Implement rate limits to address mis-use causing downtime

See: https://discord.com/channels/1358869848138059966/1437868748890571045/1447619393667469353

> Maybe the limit should be something like 60 reqs / minute, 1000 / hour per IP? Could be just on /servers to start. Still pretty lenient but would get rid of these breaking spikes