Google Cloud Platform are investigating reported downtime in their europe-west1-b region: https://status.cloud.google.com/incident/cloud-networking/18019
We believe this to be the cause of what we saw as timeouts and errors opening connections between machines, which may have surfaced themselves as 500s and 401s on the API; and error on the dashboard. This began at 14:01 UTC.
Since 14:25 UTC we have begun successfully receiving and processing API requests.
We are continuing to monitor for updates from GCP and API performance.