Yep, a decent canary mechanism should have caught this. There's a trade off between canarying and rollout speed, though. If this was a system for fighting bots, I'd expect it to be optimized for the latter.
Even if you want this data to be very fresh you can probably afford to do something like:
1. Push out data to a single location or some subset of servers.
2. Confirm that the data is loaded.
3. Wait to observe any issues. (Even a minute is probably enough to catch the most severe issues.)
4. Roll out globally.
Even if the servers weren't crashing it is possible that a bet set of parameters results in far too many false positives which may as well be complete failure.