Microservices in the Enterprise
Download as PPTX, PDF12 likes2,668 views
This document provides an overview of microservices in the enterprise. It discusses factors driving the rise of microservices like SOA fatigue and the need for faster innovation. Examples of microservice architectures from companies like Netflix, Twitter and Gilt are presented. Key capabilities for building enterprise-ready microservices are described, including service discovery, description, deployment isolation using containers, data/verb partitioning, lightweight middleware, API gateways and observability. Open source technologies that support implementing these capabilities are also outlined. The document concludes that microservices are the future of distributed systems and enterprises should implement solutions from first principles using inspiration from internet companies.
Microservices in the Enterprise
- 1. A Practical Guidance to Microservices in the Enterprise
- 2. About Us • Emerging technology firm focused on helping enterprises build breakthrough software solutions • Building software solutions powered by disruptive enterprise software trends -Machine learning and data science -Cyber-security -Enterprise IOT -Powered by Cloud and Mobile • Bringing innovation from startups and academic institutions to the enterprise • Award winning agencies: Inc 500, American Business Awards, International Business Awards
- 3. About This Webinar • Research that brings together big enterprise software trends, exciting startups and academic research • Best practices based on real world implementation experience • No sales pitches
- 4. • Microservices overview • Some inspirational architectures • Microservices in the enterprise • Enterprise microservices patterns • Capabilities • Technologies Agenda
- 5. Microservices: What’s the fuss all about?
- 6. • SOA fatigue • Top down has proven to be impractical in the enterprise • Large monolithic applications can’t evolve fast enough • Large monolithic applications can’t scale fast enough • Docker and the container revolution • Emergence of new and exciting programming platforms (NodeJS, GO, etc) • Friction between the need for innovation and the constrained enterprise software development practices Factors Contributing to the Raise of Microservices
- 8. A Definition of Microservices Loosely Coupled Service Oriented Architecture with Bounded Contexts
- 9. Loosely Coupled SOAs • Service dependencies • Component sharing • Database sharing • Centralized ESBs • Organizational coupling • Conway’s Law: https://en.wikipedia.org/wiki/Conway%27s_law
- 10. Bounded Contexts • Inspired by domain driven design • Encapsulates the details of a single business domain • Self-contained entity for the purpose of software development • Ability to update a microservices without knowledge of its peers
- 11. Microservices in the Real World
- 12. Some Examples http://www.infoq.com/presentations/Twitter-Timeline- Scalability http://www.infoq.com/presentations/twitter- soa http://www.infoq.com/presentations/Zipkin https://speakerdeck.com/mattheath/scaling-micro-services-in-go-highload-plus- plus-2014 AWS Re:Invent : Asgard to Zuul https://www.youtube.com/watch?v=p7ysHhs5hl0 Resiliency at Massive Scale https://www.youtube.com/watch?v=ZfYJHtVL1_w Microservice Architecture https://www.youtube.com/watch?v=CriDUYtfrjs http://www.infoq.com/presentations/scale-gilt http://www.slideshare.net/mcculloughsean/itier-breaking-up-the-monolith- philly-ete
- 13. Foundational Building Blocks of Microservices Architectures ConfigurationTooling Discovery Routing Observability Datastores Operational: Orchestration and Deployment Infrastructure Development: Languages and Container
- 14. Netflix OSS Microservices Architecture Edda Archaius Configuration Asgard Aminator Tooling Eureka Prana Discovery Denominator Zuul, Netty Ribbon 2.0 Routing Hystrix Pytheus SALP Observability Ephemeral datastores using Dynomite, Memcached, Astyanax, Staash, Priam,Cassandra Manual Orchestration with Asgard and deployment on AWS or Eucalyptus Java, Groovy, Scala, Clojure, Python, Node.js with AMI and Docker Containers
- 15. Twitter Microservices Architecture Decider ConfigurationTooling Finagle Zookeeper Discovery Finagle Netty Routing Zipkin Observability Custom Cassandra-like datastore: Manhattan Orchestration using Aurora deployment in datacenters using Mesos Scala with JVM Container
