Haproxy needs to have gathered facts from all hosts that it will use
information about. In this case it must talk to all of the api hosts
as well as the database and rabbitmq hosts before generating the
configuration file.
Change-Id: I99b7dbebd5a6193e192ee258ddf576d18db90ed7
Closes-Bug: #1482406
Deploy the core services for murano-api and murano-engine containers.
Implements: bp ansible-murano
Change-Id: Ibcc618665a3509465ba8f9249a417e8848087eef
The stack-create operation fails. The current implementation of
Heat Ansible role does not implement trusts. Tell heat it should use
the non-trusts authentication model.
Also add a [keystone_clients] section which is required in newer
versions of heat to properly resolve the keystone v3 API endpoint
required for trust communication.
This fix spawns a Tech-Debt Bug to implement trusts in heat: #1492736
Change-Id: I06dd6436cc34a37b4966fddc888fade982ea51bd
Closes-Bug: #1492648
The default incoming database connection count is limited to 151
connections. These are not sufficient in a 100 node 3 controller
deployment to launch several heat stacks simulatenously.
Data measured on bare metal shows that 250 connections are needed
to launch one 25 VM heat stack with 3 controller nodes with 209
tasks (default enablement).
Change-Id: If60b540010d6d173a393fa91fa30cb3ba572cfc0
Closes-Bug: #1492719
Currently bootstrap containers are waited to exit but are not
checked for exit status and ansible runs further tasks. If
bootstrapping fails we notice it at much later time.
Change-Id: I137fc11b0f9d1f03d2ded08a213e8dbd62741f92
Closes-Bug: #1492337
Default value for auth_strategy changed for Cinder in Liberty
release from 'noauth' to 'keystone'. Therefore relying on
default value makes Cinder fail in Kilo. This patch covers
both releases by setting auth strategy to 'keystone' explicitely
Change-Id: Ic1f45fd43e0a94ca1dbfc63e90e2756c40453352
Closes-Bug: #1492233
The register operation was using database_user which is also a global
variable in group_vars/all.yml. This results in Heat being undeployable
because database_user is overwritten.
Change-Id: I78b2db34b7c99142040390c7258b7e81d3770835
Closes-bug: #1491980
When running register task for MariaDB we should firstly check
if operation failed before trying to parse result from json output
Change-Id: Ic331ec20990277d30b147826ff24affb27d52383
Closes-Bug: #1491751
If a bootstrap fails at some stage and leaves a rabbitmq_data
container on a host, it will not bootstrap again and rabbitmq
fails to start due to root permissions on /var/lib/rabbitmq.
This patch improves the check for a valid, existing
rabbitmq_data
Closes-Bug: #1490602
Change-Id: I6a00aaef38b4ab6a60dcfb5ed2f808513b9f8c84
Nova-compute is curerntly generating a backtrace when trying to nova
boot a VM. The backtrace is described in the bug log. Bind-mounting
/sys/fs/cgroup:/sys/fs/cgroup in the libvirt container resolves the
problem.
Change-Id: I5a56b61fd260728fb76e19bcbdf403b80a4ec5a6
Closes-Bug: #1490326
Initially it did not use this client due to the fact that the client
didn't seem to want to work in the containers. Now it works fine and
it should be used instead of the deprecated keystone client
Change-Id: Ib0e172c677e200e5df01c478b7ea257b30c0df0b
Closes-Bug: #1490126
This changes bootstrapping of the Heat container to bootstrap
the Heat container with a heat domain user. This requires some
work from bootstrap.yml to pass in several environment variables
needed by the heat domain setup script.
Co-Authored-By: Sam Yaple <sam@yaple.net>
Change-Id: Iab05983754fa514835cb5ff54d775faa18773110
Partially-implements: blueprint ansible-heat
This currently deploys the core services for a working Swift which are
account/container/object/proxy.
I've included some basic docs in docs/swift-related.rst, which gives
usage instructions and more context on this patch. These are really to
give an overview of the state of Swift in Kolla as of now, so unless
there's some major inaccuracy there please don't nitpick it.
Change-Id: Id0c54be3e24c46459c40b16b7020f05bddbe1b19
Implements: blueprint ansible-swift
nova.conf was not using the proper port value for glance and would
only work with the default port in that situation.
Additionally, the state_path defaults to the python directory rather
that the expected /var/lib/nova/ directory.
Paritially-Implements: blueprint update-configs
Change-Id: I9cc2117c1d786fc1b81c53848e543d27afaf8979
Use cinder-volumes as name of LVM volume group expected on the host
instead of the nonsense that was there before.
Change-Id: I35221ac0ad3464581f9ba95c3127832e0af7ad64
Closes-Bug: #1487800
Note, this is not in the inventories/site.yml as it will be a
transparent dependency of Swift.
Change-Id: Ic8c80072596594e646ece74ac0ce74c7000398ff
Partitially-Implements: blueprint ansible-swift
In MariaDB we adjust the dependecies of what we install there as well
adding only what is appropriate for the install.
In Ansible we adjust some templates to all work around differences
between the different Linux families.
Change-Id: Ibc26e2f4d4a732630632d3ed27fb595b6fe019d2
Partially-Implements: blueprint install-from-ubuntu
In rabbitmq we removed the requirement for hostname since that is
already installed.
Change-Id: Ia2355e332d53763db630d19663c2ddc8bf3aa425
Paritially-Implements: blueprint install-from-ubuntu
The register operation was registering user_database which is a global
variable in group_vars/all.yml. This results in glance not being deployable
because user_database is overwrriten at keystone task registration time.
The result of this is the playbooks fail to deploy.
Change-Id: Ic69725fb09aa0a368ec5b0d8f47bfc768c29c70c
Closes-Bug: #1486876
If ip_vs is not loaded keepalived will enter a restart loop. To fix that
load ip_vs module every time we start the container.
Change-Id: I6b39d8654cc6a6da7a9162e5c3efab073746b246
Closes-Bug: #1486276
Due to the differences between Ubuntu and Centos around Apache2 I have
had to copy both start.sh and config-external.sh to be unique files
for Ubuntu so as not to affect Centos either. This can be removed when
we change the way config-external works.
Change-Id: I14dc11e7bb6284b897259a53acc914ae8f254d3c
Partially-Implements: blueprint install-from-ubuntu
Rabbitmq containers mount /var/lib/rabbitmq from 'data'
containers with 'root:root' permissions because
ansible does not wait for rabbitmq bootstrap container -
the latter corrects access rights but sometimes gets
killed just after start.
Change-Id: I5cd1ce9810a1fd457dd34a1283e448204964c17d
Closes-Bug: #1487422
Bootstrap task runs bootstrap container on one host but
waits for its exit on all machines in inventory. This
gives error about non-existing container id.
Change-Id: Ie06c4da470c2e109d13b029fa674373b255653f3
Closes-Bug: #1487387
