The processor is not a web app, so it has no health check endpoint that can be queried via HTTP. It would be good to have some kind of equivalent that runs once a minute so that we have a signal when processors are not healthy that can trigger terminating a processor instance.

Currently, the processor instance has multiple processes running inside of it:

• a multi-threaded processor app
• a disk cache manager

We could adjust the processor app and have it run through health check stuff in the loop where it acquires crash report ids.

We could add a third process that is a webapp that supports Dockerflow endpoints.

What do other non-webapp services do? Are there conventions/standards we should use?