GAE-friendly Chef API Client
This library implements a thin wrapper around the Chef server API using only PyCrypto functions for RSA signing.
Installation
We're on pypi:
pip install chef_gae
Usage/Samples
See code examples below.
You can find a list of endpoints on the Chef server API docs.
###Get all nodes
from gae_chef import Chef
rsa_key = open('chef_client_key.pem').read()
client = Chef(server_url=chef_server_url,
rsa_key=rsa_key,
chef_username=chef_username,
org_name=your_org_name)
#create a node
nodes = client.get_endpoint('GET', '/nodes')
#return value is a dictionary
for node in nodes:
print node['name']###Create a Node
from gae_chef import Chef
rsa_key = open('chef_client_key.pem').read()
client = Chef(server_url=chef_server_url,
rsa_key=rsa_key,
chef_username=chef_username,
org_name=your_org_name)
#data arg is a JSON-serialiazble dictionary
node_data = {'name': 'hello'}
new_node = client.get_endpoint('POST', '/nodes', data=node_data)
print new_node['uri']###Update a node WARNING: The Chef Server API is finicky about accepting updates. If you're planning to PUT data (eg updating a node), you'll probably want to GET the current status, then PUT the resulting object back. Example below:
from gae_chef import Chef
rsa_key = open('chef_client_key.pem').read()
client = Chef(server_url=chef_server_url,
rsa_key=rsa_key,
chef_username=chef_username,
org_name=your_org_name)
#get node first
node = client.get_endpoint('GET', '/nodes/hello')
#set some attribute from returned data
node['normal']['cool_users'] = ['ariel', 'christian', 'michael']
#update with modified returned data
client.get_endpoint('PUT', '/nodes/hello', data=node)You can also hit undocumented endpoints, like the ones surrounding access control. Same warning as above applies if you plan to update access control:
from gae_chef import Chef
rsa_key = open('chef_client_key.pem').read()
client = Chef(server_url=chef_server_url,
rsa_key=rsa_key,
chef_username=chef_username,
org_name=your_org_name)
access_list = client.get_endpoint('GET', '/nodes/hello/_acl')Testing
To run tests, install the packages in tests/requirements.txt. Then you can simply run py.test.
