queueround

Queueround is a queuing discipline that applies a round-robin principle to selecting the next item to be serviced from multiple input queues. In its basic form, service capacity is divided into rounds. During each round, the scheduler visits the queues in a fixed order and serves at most one item from each non-empty queue, continuing to the next queue after serving or when the round ends. The cycle repeats, cycling through the queues until all requests are processed. Variants may introduce weights, creating a weighted queueround where some queues receive more service slots per round based on priority or demand.

Queueround is designed to promote fairness and predictability, especially when multiple queues compete for shared service capacity. By giving each queue a turn, it reduces starvation risk for low-volume or bursty queues compared with a single FCFS queue. However, it can increase latency for long tasks and may be inefficient when service times vary greatly between queues. Implementation overhead includes maintaining order, skipping empty queues, and updating round boundaries.

Applications appear in computer networks for packet scheduling, in operating systems for multi-queue process scheduling, in messaging systems to distribute work across partitions, and in customer service environments to rotate through service lines. In practice, queueround-like policies are often hybridized with other disciplines to balance fairness, latency, and throughput.

Related concepts include round-robin scheduling, fair queuing, deficit round robin, and generalized processor sharing. Unlike strict FCFS in a single queue, queueround emphasizes fairness across multiple queues rather than timing within one queue.

The term queueround is used variably in technical discussions and does not have a single standardized definition; implementations differ in details such as ordering, rounding cadence, and weighting.