to maintenance mode while upgrading to the docker architecture.
Closes-Bug: 1706594
Change-Id: Ibd69670d1617b73a8621b8fd4243041891a96747
The non-HA version of those two A/P Cinder services currently runs
non-containerized, as explained in I9ac74d6717533f59945694b4a43fe56d7ca768c6
and Ib10e4f18d967d356a15b97f58c488f8402a73356.
Disable their HA counterpart until the non-HA version is re-enabled.
Change-Id: I2aa49330fa361e330448dc9aa88e3812d9a7d464
Adds upgrade_tasks to remove the pacemaker resources using the
ansible-pacemaker module.
Resources are disabled and removed in step2 (called only on
bootstrap node) and then the cluster stop is moved to step3
The existing systemd/service call is kept but only to disable
services after they are disabled/deleted from the cluster.
Related-Bug: 1701485
Co-Authored-By: Damien Ciabrini <dciabrin@redhat.com>
Change-Id: Ia597d240ea5834c50a8f6c4fac0b6ed417b8535c
This adds a docker-ha.yaml that can be passed to the deployment
environments in order to get a containerized HA deployment.
Until we make the containerized deplyment the default the operator
must first include docker.yaml and *then* docker-ha.yaml in order
to get a containerized overcloud with an HA control plane.
We also make sure that the ClusterCheck service is set to None
by default and is part of the Controller roles.
Change-Id: I13204d70aad8dfeaf2bcf2ae30a1bb4715167659