Balancing on the Edge: Fast Service or No Service.

Edge computing operates between the cloud and end-users, and strives to provide computing services at a high rate. Since the computing and storage resources are limeted, directing more resources to some computing jobs will block (and pass to the cloud) the execution of others. We evaluate the system performance with two metrics: job computing time and job blocking probability. Edge nodes often operate in highly unpredictable environments, and replicating job execution when resources allow improves the job mean execution time. We show that the job computing time increases with the number of groups, but the job blocking probability does not. That is, there is a tradeoff between job computing time and blocking probability. This paper adopts the average system time as a single system's performance indicator to evaluate the tradeoff. We conclude that the optimal number of groups that maximizes the system service rate changes with arrival rate and cloud time.