Between 19:13 and 19:30 UTC, all of the redundant query engines on one of our partitions crashed, causing all queries to that partition to fail. They automatically recovered, and querying is back up. Customers may also have seen a delay in ingest until 19:30.