Micro-Service Orchestration Deep Dive
summary
Compares Apache Mesos (and DC/OS/Marathon frameworks) to Kubernetes with concrete differences: Mesos uses a two-level scheduler and pushes state to agents with high-scale deployments (examples cited at tens to >100,000 nodes), Kubernetes centralizes control through the API server, uses one scheduler plus controllers and the reconciliation loop pattern, and introduces pods with a single IP and CRDs/operators. Notes Mesos excels at extreme scale and legacy flexibility; Kubernetes wins on documented scaling targets, developer UX, extensibility and unified scheduling.
AI-written summary · about
chapters (10)
- 0:00Introduction and background
- 7:49Legacy data centers vs orchestration
- 16:24Mesos architecture deep dive
- 23:32Kubernetes architecture overview
- 30:57Pods and controllers in Kubernetes
- 38:03Scheduling and reconciliation loops
- 45:02Containers pods and security
- 50:20Extensibility and custom resources
- 53:14Scaling and resource management
- 56:01Q&A and closing remarks
related talks
32mTechnical View: Comparison of Container Orchestration and Management Systems
53mDocker as the Building Block for Datacenter-Scale Applications
22mKeynote: The Future of Apache Mesos and DC/OS
53mMesos vs Kubernetes vs Swarm: Fight!
51mDeploying And Scaling Microservices
37mKnee Deep in Microservices
37mDistributed applications and Kubernetes: Frameworks or service meshes or both
33mContainer Networking for Micro-Services: An Apache Mesos Networking Deep Dive