Ursula provides a series of Ansible playbooks for installing, managing, and maintaining OpenStack powered clouds.
Ursula was originally created by a team at Blue Box and is released under the MIT License (MIT).
The ceph-common
, ceph-monitor
, and ceph-osd
roles were originally taken from
ceph/ceph-ansible, but have since been
modified. ceph/ceph-ansible
is released under the Apache License.
The following system packages ( or their equivalents for your OS ) are
required to run ursula
:
- python-pip
- python-dev
- libxml2-dev
- libxslt-dev
- libffi-dev
- libssl-dev
We recommend using virtualenv or virtualenvwrapper to isolate your Python environment.
If you're new to python, the following will install virtualenvwrapper
and set
up a virtualenv
for ursula:
$ pip install virtualenvwrapper
$ source /usr/local/bin/virtualenvwrapper.sh
$ mkvirtualenv ursula
Note: If you're using OSX El Capitan, you need to use pip install --ignore-installed six virtualenvwrapper
to get pip to not attempt to uninstall the existing version of six which the system will not allow.
You will want to add source /usr/local/bin/virtualenvwrapper.sh
to your shell startup file, changing the path to virtualenvwrapper.sh
depending on where it was installed by pip.
echo " " >> .bash_profile (for OSX; .bashrc for various linux flavors)
echo "#sourcing statement for virtualenvwrapper" >> .bash_profile
echo "source /usr/local/bin/virtualenvwrapper.sh" >> .bash_profile
From now on to work with ursula you can run $ workon ursula
to
enter the virtualenv
Now that your python environment is ready, you can clone ursula and install its prerequisites.
You'll need a modern version of pip, so if you're using a version <7, run:
$ pip install -U pip
Now you can continue cloning and installing ursula:
$ cd ~/development
$ git clone [email protected]:blueboxgroup/ursula.git
$ cd ursula
$ pip install -r requirements.txt
These steps will have installed ursula-cli
, the various openstack clients, and our
patched fork of Ansible
.
Ursula was designed by Blue Box to manage a large
number of OpenStack deployments. In order to do this efficiently we've made
some changes to how ansible
works. As part of these changes we have a
wrapper tool called ursula-cli
which was installed during the
pip install -r requirements.txt
above.
Make sure ursula-cli
is installed in your environment:
ursula -h
usage: ursula [-h] [--ursula-forward] [--ursula-test] [--ursula-debug]
environment playbook
A CLI wrapper for ansible
...
...
There are two mandatory fields required by ursula-cli
. The first is
environment
which will require some further explanation.
The second is playbook
which will almost always be site.yml
.
One of the modifications that we have made to Ansible
is the ability to have
a seperate path that includes all of the configuration options for your
OpenStack deployment(s). An example of this can be found in /envs/example
If you look in the /envs/example
path, you'll see a defaults.yml
file and a
series of directories each representing a different OpenStack deployment.
We then utilize the standard Ansible
features by having group_vars
,
host_vars
, and a hosts
file.
There are also some vagrant.yml
files scattered around. These are helper
files to make using Vagrant
even easier with ursula
to test your
environments in VMs.
The simplest example deployment is allinone
which is a single server
deployment that acts as both a controller
and a compute
node.
Whether or not you're using Vagrant
if you look in the
envs/example/allinone/vagrant.yml
file it will give you some hints on what
your server should look like. If you do not wish to use Vagrant
then you
should install Ubuntu 12.04 on a server and configure its networking as
described in the vagrant.yml
file.
Next, look in the hosts
file. It's very simple in this case due to the fact
we have only a single server. This file combined with the site.yml
playbook
tells Ansible what roles to apply to which servers.
Finally, we have the group_vars/all.yml
file. This contains values that will
override the defaults.yml
in the parent directory. For example, we're
disabling Percona replication by setting percona.replication: False
.
For the sake of simplicity, I recommend using Vagrant
rather than Manual
for your first install.
If you want to install manually, do not use envs/examples/*
without modifications as it contains several circuit breakers such
as invalid certificates and your installs will fail.
If you're not running Vagrant
and have installed ubuntu onto a server and
configured the networking then we need to tell our system how to talk to this
new server. The easiest way is via an entry in your ssh config file in
~/.ssh/config
.
Host allinone
HostName 172.16.0.100
User ubuntu
IdentityFile ~/.ssh/private_key
$ ursula envs/example/allinone site.yml
If you're running Vagrant
, we have a wrapper script that stands up the
appropriate vagrant environment, saves it as an ssh config, and then calls
ursula
for you.
To deploy your allinone
environment via Vagrant
simply run:
$ ursula --provisioner=vagrant envs/example/allinone site.yml
Note: The default OS for ursula is Ubuntu Trusty. If you want Precise, set
the env var URSULA_BOX_NAME
to the name of your precise vagrant
box before running vagrant.
See the /doc directory of this repo.