Cronitor is great. A few months back I used cronitor for a use case that ended up hitting their ping endpoint at about 10-15 QPS, and they said there was some capacity constraint they were running into (not exhausting though). I would love to hear more about the architecture behind cronitor.