- 16. Gilt Microservices Architecture Decider Configuration Ion Cannon SBT Rake Tooling Finagle Zookeeper Akka Finagle Netty Discovery Routing Zipkin Observability Datastores per Microservice using MongoDB, Postgres, Voldemort Deployment on AWS Scala and Ruby with Docker Containers
- 17. Hailo Microservices Architecture Configuration Hubot Janky Jenkins Tooling go-platform Discovery go-platform RabbitMQ Routing Request trace Observability Datastore based on Cassandra Deployment on AWS Go using Docker
- 18. Microservices in the Enterprise
- 19. Challenges for Adopting Microservices in the Enterprise • Open source technology adoption • Organizational boundaries • Strict business processes • Traditional SOA mindset • Limited cloud adoption
- 20. Benefits of Adopting Microservices in the Enterprise • SOA that works • Building products instead of projects • Agility • Speed to market • Innovation • Remove friction for the adoption of new technologies
- 21. Microservices vs. SOA • SOA promise == Microservices reality • Federated innovation vs. Designed by committee • Small functional services vs. Large business services • REST and lightweight RPC vs. SOAP and WS-* • Lightweight middleware vs. ESBs • Decentralized governance vs. Centralized service repository • Development agility vs. Control
- 23. Relevant Capabilities of Enterprise Microservices Architectures • Service discovery • Service description • Deployment isolation • Lightweight middleware • Service gateway • Data Source partition • Verb partition
- 25. Capabilities • Removing coupling between microservices and client apps • Dynamically registering microservices in an enterprise topology • Allow client applications and other services to dynamically discover microservices and adapt to changes • Avoid the centralized registry pattern of traditional SOAs
- 26. Enterprise Microservices Discovery Pattern
- 27. Technologies • Consul.io ( https://consul.io/ ): DNS-style service discovery and configuration • Netflix’s Eureka ( https://github.com/Netflix/eureka ): AWS service registry used for locating services for the purpose of load balancing and failover • Zookeeper (https://zookeeper.apache.org/ ): Centralized service used for maintaining highly available configuration information • Etcd (https://github.com/coreos/etcd ): Distributed key value store optimized for service discovery
- 29. Capabilities • Express features of microservices in a descriptive format that can be understood by client applications • Manage microservices metadata • Simplify the creation of client artifacts • Manage versions of microservices
- 30. Enterprise Microservices Description Pattern
- 31. Technologies • Swagger (http://swagger.io/ ): Description language and description modeling tooling for RESTful services • API Blueprints (https://apiblueprint.org/ ): Description language and description modeling tooling for Web APIs: • Apache Thrift IDL (https://thrift.apache.org/ ): Highly scalable, cross language service development • Google’s gRPC IDL(http://www.grpc.io/ ): HTTP2 framework for cross platform service development
- 33. Capabilities • Enable internal communication between microservices • Provide high performance interactions between large number of microservices • Enable seamless cross language communication between microservices and client applications • Facilitate rapid microservices implementations across different languages
- 34. Enterprise Microservices IPC Pattern
- 35. Technologies • Apache Thrift (https://thrift.apache.org/ ): Highly scalable, cross language service development • Google’s gRPC (http://www.grpc.io/ ): HTTP2 framework for cross platform service development • Akka (http://akka.io/ ) : Framework for building highly concurrent, distributed applications • Twitter’s Finagle(https://twitter.github.io/finagle/ ): RPC framework for JVM services • Netty(http://netty.io/ ): Asynchronous, even driven framework for client server solutions
- 37. Capabilities • Isolate the infrastructure between microservices • Enable continuous deployment practices • Allow microservices portability across platforms
- 38. Enterprise Microservices Deployment Container Pattern
