tuskar/doc/source/install.rst

Developer Installation Guide

The Tuskar source code should be pulled directly from git.

git clone https://git.openstack.org/openstack/tuskar

Dependencies

Setting up a local environment for development can be done with tox.

# install prerequisites
* Fedora/RHEL:
$ sudo yum install python-devel python-pip libxml2-devel \
    libxslt-devel postgresql-devel mariadb-devel

* Ubuntu/Debian:
$ sudo apt-get install python-dev python-pip libxml2-dev \
    libxslt-dev libpq-dev libmysqlclient-dev

Note

If you wish you run Tuskar against MySQL or PostgreSQL you will need also install and configure these at this point. Otherwise you can run Tuskar with an sqlite database.

To run the Tuskar test suite you will also need to install Tox.

$ sudo pip install tox

Note

An issue with tox requires that you use a version <1.70 or >= 1.7.2.

Now create your virtualenv.

$ cd <your_src_dir>/tuskar
$ tox -e venv

Note

If pip install fails due to an outdated setuptools, you can try to update it first.

$ sudo pip install --upgrade setuptools

To run the test suite use the following command. This will run against Python 2.6, Python 2.7 and run the flake8 code linting.

$ tox

Note

If you only have access to Python 2.6 or 2.7 locally pass in -e py26 or -e py27 respectively.

Configuration

Copy the sample configuration file:

$ cp etc/tuskar/tuskar.conf.sample etc/tuskar/tuskar.conf

We need to tell tuskar where to connect to database. Edit the config file in database section and change

#connection=<None>

to

connection=sqlite:///tuskar/tuskar.sqlite

Note

If you are using a different database backend, you will need to enter a SQLAlchemy compatible conection string <http://docs.sqlalchemy.org/en/latest/core/engines.html #database-urls> for this setting.

We need to initialise the database schema.

# activate the virtualenv
$ source .tox/venv/bin/activate

# if you delete tuskar.sqlite this will force creation of tables again - e.g.
# if you added a new resource table definitions etc in an existing migration
# file
$ tuskar-dbsync --config-file etc/tuskar/tuskar.conf

You can verify this was successful (in addition to seeing no error output) with.

$ sqlite3 tuskar/tuskar.sqlite .schema

Then, launch the app.

$ tuskar-api --config-file etc/tuskar/tuskar.conf

You can then verify that everything worked by running.

$ curl -v -X GET -H 'Accept: application/json' http://0.0.0.0:8585/v2/plans/ | python -mjson.tool

This command should return JSON with an empty result set.

Running Tuskar API

Whenever you want to run the API again, just switch to the virtualenv and run tuskar-api command.

$ source .tox/venv/bin/activate
$ tuskar-api --config-file etc/tuskar/tuskar.conf

Loading Initial Roles

Tuskar needs to be provided with a set of roles that can be added to a deployment plan. The following steps will add the roles from the TripleO Heat Templates repository.

$ git clone http://git.openstack.org/cgit/openstack/tripleo-heat-templates/
$ cd tripleo-heat-templates
$ tuskar-load-roles --config-file etc/tuskar/tuskar.conf \
    -r compute.yaml \
    -r controller.yaml

After this, if the Tuskar API isn't running, start it with the above command and the following curl command should show you the loaded roles.

$ curl -v -X GET -H 'Accept: application/json' http://0.0.0.0:8585/v2/roles/ | python -mjson.tool

Keystone Configuration

By default, Tuskar is configured to skip authentication for REST API calls. Keystone authentication can be enabled by making the appropriate changes to the tuskar.conf file as described in the keystone documentation <http://docs.openstack.org/developer/ keystone/configuringservices.html>

Contributing

For additional developer information, take a look at the contributing guide <contributing>.