How to leverage ML-based optimization to balance Kubernetes performance, resilience and cost-efficiency

Stefano Doni (Akamas CTO) was one of the guest experts for the CMG virtual event on “Cloud’s top trends for 2022”.

During his speech, Stefano describes how to address the challenge of ensuring that cloud-native applications get optimized in terms of service quality, resilience and cost-efficiency.

Stefano illustrates how ML-powered optimization makes it possible to automatically tune Kubernetes microservices applications, without spending hours of manual tuning. A real-case is discussed where the cloud bill was significantly reduced while better performance and resilience was achieved by simply letting ML automatically tune Kubernetes pod and application runtime configurations once the cost-reduction goal and the constraints reflecting SLOs in place were stated.