RETIRED, Disk image elements for deployment images of OpenStack
Go to file
Yolanda Robla 95945ef3db Add a dependency on block-device-mbr element
When specifying the layout for the hardened image, we need
also to specify the device type, that in our case will be
mbr. So add this as a dependency of overcloud-secure to
properly set the type.

Change-Id: Ic0ad151cc12723e437a1123f90c6663198b9564d
(cherry picked from commit 950a63f77c)
2018-04-05 13:40:07 +02:00
doc/source Switch from oslosphinx to openstackdocstheme 2017-07-06 12:38:48 -07:00
elements Add a dependency on block-device-mbr element 2018-04-05 13:40:07 +02:00
releasenotes Use Python to compute release notes version 2017-08-30 14:32:04 -07:00
tests Add unit test to enforce unique file names. 2014-06-23 12:17:50 -04:00
tools Remove unused tox_install.sh 2017-12-01 07:31:07 +01:00
zuul.d Zuul: Remove project name 2018-02-01 03:37:10 -08:00
.gitignore Add reno support 2017-01-29 14:40:35 +00:00
.gitreview Update .gitreview for stable/queens 2018-03-03 11:43:32 +00:00
.testr.conf Add a .testr.conf configuration and ignore .testrepository. 2012-12-14 20:17:39 +13:00
HACKING.rst Create HACKING.rst for elements 2014-11-21 10:35:17 +00:00
LICENSE Fix copyrights for HP work. 2012-11-15 16:20:32 +13:00
MANIFEST.in Package with pbr 2013-08-06 12:50:39 +12:00
README.md Dummy patch to build a new rpm in RDO 2017-05-18 18:43:56 -04:00
babel.cfg Make it possible for openstack-CI to run tests 2013-02-04 22:26:17 -08:00
requirements.txt Update pbr version to match global requirements 2015-07-02 16:53:06 +01:00
run-flake8 PEP8 on the doc and elements files 2014-08-19 10:02:00 +02:00
setup.cfg Remove hardcoded version. 2014-08-20 11:01:36 +12:00
setup.py Align build files and requirements 2013-09-28 23:18:29 -04:00
test-requirements.txt Switch from oslosphinx to openstackdocstheme 2017-07-06 12:38:48 -07:00
tox.ini Add reno support 2017-01-29 14:40:35 +00:00

README.md

Team and repository tags

Team and repository tags

Image 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:

virtualenv .
source bin/activate
pip install dib-utils pyyaml
git clone https://git.openstack.org/openstack/diskimage-builder.git
git clone https://git.openstack.org/openstack/tripleo-image-elements.git
export ELEMENTS_PATH=tripleo-image-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)
  • 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

Copyright

Copyright 2012,2013 Hewlett-Packard Development Company, L.P. Copyright (c) 2012 NTT DOCOMO, INC.

All Rights Reserved.

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.