Download as PDF, PPTX
Uploaded byJonas Bonér
Go Reactive: Building Responsive, Resilient, Elastic & Message-Driven Systems
The document discusses the principles of reactive systems, emphasizing responsiveness, resilience, and elasticity in building modern applications. It covers how to manage failures, the importance of asynchronous message passing, and the necessity of a message-driven architecture to handle scalability effectively. Additionally, it critiques traditional distributed computing approaches while advocating for immutable state and specific design strategies to enhance system reliability and performance.
More Related Content
Similar to Go Reactive: Building Responsive, Resilient, Elastic & Message-Driven Systems
Go Reactive: Building Responsive, Resilient, Elastic & Message-Driven Systems
- 1. Go Reactive BuildingResponsive, Resilient, Elastic & Message-Driven Systems Jonas Bonér CTO Typesafe @jboner
- 3. The rules ofthe game have changed
- 4.
- 5. 3 Yesterday Today Single machines Clusters of machines
- 6. 3 Yesterday Today Single machines Clusters of machines Single core processors Multicore processors
- 7. 3 Yesterday Today Single machines Clusters of machines Single core processors Multicore processors Expensive RAM Cheap RAM
- 8. 3 Yesterday Today Single machines Clusters of machines Single core processors Multicore processors Expensive RAM Cheap RAM Expensive disk Cheap disk
- 9. 3 Yesterday Today Single machines Clusters of machines Single core processors Multicore processors Expensive RAM Cheap RAM Expensive disk Cheap disk Slow networks Fast networks
- 10. 3 Yesterday Today Single machines Clusters of machines Single core processors Multicore processors Expensive RAM Cheap RAM Expensive disk Cheap disk Slow networks Fast networks Few concurrent users Lots of concurrent users
- 11. 3 Yesterday Today Single machines Clusters of machines Single core processors Multicore processors Expensive RAM Cheap RAM Expensive disk Cheap disk Slow networks Fast networks Few concurrent users Lots of concurrent users Small data sets Large data sets
- 12. 3 Yesterday Today Single machines Clusters of machines Single core processors Multicore processors Expensive RAM Cheap RAM Expensive disk Cheap disk Slow networks Fast networks Few concurrent users Lots of concurrent users Small data sets Large data sets Latency in seconds Latency in milliseconds
- 14. Reactive “Readily responsiveto a stimulus” - Merriam Webster
- 15. The Principles ofReactive Systems Reactive Applications 6
- 17.
- 18. Responsive “Quick torespond or react appropriately” - Merriam Webster
- 19. 9 The systemshould always be responsive
- 20. 9 The systemshould always be responsive 1. Blue skies 2. Heavy load 3. Failures
- 21. Resilient “The abilityof a substance or object to spring back into shape” “The capacity to recover quickly from difficulties” - Merriam Webster
- 22.
- 23. Think Vending Machine Coffee Programmer Machine
- 24. Think Vending Machine Coffee Inserts coins Programmer Machine
- 25. Think Vending Machine Coffee Inserts coins Add more coins Programmer Machine
- 26. Think Vending Machine Coffee Inserts coins Add more coins Programmer Machine Gets coffee
- 27. Think Vending Machine Coffee Programmer Machine
- 28. Think Vending Machine Coffee Inserts coins Programmer Machine
- 29. Think Vending Machine Coffee Inserts coins Out of coffee beans error Programmer Machine
- 30. Think Vending Machine Coffee Inserts coins Out of coffee beans error WRONG Programmer Machine
- 31. Think Vending Machine Coffee Inserts coins Programmer Machine
- 32. Think Vending Machine Coffee Inserts coins Out of coffee beans failure Programmer Machine
- 33. Think Vending Machine Service Guy Coffee Inserts coins Out of coffee beans failure Programmer Machine
- 34. Think Vending Machine Service Guy Coffee Inserts coins Out of coffee beans failure Programmer Machine Adds more beans
- 35. Think Vending Machine Service Guy Coffee Inserts coins Programmer Machine Gets coffee Out of coffee beans failure Adds more beans
- 36. The Right Way Client Service
- 37. The Right Way Request Client Service
- 38. The Right Way Request Client Service Response
- 39. The Right Way Request Validation Error Client Service Response
- 40. The Right Way Request Validation Error Client Service Response Application Failure
- 41. The Right Way Supervisor Request Validation Error Client Service Response Application Failure
- 42. The Right Way Supervisor Request Validation Error Client Service Response Application Failure Manages Failure
- 44. Resilience Is ByDesign
- 46. Use Bulkheads 1.Isolate the failure 2. Compartmentalize 3. Manage failure locally 4. Avoid cascading failures
- 47. Use Bulkheads 1.Isolate the failure 2. Compartmentalize 3. Manage failure locally 4. Avoid cascading failures
- 48.
