A Django app for performing Akamai purge requests
django-akamai serves as a means to perform purge requests from Django apps using the Akamai REST API. Purge requests are performed on demand or, optionally, placed in a queue using Celery.
This library uses the edgegrid-python client for authentication. If the ~/.edgerc config file contains a CCU section those credentials will be used automatically.
To specify a different location for the edgerc file, you may use these Django settings:
If you prefer to keep the values in your Django settings you may specify them directly:
AKAMAI_CCU_CLIENT_SECRET AKAMAI_CCU_HOST AKAMAI_CCU_ACCESS_TOKEN AKAMAI_CCU_CLIENT_TOKEN
For simplicity and security use of the .edgerc file is recommended.
Consult Akamai’s documentation for full information about the API:
You may import PurgeRequest from django_akamai.purge and provide it with one or more URLs to invalidate or delete.
Note that Akamai’s API specifies a byte limit on the number of requests and a single purge call may require multiple HTTP requests to complete.
TODO: discuss options for rate-limiting
>>> pr = PurgeRequest() >>> pr.add("http://www.example.com/url-1.html") >>> pr.add(u"http://www.example.com/url-2.html") >>> for url_batch, response in pr.purge(): print(resp.status_code, len(url_batch)) 201 2 >>> print pr.urls 
django_akamai.signals defines two signals to directly issue a purge request or, when Celery is available, queue the request.
When sending these signals from other apps, you can pass in a variety of things as the sender for convenience. Sender can be a single URL string, a list of URL strings, an individual Django object, or a QuerySet. If passing in an object or QuerySet, then get_absolute_url() must be defined on every object.
Example of signalling to immediately perform the request:
>>> from django_akamai.signals import purge_request, queue_purge_request >>> obj = MyObject.objects.get(pk=3) >>> obj.get_absolute_url() u'http://www.example.com/blahblah.html' >>> purge_request.send(obj)
Or, to queue the request using Celery:
To use the task directly, import PurgeRequestTask from tasks.py thusly:
>>> from akamai.tasks import PurgeRequestTask >>> obj = MyObject.objects.get(pk=3) >>> PurgeRequestTask.delay(obj)