Cloud Solution Day 2016: Service Mesh for Kubernetes
Download as PPTX, PDF0 likes702 views
This document discusses service mesh for Kubernetes containers. It introduces Linkerd as a service mesh that provides features like load balancing, tracing, circuit breakers, retries, and more. It also discusses Kubernetes monitoring and logging capabilities as well as existing cluster add-ons for indexing and visualization. The document provides an overview of cloud native computing and common projects under the Cloud Native Computing Foundation, including Kubernetes for orchestration, Prometheus for monitoring, and Fluentd for logging.
Cloud Solution Day 2016: Service Mesh for Kubernetes
- 1. Service Mesh for Kubernetes Cloud Solution Day Vietnam - 2016
- 2. Who? Vincent De Smet (@vincentdesmet) - DevOps @ Honestbee (we’re hiring!) - Docker Saigon Organizer (Nov ‘15 - May ‘16) - Docker Captain (May ‘16 - Now) - Cloud Native Singapore Organizer (July ‘16 - Now)
- 3. Build a practical Vehicle Routing Solver
- 4. Kubernetes? - Google > Cloud Native Foundation - Container Orchestration?
- 5. Cloud Native? Creating a stable, operable and well integrated group of projects for "born in the cloud" applications
- 6. Evolution ~> distributed service oriented architecture Loosely Coupled Services Many Small Servers ~2000 Today Monolithic Big Servers Slow changing Rapidly updated
- 7. Docker - Cliff To Prod - Peer Discovery - Rolling Updates - Secrets / Config - Monitoring - Logging - Networking - ...
- 8. CNCF Hosted projects Current: - Orchestartion (Kubernetes) - Monitoring (Prometheus) - Distributed Tracing (OpenTracing) - Logging (Fluentd) Potential projects: - Networking (Flannel,...) - Configuration (etcd) - Event Streaming (Twitter Heron)
- 9. Kubernetes Monitoring - Integrated Discovery - Pull based
- 10. Kubernetes Logging - DeamonSet - Docker - json driver - Mount Log Directory - Add Meta Data
- 11. Existing Cluster Add-ons: Indexing & Visualisation - Analysis - Aggregation - Exploration
- 12. Linkerd - Twitter Finagle - Circuit breaking - Dynamic routing - ... - OpenZipkin support
- 13. Demo
- 14. Linkerd Features - Load balancing - Tracing - Circuit Breakers - Retries - Canary deploys - Load shedding - Error Tracking - Metrics - Service Discovery - Logging - Timeouts - Expirations - Security Policies - Back-offs - Retry Budgets - Dynamic Routing
- 15. Links - Buoyant blog series - Tracing Integration
Editor's Notes
- #2: Inspiration for this talk: https://blog.buoyant.io/2016/10/04/a-service-mesh-for-kubernetes-part-i-top-line-service-metrics/
- #6: At its root, Cloud Native is structuring teams, culture and technology to utilize automation and architectures to manage complexity and unlock velocity. Operating in this mode is as much a way to scale the people side of the equation as much as the technology side.
- #10: Integrated: New containers / services ran on the cluster are automatically discovered and registered.
- #13: service discovery, routing, failure handling, and visibility to modern software applications