Comment by ykumar6

The blog post talks about how the api server was dropping web hooks, and there is no re-try mechanism built into most web hook senders.

Scaling out an api server is not resilient enough, and there needs to be a queue-based system fronting your web hooks so they aren't dropped