Sammanfattning

This thesis compares three fair-scheduling algorithms, Round robin, Weighted fairqueuing and the Completely fair scheduler, and two message brokers, Kafka andRabbitMQ, in resource-intensive environments. In controlled docker setups, fairness is measured using Jain’s index, and efficiency is evaluated via throughput, waiting time and context-switch overhead. Round robin achieved the highest fairness of0.831, but the lowest throughput. WFQ recorded fairness of 0.593 with throughputof 0.095 while CFS combined fairness of 0.807 with high throughput 0.182.Both message brokers reached near-perfect fairness of about ≈1.00 on the Jain’s fairness index, and RabbitMQ delivered higher throughput than Kafka.

Utforska vidare

Liknande uppsatser

Uppsatser med liknande ämnen och nyckelord.