Adaptive Monitoring for Cloud-Native Microservices

Monitoring is essential in cloud environments, ensuring performance, efficient resource use, and application oversight. In microservices, it aids debugging and issue resolution, but its cost and resource impact must be minimized, especially in edge-cloud scenarios. In this paper, we propose the Adaptive Monitoring Controller (AdaMC) as a novel adaptive sampling approach for monitoring systems managing dynamic microservices and operated in edge-cloud environments. Our contribution is threefold. First, we present the problem of scrape frequency adaption for microservices. Second, we propose the AdaMC framework as a practical solution to the problem. Third, we evaluate AdaMC using real-world cloud scaling patterns, and different scenarios and compare the results with those of Prometheus monitoring tool. Our tests confirm that this dynamic adjustment can lead to significant resource savings and provide frequent monitoring depending on the behavior of the microservices.