Change-Id: I13ee17a96033da75cbb377bce483f027127d646d

Hardcoding the first etcd host creates a single point of failure.

Change-Id: I0f83030fcd84ddcdc4bf2226e76605c7cab84cbb

etcd-compatible tooz drivers do not support multiple endpoints via
backend_url. We can put a loadbalancer in front of etcd and configure
backend_url to use the VIP instead. The issue with hard coding the first
host is that we break coordination if we take this host offline. In the
case of cinder, we would not be able to perform any volume related
operations.

Co-Authored-By: Mark Goddard <mark@stackhpc.com>
Change-Id: Ib684501ba03c386dc5ac71e5cbea05c99f191665

Depends-On: https://review.opendev.org/c/openstack/ansible-collection-kolla/+/872590
Change-Id: Ia8283f28147fbbd2d24028d01e021f95598b86fb

Rocky9 support landed in Zed release, we should start test them in
Antelope.

Change-Id: If4ca6aab660793015d577c3dfbeb7c75ca08c3fb

Change-Id: I115b491eca413437926f5bcaf53336151f9a7c0b

Change-Id: I99145322f65468e9926b1412844ad4ccaa6829d7

This patch fixes kolla_docker module
as it did not take into account common_options
parameter. From patchset it's visible that module's
default values are used always - even if user overrided
some param in common_options dict.

Closes-Bug: #2003079

Change-Id: I677fde708dd004decaff4bd39f2173d8d81052fb

It is deprecated in 20.10 and removed in 23.0 (and 23.0 is out) [1], [2].

[1]: https://docs.docker.com/engine/deprecated/#kernel-memory-limit
[2]: https://docs.docker.com/engine/api/version-history/#v142-api-changes

Change-Id: Ia6fa85172aad7bcd5f958922d3c224ef79882e6c

Change-Id: I0ff303a2fad2edbcedbe88486b272d2efa765d8d

This patch add connection local for above mentioned task as
kolla-ansible can be executed in docker container as in
my case.

When there is no connection: local, ansible is trying to connect
to localhost via ssh where specified python script is not available.

After connection: local everything is working as expected as file
is found inside container

Closes-Bug: #2004224

Change-Id: I219a958b4f101efb71a2935e6d910dae5c65f0be

Support to deploy skyline by kolla-ansible.

Implements: blueprint skyline
Depends-On: https://review.opendev.org/c/openstack/kolla/+/826948

Change-Id: Ice5621491a432ba32138abd6f62d1f815cc219e0

neutron_tls_proxy and glance_tls_proxy are using haproxy container
image. Pin them to haproxy_tag directly.

Change-Id: I73142db48ebe6641520d21b560f16de892e07c34

Users running on a Focal host will now fail in prechecks.

Change-Id: Icaef4b25458490e46f623b055658abc678d2f1c6

As per the RBAC new direction in Zed cycle, we have dropped the
system scope from API policies and all the policies are hardcoded
to project scoped so that any user accessing APIs using system scope
will get 403 error. It is dropped from all the OpenStack services
except for the Ironic service which will have system scope and to
support ironic only deployment, we are keeping system as well as project
scope in Keystone.

Complete discussion and direction can be found in the below gerrit
change and TC goal direction:

- https://review.opendev.org/c/openstack/governance/+/847418
- https://governance.openstack.org/tc/goals/selected/consistent-and-secure-rbac.html#the-issues-we-are-facing-with-scope-concept

As phase-2 of RBAC goal, services will start enabling the new
defaults and project scope by default. For example: Nova did in
- https://review.opendev.org/c/openstack/nova/+/866218

Kolla who start accessing the services using system scope token
- https://review.opendev.org/c/openstack/kolla-ansible/+/692179

This commit partially revert the above change except keeping
system scope usage for Keystone and Ironic. Rest all services are changed
to use the project scope token.

And enable the scope and new defaults for Nova which was disabled
by https://review.opendev.org/c/openstack/kolla-ansible/+/870804

Change-Id: I0adbe0a6c39e11d7c9542569085fc5d580f26c9d

Change-Id: I81b53b6b219865f6f16d58097e6c055058c06d4c