We have identified the problem which caused instances not being able to be created in the US compute shard. A mitigration has been applied an we are actively monitoring instance creation and job ingestion.