In order to give the user an up-to-date overview of the changes it is considered to gather some data from a set of nodes (< 1000). Gathered data should be stored in order to let users perform some queries and generate graphs. Suitable components would be for instance:
This might be an excellent opportunity to learn about some techniques like NoSQL and Celery. For instance Couchdb appears to be more appropriate option than a RDBMS (like Postgres) for storing time-series data in django. However, unless the potential bandwidth is not extraordinary, the quest may be solved with a more pedestrian approach. Particularly, for this purpose a relational database like PostgreSQL would be sufficient and a cron triggered django management command gathers the data and stuffs it into the models.