Process status tracking doesn't work

[SandersAaronD]

Currently a PR in to make sure default process status is "running".

To document expected behavior that we can build workflows around and that match other solutions:
When a process is added to the database, its status is "running".
An API call from that same process can set that process' status to "crashed" or "finished".
If it has been more than some interval of time, by default an hour, since that process has sent any data, that process should be set to "timed out". (Current behavior is that this happens to processes an hour after start, which is not correct).

This last thing requires something clever, like a reaper thread that is checking periodically if a given process has received data. It is much less important to have it in; by default for an MVP, we can expect to force users to mark processes "crashed" and consider them to remain running otherwise.