Dive deep into the architecture of open source orchestration platforms in this 42-minute conference talk from Docker. Explore higher-level abstractions like services and jobs in distributed systems, examining the actors and APIs that form the foundation of orchestration platforms. Analyze scaling challenges, discover extension points, and gain insights from an orchestration platform at Amazon. Learn about common architectures, interfaces, control loops, and the differences between orchestration and scheduling. Investigate networking stress, platform life cycles, and new abstractions. Witness a live demo of a homemade abstraction deployed on a multi-cloud Swarm cluster. Gain valuable knowledge on implementing APIs, running services, and understanding the complexities of orchestration in modern distributed systems.
Getting Deep on Orchestration - APIs, Actors, and Abstractions in Distributed Systems
Overview
Syllabus
Introduction
Before 2005
First time you really encounter orchestration
Amazon Apollo
Orchestration
DevOps Revolution
Common architectures
Interfaces
Control Loops
Orchestration vs Scheduler
Orchestration Patterns
What Breaks
Networking Stress
Platform Life Cycle
New Abstractions
Failure Injection
Entropy
Coffee Break
Running Services
Running Services Demo
Summary
What you need
Implementation
APIs
Be honest with yourself
Conclusion
Taught by
Docker
