Python bindings to the OpenStack Nova API

This is a client for the OpenStack Nova API. There's a Python API (the novaclient module), and a command-line script (nova). Each implements 100% of the OpenStack Nova API.

See the OpenStack CLI guide for information on how to use the nova command-line tool. You may also want to look at the OpenStack API documentation.

The project is hosted on Launchpad, where bugs can be filed. The code is hosted on Github. Patches must be submitted using Gerrit, not Github pull requests.

python-novaclient is licensed under the Apache License like the rest of OpenStack.

Command-line API

Installing this package gets you a shell command, nova, that you can use to interact with any OpenStack cloud.

You'll need to provide your OpenStack username and password. You can do this with the --os-username, --os-password and --os-tenant-name params, but it's easier to just set them as environment variables:

export OS_USERNAME=openstack
export OS_PASSWORD=yadayada
export OS_TENANT_NAME=myproject

You will also need to define the authentication url with --os-auth-url and the version of the API with --os-compute-api-version. Or set them as an environment variables as well:

export OS_AUTH_URL=http://example.com:8774/v1.1/
export OS_COMPUTE_API_VERSION=1.1

If you are using Keystone, you need to set the OS_AUTH_URL to the keystone endpoint:

export OS_AUTH_URL=http://example.com:5000/v2.0/

Since Keystone can return multiple regions in the Service Catalog, you can specify the one you want with --os-region-name (or export OS_REGION_NAME). It defaults to the first in the list returned.

You'll find complete documentation on the shell by running nova help

Python API

There's also a complete Python API, but it has not yet been documented.

Quick-start using keystone:

# use v2.0 auth with http://example.com:5000/v2.0/")
>>> from novaclient.v1_1 import client
>>> nt = client.Client(USER, PASS, TENANT, AUTH_URL, service_type="compute")
>>> nt.flavors.list()
[...]
>>> nt.servers.list()
[...]
>>> nt.keypairs.list()
[...]