Designing Cloud Native Applications with Kubernetes
21 likes52,425 views
The document discusses the key principles and benefits of designing cloud-native applications using Kubernetes, emphasizing the importance of automation, microservices, and containerization. It outlines key characteristics of cloud-native applications, Kubernetes abstractions, container design principles, and various design patterns. Additionally, it highlights the advantages of using Kubernetes, such as dynamic application placement, declarative service deployments, and application resilience.
Designing Cloud Native Applications with Kubernetes
- 1. Designing Cloud Native Applications with Kubernetes June 2017 @bibryam
- 2. 2 @bibryam Bilgin Ibryam ● Architect at Red Hat ● ASF Member ● Committer for Isis,Camel,OFBiz ● Microservices & Cloud Native ● Twitter: @bibryam ● Email: bibryam@gmail.com ● Blog: http://ofbizian.com ● Github: https://github.com/bibryam
- 3. 3 @bibryam Agenda ● Cloud Native Ecosystems ● Kubernetes Abstractions & Primitives ● Container Design Principles ● Kubernetes Design Patterns ● Benefits of using Kubernetes
- 4. 4 @bibryam Microservices Architecture ...trades code complexity for operational complexity
- 5. 5 @bibryam Cloud Native Reach
- 6. 6 @bibryam A Good Definition Cloud Native is structuring teams, culture and technology to utilize automation and architectures to manage complexity and unlock velocity. Joe Beda, Heptio
- 7. 7 @bibryam Common Characteristics Applications adopting the principles of Microservices packaged as Containers orchestrated by Platforms running on top of Cloud infrastructure, developed using practices such as Continuous Delivery and DevOps.
- 8. 8 @bibryam Brief Cloud (Native) History
- 9. 9 @bibryam Cloud Native Landscape
- 10. 10 @bibryam Cloud Native Ecosystems
- 11. 11 @bibryam Ecosystem Concerns ● Governance model ● Supporting organizations ● User community ● Culture, ways of working ● Technology (maturity * potential) ● Complementary tools ● Documentation, conferences, books ● Coolness factor and developer happiness
- 14. 14 @bibryam Potential Future Projects ● Tracing: OpenTracing ● Messaging: NATS ● Stream Processing: Heron ● Networking: Flannel, Calico, Weave ● Configuration: etcd ● Database: CockroachDB ● Storage: Minio
- 15. 15 @bibryam Agenda ● Cloud Native Ecosystems ● Kubernetes Abstractions & Primitives ● Container Design Principles ● Kubernetes Design Patterns ● Benefits of using Kubernetes
- 16. 16 @bibryam A Kubernetes Microservice
- 18. 18 @bibryam Common Abstractions & Primitives ● Application packaging (Container) ● Deployment unit (Pod) ● Recurring execution (CronJob) ● Service discovery & load balancing (Service) ● Application placement (Scheduler) ● Artifact grouping (Label) ● Resources isolation (Container/Namespace)
- 19. 19 @bibryam OOP Abstractions & Primitives
- 20. 20 @bibryam Java Abstractions & Primitives
- 21. 21 @bibryam K8S Abstractions & Primitives
- 22. 22 @bibryam Local & Distributed Primitives
- 23. 23 @bibryam Agenda ● Cloud Native Ecosystems ● Kubernetes Abstractions & Primitives ● Container Design Principles ● Kubernetes Design Patterns ● Benefits of using Kubernetes
- 24. 24 @bibryam (SOLID) Principles of Container Design ● Single Concern Principle (SCP) ● Self-Containment Principle (S-CP) ● Image Immutability Principle (IIP) ● High Observability Principle (HOP) ● Lifecycle Conformance Principle (LCP) ● Process Disposability Principle (PDP) ● Runtime Confinement Principle (RCP)
- 25. 25 @bibryam Single Concern Principle Patterns ● Sidecar ● Ambassador ● Adapter ● Init-container ● Defer-container
- 26. 26 @bibryam Self-Containment Principle AntiPattern ● Locomotive
- 27. 27 @bibryam Image Immutability Principle ● Dev/prod parity
- 28. 28 @bibryam High Observability Principle ● Spring Boot Actuator ● Dropwizard Metrics ● WildFly Swarm Monitor ● MicroProfile Healthchecks
- 29. 29 @bibryam Lifecycle Conformance Principle ● SIGTERM ● SIGKILL ● PreStop ● PostStart
