Runner, Cron: Add docker health check

[JulianFP]

[JulianFP]

The problem with docker health checks on the runner side is that since the runner is just an http client there is no endpoint that could be tested with curl or anything like that. Somehow the health check command would need to find out if the heartbeat was send recently. I have the following ideas floating around:

• health check reads the log messages. For this every heartbeat would have to write a line into the log. This would make log files explode in size and make them unreadable.
• runner writes timestamp of latest heartbeat into some file (and rewrites that timestamp each time instead of adding a new one). Health check reads that file. It seems a bit hacky though and I would prefer to not touch the application code itself for implementing a docker-specific thing
• health check asks the backend. The problem is that if the runner is running on a different machine, then docker will not know under which url the backend runs. Also this would make the runners health dependent on the backends health.

Because of these issues and the fact that the health of runners isn't that important (there are supposed to be multiple runners for redundancy anyway) I am adjusting the priority of this