Add token authentication to pulpcore
Nobody has claimed this yet.
- Dominant language
- Python
- Stars
- 598
- Forks
- 168
- Avg merge
- 1d 4h
- Merged PRs (30d)
- 86
Description
Author: @gerrod3 (gerrod)
Redmine Issue: 8939, https://pulp.plan.io/issues/8939
Background
Token authentication hands out a secret token to be used by a user to authenticate themselves. These tokens are passed in through the Authorization HTTP header with each request usually in the form of TOKEN {USER_TOKEN}. Tokens need to be kept secret like passwords and should only be used with https. Token auth can be implemented many different ways, but the general workflow follows:
- User visits a token view with basic credentials to receive a token
- Server generates a token that can be used to authenticate that user for future requests
- User uses token for token for authentication on future requests
- Token expires after set time or user deletes/generates a new token
Token auth can easily be added using the pre-built token authentication available in DRF: https://www.django-rest-framework.org/api-guide/authentication/#tokenauthentication. Two popular methods for token authentication are simple HTTP tokens and JSON Web token (JWT).
Simple HTTP Tokens
Basic token authentication comes included in DRF and can be added by including rest_framework.authtoken in the INSTALLED_APPS list and by adding TokenAuthentication to DRF's DEFAULT_AUTHENTICATION_CLASSES setting. This creates a model for storing tokens in the database. obtain_auth_token is a default view for generating tokens that can be added for users to receive their tokens.
Pros
- Simple to add and customize
- Admins and users can track tokens that have been deployed
- Lots of third party libraries that are built on top to add more functionality: Django-Rest-Durin, drfpasswordless, django-rest-knox, Djoser
Cons
- Adds an extra model to the database to maintain.
JSON Web Tokens
Background: https://jwt.io/introduction. JWTs consist of three encoded strings separated by dots which are: the header, the payload and the signature. JWTs use the signatures of the token to validate the authenticity of a token and have no need for a database to store them. The header tells how the token is encoded and the payload contains information, called claims, about the token like the user and expiration time for the token. The signature is created from the private key of the server and the encoded strings of the header and payload to validate the token. JWTs usually use the Bearer schema inside the Authentication header, e.g.: Authentication: Bearer {USER_TOKEN}. Since the signature validates the token an expiration is added to the payload to be able to invalidate it. djangorestframework-simplejwt implements JWTs for DRF by generating two tokens for users. One is a short-lived access token to authenticate and the other is a longer-lived refresh token that can be used to get another acess token.
Pros
- Simple to add using djangorestframework-simplejwt
- No need for database models
- Tokens expiration naturally built in
Cons
- No way to manually delete/track tokens without a database table
- Tokens need to be continuously fetch since long-lived tokens are ill-advised
Implementation Options
- Just use DRF's token auth or a pre-built DRF token auth package
- Just use JWT from djangorestframework-simplejwt
- Build a custom implementation on top of DRF tokens
Contributor guide
First steps
- Read the whole issue, then the project's contributing guide.
- Comment on the issue to say you are picking it up — it saves two people doing the same work.
- Fork the repository and make your change on a branch.
- Open a pull request that references the issue number.
Research direction
Start with the issue's DRF TokenAuthentication and djangorestframework-simplejwt references, then compare the three implementation options listed. Before coding, establish which token format, issuance flow, expiration behavior, and deletion or rotation behavior the project wants. Done should mean users can obtain tokens and use them for authenticated requests over HTTPS with the selected lifecycle rules.
Written by the indexing model from the issue text.
Assessment
- Tech stack
- django, python
- Domain
- api, authentication, backend
- Issue type
- Feature
- Difficulty
- 5/5
- Estimated time
- Over a week
- Activity status
- Stale
- Clarity
- Needs clarification
- Newbie friendliness
- 25/100