[04:42:34] <ori>	 Krinkle: has anyone ever attempted to correlate backend latency with service / server uptime?