Server Debugging Process Improvements

[runt1me]

As mentioned in the closing comments and other comments on https://github.com/runt1me/stormcloud/issues/32, it would behoove us to have a more robust process for recognizing unhandled exceptions (server code 500 series) and mitigating them in code.

Ideally, we would have:

• Logs of the traffic (HTTP request or otherwise) that go to the web server. Maybe full pcaps if that's not too disk intensive?
• Python stack traces get pulled out of the stormcloud-error logs
• Some kind of script to aggregate this? Remote logging perhaps?

There are a million directions we can go for this and it's certainly not necessary to do all of it before we push out the MVP. Just noting my thoughts down here for a rainy day.

Estimated story points: 10

[EccentricKnight]

Dropping pre-release as this feels unspecific and not a showstopper.