Scheduling at the Edge of Reason: Multi-Cluster AI & OCM
summary
Demonstrates a dynamic scoring framework in Open Cluster Management that deploys agents as add-ons to feed a scoring API which generates continuous Placement API decisions for cross-cluster job dispatching; shows Federated Learning implemented as a CRD using a Flower add-on to aggregate edge training and avoid egress costs; mentions MultiKueue integration for batch jobs, a v1beta1 add-ons release, KEP 5313 placement decision API, project age five and version moving from 1.0 to 1.3 in a 5 minute lightning demo.
AI-written summary · about
related talks
05mProject Lightning Talk: Scheduling AI Workload Among Multiple Clusters
30mScore-Driven Multi-Cluster Management: An Evaluation Framework for…
27mBalancing Cost and Efficiency: Day2 Optimization of Multi-Cluster AI Infrastructure
04mProject Lightning Talk: Empowering Federated Learning with Multi-Cluster Management for Privacy and Efficiency
27mGen-AI at Scale: Simplifying Orchestration of Healthcare Applications Across Multi-Cluster Environme
16mCASSINI: Network-Aware Job Scheduling in Machine Learning Clusters
24mDemo of Kubernetes at the Edge for AI Model Deployments
39mScalable Multi-Node AI Workloads in Multi-Tenant AI Clouds U...