Not sure if it really counts as a bug, but the system appears to stop processing on server restart. Is there any feasibility for a state save or similar to protect against this? We run a dedicated cluster, so there's at least a couple reboots a day, and I like to have the time higher for research for balance reasons.