- 49. Supervisor hierarchies A Foo Bar B C B E A D C Automatic and mandatory supervision
- 50. Supervisor hierarchies A Foo Bar B C E A D C Automatic and mandatory supervision B
- 51. Supervision in Akka Every single actor has a default supervisor strategy. Which is usually sufficient. But it can be overridden. 19
- 52. Supervision in Akka class Supervisor extends AbstractActor { private SupervisorStrategy strategy = new OneForOneStrategy( 10, Duration.create(1, TimeUnit.MINUTES), DirectiveBuilder. match(ArithmeticException.class, e -> resume()). match(NullPointerException.class, e -> restart()). matchAny( e -> escalate()). build()); … // rest of actor omitted } } 19
- 54. Resilience requires a Message-Driven Architecture
- 55. Elastic “Capable ofready change or easy expansion or contraction” - Merriam Webster
- 56.
- 57. UP Scale anddown
- 58. 1. Minimize Contention 2. Maximize Locality of Reference 23 We need to
- 59. Common points of contention Physical Application
- 60.
- 61.
- 62.
- 63.
- 64.
- 65.
- 66. Needs to beasync and non-blocking 28 all the way down…
- 67. Needs to beasync and non-blocking 28 all the way down…
- 68. 29 Single WriterPrinciple
- 69. 29 Single WriterPrinciple Producers IO device SERIAL & CONTENDED
- 70. 29 Single WriterPrinciple Producers IO device SERIAL & CONTENDED Producers Actor or Queue IO device BATCHED & UNCONTENDED
- 71. The Role ofImmutable State 30
- 72. The Role ofImmutable State 30
- 73. The Role ofImmutable State • Great to represent facts • Messages and Events • Database snapshots • Representing the succession of time 30
- 74. The Role ofImmutable State • Great to represent facts • Messages and Events • Database snapshots • Representing the succession of time • Mutable State is ok if local and contained • Allows Single-threaded processing • Allows single writer principle • Feels more natural • Publish the results to the world as Immutable State 30
- 75.
- 76.
- 77. ON DEMAND 33 scale
- 78.
- 79.
- 80. • Mobile •Cloud Services, REST etc. • NOSQL DBs • Big Data 35 Distributed Computing is the new normal
- 81. What is theessence of distributed systems? 36
- 82. What is theessence of distributed systems? To try to overcome that 1. Information travels at the speed of light 2. Independent things fail independently 36
- 83. No difference Betweena Slow node Dead node 37 and a
- 84.
- 85. The network is 38 Inherently Unreliable http://aphyr.com/posts/288-the-network-is-reliable
- 86. 39 Graveyard ofdistributed systems
- 87. 39 Graveyard ofdistributed systems • Distributed Shared Mutable State EVIL N • (where N is number of nodes)
- 88. 39 Graveyard ofdistributed systems • Distributed Shared Mutable State EVIL N • (where N is number of nodes) • Serializable Distributed Transactions
- 89. 39 Graveyard ofdistributed systems • Distributed Shared Mutable State EVIL N • (where N is number of nodes) • Serializable Distributed Transactions • Synchronous RPC
- 90. 39 Graveyard ofdistributed systems • Distributed Shared Mutable State EVIL N • (where N is number of nodes) • Serializable Distributed Transactions • Synchronous RPC • Guaranteed Delivery
- 91. 39 Graveyard ofdistributed systems • Distributed Shared Mutable State EVIL N • (where N is number of nodes) • Serializable Distributed Transactions • Synchronous RPC • Guaranteed Delivery • Distributed Objects • “Sucks like an inverted hurricane” - Martin Fowler
- 92.
- 93. Instead Embrace theNetwork 40 and be done with it Use Asynchronous Message Passing
- 94.
- 95.
- 96. 42 Scaling Upand Out is essentially the same thing
- 98. Elasticity requires a Message-Driven Architecture
- 100. Asynchronous Message-Passing isthe enabler
- 101. 45 Typesafe ReactivePlatform • Purely asynchronous and non-blocking web framework • No container required, no inherent bottlenecks in session management • Actors are lightweight, isolated and and communicate via asynchronous message passing • Supervision and clustering in support of fault tolerance • Asynchronous and immutable programming constructs • Composable abstractions enabling simpler concurrency and parallelism
- 102.
- 103. 47 Finance Internet/SocialMedia Mfg/Hardware Government Retail
- 104. Go Reactive BuildingResponsive, Resilient, Elastic & Message-Driven Systems Jonas Bonér CTO Typesafe @jboner