Skip to content
This repository has been archived by the owner on Apr 6, 2021. It is now read-only.

Separate configuration from living data in the PAM module #166

Closed
ThomasHabets opened this issue Oct 10, 2014 · 1 comment
Closed

Separate configuration from living data in the PAM module #166

ThomasHabets opened this issue Oct 10, 2014 · 1 comment

Comments

@ThomasHabets
Copy link
Contributor

Original issue 167 created by raphink on 2012-04-21T09:02:48.000Z:

Currently, the PAM module relies on one file per user: ~/.google_authenticator.

While this is simple, it mixes the configuration and living data in one file. The problem I have is when deploying this file on machines automatically: the file I deploy is a configuration file, containing the secret key, parameters and scratch codes available.

If for example I use one of the scratch codes, the file gets modified, but the configuration manager (puppet for example) will replace it with the same scratch codes next time it runs.

Ideally, the PAM module would use two files:

  • One for static configuration (secret key, parameters, scratch codes);
  • One for living data (used scratch codes, timestamps for rate limit, etc.).
Sign up for free to subscribe to this conversation on GitHub. Already have an account? Sign in.
Projects
None yet
Development

No branches or pull requests

1 participant