Uploaded byThang Chung
Avoid SPOF in Cloud-native Apps
The document discusses avoiding single points of failure in cloud-native applications, emphasizing the importance of resilient design, microservices architecture, and dynamic scaling. Key factors for cloud-native application development are outlined, including dependency management, API-first design, and fault tolerance. The document also addresses high-availability setups in Kubernetes, highlighting the configurations required to eliminate single points of failure.
Avoid SPOF in Cloud-native Apps
- 1. Avoid Single Pointof Failure in Cloud- native Application Phuong Le - Thang Chung Dec 2018
- 2. $ whoami • PhuongLe - Engineering Manager at NashTech - Over 15-years’ experience in the full life-cycle of the software. I have passion on people development and help them grow in their career path • Thang Chung - Solution Architect at NashTech - Over 10 years in software development industry, mainly focus on modern software architectures such as n-tiers, microservices, and cloud-native application. - https://github.com/thangchung - https://medium.com/@thangchung
- 3. • Key Takeaways -Cloud Native Fundamentals - Single Point of Failure in Cloud-native Apps • Q&A
- 4. Cloud Native Fundamentals “Itis the first step that is troublesome” - proverb
- 5. Cloud-native Apps Maturity •Microservices architecture and principles • API first design • Scale dynamically • Dynamic infrastructure migration without down-time L3: Cloud Native • Fault tolerant and resilient design • Metrics and monitoring build-in • Run anywhere, and cloud agnostic L2: Cloud Resilient • Loosely coupled systems • Horizontally scalable (services by name) • Follow 12 factors Apps • Leverage platform for high availability • Design for failure (include proactive testing for failure) L1: Cloud Friendly • No file system • Self-contained application • Run on VM with managed Ports and Addressing • Consume platform services L0: Cloud Ready Allan Beck & John McTeague (JPMorgan Chase & Co.) • Microservices & APIs • Resilience: monitoring, logging and exception handling • DevsOps • Cloud agnostic • 12 factors apps • Stateless & Scaling • Containers & Compute Units • Platforms & Services
- 6. Cloud Native ApplicationMaturity (cont.) Reference at Steinzeit war gestern! Die vielfältigen Wege der Cloud-nativen Evolution
- 7. Cloud Native ApplicationDevelopment Reference from Cloud Native Application development - A New Computing Paradigm by Oracle 2017
- 8. Design Cloud-native AppsFactors • Factor 1: One Code-base, One Application • Factor 2: API first • Factor 3: Dependency Management • Factor 4: Design, Build, Release and Run • Factor 5: Configuration, Credentials and Code • Factor 6: Logs • Factor 7: Disposability • Factor 8: Backing Services • Factor 9: Environment Parity • Factor 10: Administrative Processes • Factor 11: Port Binding • Factor 12: Stateless Processes • Factor 13: Concurrency • Factor 14: Telemetry • Factor 15: Authentication & Authorization
- 9. Single Point ofFailure in Cloud-native Apps “Every man is the architect of his own fortune” - proverb
- 10. Kubernetes High-Availability K8s HAis not just about the stability of Kubernetes itself. It is about setting up Kubernetes, along with supporting components such as etcd, in such a way that there is no single point of failure Lucas Käldström
- 11.
- 12. Factors can leadthe single point of failure •Load balancer to cluster •Kube-dns •Etcd •A single master cluster
- 13. Achieve Kubernetes HA •SetupEtcd cluster •Setup load balancer - Keepalived - HA proxy - NginX •Setup 3 master nodes •Join worker nodes
- 14.
- 15.