Has anyone experienced memory-leak like behavior in long-running processor worker? I had a few workers running for 10 days and the resource usage graph seems very suspicious: