A configuration change in our gcp-eu-west3 cluster caused connectivity issues from 10:39 to 10:46 UTC affecting 100% of requests during this interval. We've identified the problem and are introducing safeguards to prevent this from occurring in the future.