Service Mesh - The New Single Point of Failure

Explore the intricacies of service mesh implementation and its potential pitfalls in this 44-minute conference talk featuring Mitch Connors from Google, Sabeen Syed from HashiCorp, and Thomas Rampelberg from Buoyant. Delve into the architectural decisions behind popular service meshes like Istio, Linkerd2, and Consul, and understand their real-world implications for users. Learn about crucial aspects such as environment verification, safe upgrades, security updates, configuration stability, troubleshooting, and responsibility division. Gain insights on managing complexity, automated ACL setup, RBAC bridge, analyzers, events, and federation. Discover strategies for proactive alerts, mesh debugging, remote development, and designing for failure. Understand the distinctions between data plane and control plane, and explore enterprise-grade software use cases. Get involved in the ongoing development of service mesh technology and learn how to address the challenges of building resilient and scalable service meshes without relying on pre-existing mesh infrastructure.

Intro
About Sabeen
About Mitch
About Thomas
Complexity
Service Mesh
Microservices
Prometheus Integration
Mixer Architecture
Multiple Console Installations
Multiple Cluster Installations
Single Point of Failure
Upgrades
Managing Complexity
Automated ACL Setup
RBAC Bridge
Analyzers
Events
Federation
Its not done
The unsung heroes
Enterprise grade software use case
A difficult problem
An anecdote
A matter of maturity
How to dump service discovery information
Proactive Alerts
Mesh Debugging
Remote Development
Failure Static
Design for Failure
Data Plane vs Control Plane
Wrap Up
Getting Involved
Outro