Kubernetes on EGO -- Bringing Enterprise Resource Management and Scheduling to Kubernetes
summary
Describes IBM’s EGO resource manager (with features like time-based scheduling, placement policies, dynamic load-balancing, preemption and reclaim) and explains a connector that lets Kubernetes run on top of EGO. Covers concrete integration details showing how Kubernetes leverages EGO’s preemption/reclaim and shares scheduling with other workload-specific schedulers, plus lessons learned from operating them together.
AI-written summary · about
chapters (7)
- 0:00Start
- 9:08Architecture Overview
- 11:01EGO: Enterprise Resource Manager
- 11:26EGO: Hierarchical Consumer
- 12:35EGO: Enterprise Sharing Policies
- 20:15EGO: Smart Preemption
- 23:14EGO: Rich Variety of Resource Attributes and Resource Requirement Language
related talks
24mKubernetes-Powered Platform Engineering: From IaC To API-Driven Infrastructure
36mCapacity Scheduling for Elastic Resource Sharing in Kubernetes
23mCloud Foundry and Kubernetes: A Peanut Butter & Jelly Application
01h 01mKubernetes – going beyond the basics
46mExtend Your Kubernetes With the Power of Open Source
24mEdge Relocation - Smart Workload Migration Across Multiple Clusters
35mOrchestrating Interconnected Apps Across Geographically Distributed Kubernetes Clusters
01h 08mIntro + Deep Dive - Provider IBM Cloud