Hi All, I am making some first steps in putting the Slurm API to use in our clusters (at the CS and Engineering school of the Hebrew University in Jerusalem), and at this point I am looking at "read only" applications. For example, using Telegraf to read Cluster usage data into an SQL database (Postgres). I would like to use a better security mechanism than the standard JWT mechanism, that uses short and short-lived keys, provided by scontrol - which is already a problem when using Telegraf and will be more problematic when we would like to allow users to use the API to manage their jobs, etc.
So I wonder if some of you could share their experience and best practices for creating a more robust security scheme for the API. Basically we do not intend to give users direct access to the API, but we will have to provide indirect access which is gnostic of the user identity. Many thanks, Oren
