lastcomefirstserved

Last come, first served, often abbreviated LCFS, is a service discipline in queuing theory and certain operation contexts in which the most recently arrived customer is served before those who arrived earlier. It stands in contrast to the common first-come, first-served (FCFS) approach. LCFS can take non-preemptive forms, where service for a current job is completed before attending a newer arrival, or preemptive forms, where a new arrival can interrupt current service to be served next.

In queuing models, LCFS alters waiting-time distributions and fairness characteristics. Non-preemptive LCFS tends to favor newer arrivals, potentially reducing their waiting times while increasing delays for those who have been waiting longer. LCFS with preemption (often called LCFS-PR) can further reduce the response time of very short tasks, as new tasks can take priority immediately, but it introduces preemption overhead and can exacerbate starvation for older tasks. These properties have been studied in various models, including single-server queues with different arrival and service-time distributions, to understand the trade-offs between responsiveness and fairness.

Applications of LCFS are limited in everyday consumer settings because of fairness concerns, but the principle appears in certain real-time or freshness-centric scenarios, and in some theoretical or experimental contexts. In computing and data management, variants of last-arrival prioritization can be used in scheduling or cache-like systems where recency is a proxy for relevance or urgency. Critics argue that LCFS can lead to unpredictable wait times and unfair treatment of early arrivals.

See also: First-Come-First-Served, priority queue, queuing theory, preemptive scheduling, LCFS-PR.