The degradation is resolved and web functions are operating normally. We will be implementing mechanisms to protect against a recurrence of this issue in future.

We observed a 2 minute period where 30-40% of web function requests failed. This was due to an infrastructure change that left our proxies unable to handle new requests.