Uploaded August 2026 | Updated September 2026, 3 weeks ago
Don't miss out! Join us at our next KubeCon + CloudNativeCon events in Shanghai, China (8-9 September, 2026) and Salt Lake City, United States (Nov 9–12, 2026). Connect with our current graduated, incubating, and sandbox projects as the community gathers to further the education and advancement of cloud native computing. Learn more at kubecon.io
Sustainability by Design: Leveraging DRA for Energy-Efficient Kubernetes Clusters - Sunyanan Choochotkaew, IBM Research & Faseela Kundattil, Ericsson
Imagine a city where every machine automatically chooses the cleanest, most efficient energy source before starting work—so the whole system runs smarter and greener. Today, Kubernetes schedules workloads with little awareness of energy use. Workloads rarely see the energy characteristics of the hardware they run on, especially accelerators, and even when they do, it seldom maps to user-facing cost or SLOs.
Dynamic Resource Allocation (DRA) brings this vision closer to reality. Device providers can expose energy consumption and efficiency as attributes, letting Kubernetes prioritize energy-efficient devices. Tools like Kepler can use these metrics to estimate node energy use, powering carbon-aware multi-cluster scheduling and energy-driven autoscaling (e.g., KEDA).
We’ll explore the challenges of treating energy as a schedulable resource, demo a proof-of-concept for energy-aware device selection and power capping, and look ahead to ideas like scheduling on remaining energy capacity.
Don't miss out! Join us at our next KubeCon + CloudNativeCon events in Shanghai, China (8-9 September, 2026) and Salt Lake City, United States (Nov 9–12, 2026). Connect with our current graduated, incubating, and sandbox projects as the community gathers to further the education and advancement of cloud native computing. Learn more at kubecon.io
Sustainability by Design: Leveraging DRA for Energy-Efficient Kubernetes Clusters - Sunyanan Choochotkaew, IBM Research & Faseela Kundattil, Ericsson
Imagine a city where every machine automatically chooses the cleanest, most efficient energy source before starting work—so the whole system runs smarter and greener. Today, Kubernetes schedules workloads with little awareness of energy use. Workloads rarely see the energy characteristics of the hardware they run on, especially accelerators, and even when they do, it seldom maps to user-facing cost or SLOs.
Dynamic Resource Allocation (DRA) brings this vision closer to reality. Device providers can expose energy consumption and efficiency as attributes, letting Kubernetes prioritize energy-efficient devices. Tools like Kepler can use these metrics to estimate node energy use, powering carbon-aware multi-cluster scheduling and energy-driven autoscaling (e.g., KEDA).
We’ll explore the challenges of treating energy as a schedulable resource, demo a proof-of-concept for energy-aware device selection and power capping, and look ahead to ideas like scheduling on remaining energy capacity.
