Fault Domains in Mesos - Overview and Use Cases

Explore fault domains in Apache Mesos 1.4 through this 43-minute conference talk by Vinod Kone from Mesosphere, Inc. Gain insights into the native support for fault domains, enabling highly available applications and a common language for Mesos frameworks. Discover use cases such as workload bursting from on-premises to public cloud instances and rack-aware scheduling for stateful applications. Learn about fault domain hierarchy, terminology, and implementation details for masters, agents, and frameworks. Understand the benefits of this feature for organizations and see examples using Marathon. Delve into upgrade considerations and the current state of fault domains in Mesos.

Intro
About me
Use case #1: Fault tolerant scheduling
Bad scheduling
Use case #2: Hybrid Cloud
Hybrid Cloud Scheduling considerations
Existing solutions
Goals
Solution Overview
Fault Domain Hierarchy
Terminology
Implementation details
Master changes
Agent changes
Framework changes
Examples with Marathon
Upgrades
State of the feature
Acknowledgements