When a worker is not responding to the ping, and consequently supposed MIA, we should: - ensure that the process is effectively killed and not frozen - if frozen, the process should be killed - start a new worker instance