This incident has been resolved.

We have rolled back a deploy that increased latency. Monitoring the recovery.

We're seeing increased latency with sending of messages. We're currently investigating.