RETIRED, Puppet elements for deployment of OpenStack
Go to file
Emilien Macchi df92cede00 Migrate t-p-e to zuul v3 jobs
Define the zuul v3 layout for jobs that we want to execute.

Change-Id: I04d4fcafffe656dbcdc9b12dff60f4041f26d4ba
(cherry picked from commit dd018a8375)
2017-11-18 19:22:27 +00:00
doc/source Delete unnecessary utf-8 coding 2016-12-27 10:25:03 +08:00
elements Set remove-machine-id to executable 2017-09-26 13:44:18 +02:00
releasenotes Use Python to compute release notes version 2017-08-30 21:00:08 -07:00
tests Update repository after OpenStack migration 2014-10-02 09:53:08 -04:00
tools Add reno support 2017-01-27 10:58:48 -05:00
zuul.d Migrate t-p-e to zuul v3 jobs 2017-11-18 19:22:27 +00:00
.gitignore Add reno support 2017-01-27 10:58:48 -05:00
.gitreview Update .gitreview for stable/ocata 2017-02-16 12:59:28 +00:00
.testr.conf Update repository after OpenStack migration 2014-10-02 09:53:08 -04:00
LICENSE Basic structure imported from tripleo-image-elements 2014-09-08 18:58:08 -04:00
MANIFEST.in Basic structure imported from tripleo-image-elements 2014-09-08 18:58:08 -04:00
README.md Show team and repo badges on README 2016-11-25 18:44:57 +01:00
babel.cfg Basic structure imported from tripleo-image-elements 2014-09-08 18:58:08 -04:00
requirements.txt Basic structure imported from tripleo-image-elements 2014-09-08 18:58:08 -04:00
run-flake8 Update repository after OpenStack migration 2014-10-02 09:53:08 -04:00
setup.cfg Update repository after OpenStack migration 2014-10-02 09:53:08 -04:00
setup.py Basic structure imported from tripleo-image-elements 2014-09-08 18:58:08 -04:00
test-requirements.txt Add reno support 2017-01-27 10:58:48 -05:00
tox.ini Add reno support 2017-01-27 10:58:48 -05:00

README.md

Team and repository tags

Team and repository tags

Puppet building rules for OpenStack images

These elements are used to build disk images for deploying OpenStack via Heat. They are built as part of the TripleO (https://wiki.openstack.org/wiki/TripleO) umbrella project.

Instructions

Checkout this source tree and also the diskimage builder, export an ELEMENTS_PATH to add elements from this tree, and build any disk images you need.

git clone https://git.openstack.org/openstack/diskimage-builder.git
git clone https://git.openstack.org/openstack/tripleo-puppet-elements.git
export ELEMENTS_PATH=tripleo-puppet-elements/elements
diskimage-builder/bin/disk-image-create -u base vm bootstrap local-config stackuser heat-cfntools -a i386 -o bootstrap

Common element combinations

Always include heat-cfntools in images that you intend to boot via heat : if that is not done, then the user ssh keys are not reliably pulled down from the metadata server due to interactions with cloud-init.

Architecture

OpenStack images are intended to be deployed and maintained using Nova + Heat.

As such they should strive to be stateless, maintained entirely via automation.

Configuration

In a running OpenStack there are several categories of config.

  • per user - e.g. ssh key registration with nova: we repeat this sort of config every time we add a user.
  • local node - e.g. nova.conf or ovs-vsctl add-br br-ex : settings that apply individually to machines
  • inter-node - e.g. credentials on rabbitmq for a given nova compute node
  • application state - e.g. 'neutron net-create ...' : settings that apply to the whole cluster not on a per-user / per-tenant basis

We have five places we can do configuration in TripleO:

  • image build time
  • in-instance heat-driven (ORC scripts)
  • in-instance first-boot scripts [deprecated]
  • from outside via APIs
  • orchestrated by Heat

Our current heuristic for deciding where to do any particular configuration step:

  • per user config should be done from the outside via APIs, even for users like 'admin' that we know we'll have. Note that service accounts are different - they are a form of inter-node configuration.
  • local node configuration should be done via ORC driven by Heat and/or configuration management system metadata.
  • inter-node configuration should be done by working through Heat. For instance, creating a rabbit account for a nova compute node is something that Heat should arrange, though the act of creating is probably done by a script on the rabbit server - triggered by Heat - and applying the config is done on the compute node by the local node script - again triggered by Heat.
  • application state changes should be done from outside via APIs
  • first-boot scripts should not be used.

Copyright

Licensed under the Apache License, Version 2.0 (the "License"); you may not use this file except in compliance with the License. You may obtain a copy of the License at

http://www.apache.org/licenses/LICENSE-2.0

Unless required by applicable law or agreed to in writing, software distributed under the License is distributed on an "AS IS" BASIS, WITHOUT WARRANTIES OR CONDITIONS OF ANY KIND, either express or implied. See the License for the specific language governing permissions and limitations under the License.