We identified the issue as a network event with AWS Vendor code not recovering well from the event. We've logged an internal ticket to investigate if we can recover automatically from this in the future.

Network connectivity to AWS has experienced an issue. Some background processing is delayed.