- 30. 30 @bibryam Process Disposability Principle ● Fast startup ● Graceful shutdown ● Don’t rely on a particular instance ● Be aware of shots at your cattle ● Be robust against sudden death
- 31. 31 @bibryam Runtime Confinement Principle ● resources.limits.cpu ● resources.limits.memory ● resources.requests.cpu ● resources.requests.memory
- 32. 32 @bibryam More Container Best Practices ● Small images ● Arbitrary user IDs ● Port declarations ● Volume declarations ● Image metadata ● Host and image sync https://www.slideshare.net/luebken/container-patterns https://docs.docker.com/engine/userguide/eng-image/dockerfile_best-practices http://docs.projectatomic.io/container-best-practices https://docs.openshift.com/enterprise/3.0/creating_images/guidelines.html http://blog.arungupta.me/docker-container-anti-patterns/ https://12factor.net/
- 33. 33 @bibryam Agenda ● Cloud Native Ecosystems ● Kubernetes Abstractions & Primitives ● Container Design Principles ● Kubernetes Design Patterns ● Benefits of using Kubernetes
- 34. 34 @bibryam Container Oriented Design Patterns https://www.usenix.org/system/files/conference/hotcloud16/hotcloud16_burns.pdf
- 35. 35 @bibryam More Kubernetes Patterns Configuration Management –> ConfigMaps Singleton Services -> replicas: 1 Initializers -> Init Container Daemon Services –> Daemon Sets Stateful Services –> StatefulSet https://leanpub.com/k8spatterns/
- 36. 36 @bibryam Agenda ● Cloud Native Ecosystems ● Kubernetes Abstractions & Primitives ● Container Design Principles ● Kubernetes Design Patterns ● Benefits of using Kubernetes
- 37. 37 @bibryam Benefits of using Kubernetes 1) Self Service Environments 2)Dynamically Placed Applications 3)Declarative Service Deployments 4)Blue-Green and Canary Releases 5)Application Resilience & Antifragility 6)Service Discovery & Load Balancing 7)Service Clustering 8)(Scheduled) Batch Jobs 9)Immutable Infrastructure
- 38. 38 @bibryam Self Service Environments An Environment is not a VM any longer
- 39. 39 @bibryam Dynamically Placed Applications Forget about these manual placement strategies and...
- 40. 40 @bibryam Dynamically Placed Applications … and trust the Scheduler
- 41. 41 @bibryam Declarative Service Deployments Deployment config based options
- 42. 42 @bibryam Blue-Green and Canary Releases Routing config based options
- 43. 43 @bibryam Application Resilience For true resilience you must go outside of the JVM!
- 44. 44 @bibryam Application Resilience & Antifragility ● Infinite loops: CPU shares and quotas ● Memory leaks: OOM yourself ● Disk hogs: Quotas ● Fork bombs: Process limits ● Circuit Breaker, Timeout, Retry as SideCar ● Failover and Service Discovery as SideCar ● Process Bulkheading with Containers ● Hardware Bulkheading through Scheduler ● Auto-scaling & Self-healing
- 45. 45 @bibryam Service Discovery & Load Balancing Client side – on the JVM
- 46. 46 @bibryam Service Discovery & Load Balancing Provided by the platform
- 47. 47 @bibryam (Scheduled) Batch Jobs Provided by the platform Example batch jobs in Java ● JDK Timer ● ScheduledExecutorService ● Quartz Scheduler ● Spring Batch
- 48. 48 @bibryam Immutable Infrastructure Application binaries → Container Deployment unit → Pod Resource demands → request/limit/PVC Configurations → ConfigMap/Secret Update/rollback mechanism → Deployment
- 49. 49 @bibryam Key Takeaways 1. Automate routine tasks through a cloud native platform. (placement, updates, healthchecks, self-healing, scaling) 2. Move XFR/NFR from application to platform layer. (service discovery, circuit breaker, jobs, configurations, logs) 3. Ecosystem matters (Bet on a complete ecosystem, not a single platform)
- 50. 50 @bibryam Cloud Native is not Mandatory It is not necessary to change. Survival is not mandatory. W. Edwards Deming
- 51. Q & A @bibryam Kubernetes Patterns http://leanpub.com/k8spatterns More on this topic: https://twitter.com/bibryam