- 39. Technologies • Docker (https://www.docker.com/ ): Container platforms for packaging, shipping and distributing applications • Rocket(https://coreos.com/blog/rocket/ ): Runtime for Linux containers • Google’s Kubernetes(http://kubernetes.io/ ): Platform for managing containers
- 40. Microservices: Data source partition strategy
- 41. Capabilities • Enable a standard model for accessing data via microservices • Partition microservices at the data source level • Facilitate the composition of data access microservices • Allow flexible data access models for client applications
- 42. Enterprise Microservices Data Source Partition Pattern
- 43. Technologies • Facebook’s GraphQL(https://facebook.github.io/react/blog/2015/05/01/graphql- introduction.html) : URI-centric protocol for data fetching • Odata(https://http://www.odata.org/ ): REST protocol for querying data • Netflix’s Falcor(http://netflix.github.io/falcor/ ): Data access API framework for JSON data sources
- 44. Microservices: Verb/Use Case partition strategy
- 45. Capabilities • Efficiently partition microservices by use case or function • Compose microservices to enable more complex use cases • Facilitate the functional testing of microservices
- 46. Enterprise Microservices Verb Partition Pattern
- 47. Technologies • Apache Thrift (https://thrift.apache.org/ ): Highly scalable, cross language service development • Google’s gRPC (http://www.grpc.io/ ): HTTP2 framework for cross platform service development • Akka (http://akka.io/ ) : Framework for building highly concurrent, distributed applications • Twitter’s Finagle(https://twitter.github.io/finagle/ ): RPC framework for JVM services • Netty(http://netty.io/ ): Asynchronous, even driven framework for client server solutions
- 49. Capabilities • Extend microservices with simple middleware capabilities such as routing, transformation, persistent messaging etc • Provide a standard model to enable the communication between client apps and microservices • Expand the message exchange patterns supported by microservices solutions
- 50. Enterprise Microservices Lightweight Middleware Pattern
- 51. Technologies • RabbitMQ (https://www.rabbitmq.com/) : Simple messaging infrastructure for applications • Linkedin’s Kafka(http://kafka.apache.org/ ): Scalable publish-subscribe model for applications • ZeroMQ(http://zeromq.org/ ): Embeddable networking and messaging model for applications
- 52. API Gateway
- 53. Capabilities • Abstract the communication between client applications and internal microservices • Compose microservices into client-ready services • Extend microservices with enterprise ready capabilities
- 54. Enterprise Microservices API Gateway Pattern
- 55. Technologies • Mashape’s Kong (https://getkong.org/ ): Open source management platform for APIs and microservices • Apigee(): Market leader in API management • 3Scale(http://www.3scale.net/ ): API management platform • Azure API Gateway(https://azure.microsoft.com/en-us/services/api-management/ ): Azure native service for API and microservices management • AWS API Gateway(https://aws.amazon.com/api-gateway/ ): AWS native service for API management
- 57. Capabilities • Detect and prevent failures in complex microservices topologies • Trace request flow across microservices • Monitor service dependencies real time
- 58. Enterprise Microservices Observability Pattern
- 59. Technologies • Netflix’s Hystrix(https://github.com/Netflix/Hystrix ): Framework for detecting and preventing microservices failures • Twitter’s Zipkin(https://twitter.github.io/zipkin/ ): Framework for enabling request tracing across microservices • Trace(http://trace.risingstack.com/ ): Platform for tracing interactions between microservices • Spigo(https://github.com/adrianco/spigo ): Simulate request interactions between microservices
- 60. Other relevant microservices capabilities
- 61. Other Relevant Microservices Capabilities • Failure isolation • Federated databases • Design for failure • Distributed configuration management • Testing • Security
- 62. Summary • Microservices are the future of enterprise distributed systems • Enterprise microservices solutions need to be implemented from the ground up • We can drive inspiration from internet giants • Foundational blocks of microservices architectures include: -Discovery -Description -Lightweight middleware -Partition by data source -Partition by verb -IPC communication -API Gateway -Observability • Start small, iterate….