I think you can overcome this issue creating a hypertable to track the joined data. Remember that continuous aggregates are hypertables.

Make a user defined action for every source of data to feed the hypertable independently. Each action can be scheduled independently but all feed same hypertable.