How to Build a Scalable, Distributed, Multi-Cloud API Architecture on Kubernetes
0 likes174 views
The document outlines the process of building a scalable, distributed, multi-cloud API architecture using Kubernetes, focusing on cloud-native and centralized API management. It introduces the WSO2 Microgateway, emphasizing its lightweight design, autoscaling capabilities, and ease of deployment. Key topics include deployment modes, the API operator for Kubernetes, and practical integration methods.
How to Build a Scalable, Distributed, Multi-Cloud API Architecture on Kubernetes
- 1. How to Build a Scalable, Distributed, Multi-Cloud API Architecture on Kubernetes September 02, 2020 1
- 2. Hello! Shehani Rathnayake Pubudu Gunatilaka Software Engineer, WSO2 pubudug@wso2.com shehanir@wso2.com Technical Lead, WSO2
- 3. ● Centralized API Management ● Cloud Native API Management ● Why Kubernetes? ● Deployment Modes ● API Operator for Kubernetes ● Private Jet Mode for Kubernetes ● Demo ● Q&A What’s covered... 3
- 6. Cloud Native API Management 6
- 8. Why Kubernetes? ● Auto scale API Gateways/ backend services ⦿ CPU/Memory ⦿ Custom metrics such as requests per minute ● Auto healing for the applications ● Easy to manage and deploy applications ● Zero down time rolling updates ● Canary deployments ● End to end fully automation capability 8
- 9. An Overview of an API Gateway 9
- 10. How to handle... ● Different resource usages ● Different Security enforcements ● Dynamic routing ● API mediation and transformation ● API Shaping ● Response Caching ● Private vs Public APIs ● API Gateway per department/unit 10
- 11. The Solution 11
- 13. WSO2 Microgateway ● Lightweight version of API gateway designed for MSA ● Designed to scale ⦿ ~1s startup time ⦿ Self-validating tokens ⦿ Localized rate limiting ⦿ Offline analytics ⦿ Immutable and Stateless ● Support for lifecycle management across environments ● Low Resource Environments ⦿ 40 MB of distribution size and docker base images. ⦿ < 250 MB of memory. ⦿ < 1 core CPU. 13
- 14. Cloud Native API Architecture 14
- 15. Cloud Native API Architecture 15
- 16. Multi Cloud API Architecture 16 VMs VMs
- 17. Data Plane 17
- 19. ● Scale separately ● Dedicated API microgateway for the API Private Jet Mode 19
- 20. ● Scale microservice with the API microgateway ● Dedicated API microgateway for the API Sidecar Mode 20
- 21. Shared Mode ● Multiple APIs deployed in the API microgateway 21
- 22. API Operator for Kubernetes 22
- 23. ● Makes API a first class citizen in Kubernetes eco system ● Easier to deploy ● Imposese ⦿ Security ⦿ Throttling ⦿ Mediation ⦿ Monitoring For your microservices ● Build in deployment patterns ● Fully automated experience for cloud native API management ● Single source of truth API Operator for Kubernetes 23
- 24. API CRD API Controller API Operator K8s API Server Docker Registry apictl add api -n products-api --from-file=products_swagger.yaml API Operator Overview 24
- 25. Private Jet Mode for APIs on Kubernetes 25
- 26. Private Jet Mode for APIs on Kubernetes ● WSO2 API Manager/Kubernetes integration ● Dedicated API microgateway for the API ● Single control plane to govern the data plane ● Expose via Ingress Gateway on Kubernetes ● Prerequisites ⦿ Kubernetes cluster with API Operator ⦿ WSO2 API Manager running in Kubernetes / VMs 26
- 27. The Workflow 27
- 28. Demo 28