Introduction to stream processing with Apache Flink

After a quick description of event streams, and stream processing, this presentation moves to an introduction of Apache Flink :
- basic architecture
- sample code
- windowing and time concepts
- complex event processing CEP
- streaming analytics with Flink SQL

Transcript

1. © 2017 MapR Technologies MapR Confidential 1 Introduction to Stream Processing with Apache Flink Tugdual Grall @tgrall
2. © 2017 MapR Technologies @tgrall {“about” : “me”} Tugdual “Tug” Grall • MapR : Technical Evangelist • MongoDB, Couchbase, eXo, Oracle • NantesJUG co-founder
 • @tgrall • http://tgrall.github.io • tug@mapr.com / tugdual@gmail.com
3. © 2017 MapR Technologies @tgrall 3 Open Source Engines & Tools Commercial Engines & Applications Utility-Grade Platform Services Data Processing Web-Scale Storage MapR-FS MapR-DB Search and Others Real Time Unified Security Multi-tenancy Disaster Recovery Global Namespace High Availability MapR Streams Cloud and Managed Services Search and Others Unified Management and Monitoring Search and Others Event Streaming Database Custom Apps MapR Converged Data Platform
4. © 2017 MapR Technologies @tgrall Streaming Streaming technology is enabling the obvious: continuous processing on data that is continuously produced Hint: you already have streaming data
5. © 2017 MapR Technologies @tgrall Decoupling App B App A App C State managed centralized App B App A App C Applications build their own state
6. © 2017 MapR Technologies @tgrall Event Stream = Data Pipelines
7. © 2017 MapR Technologies @tgrall Streaming and Batch 2016-3-1
 12:00 am 2016-3-1
 1:00 am 2016-3-1
 2:00 am 2016-3-11
 11:00pm 2016-3-12
 12:00am 2016-3-12
 1:00am 2016-3-11
 10:00pm 2016-3-12
 2:00am 2016-3-12
 3:00am … partition partition
8. © 2017 MapR Technologies @tgrall Streaming and Batch 2016-3-1
 12:00 am 2016-3-1
 1:00 am 2016-3-1
 2:00 am 2016-3-11
 11:00pm 2016-3-12
 12:00am 2016-3-12
 1:00am 2016-3-11
 10:00pm 2016-3-12
 2:00am 2016-3-12
 3:00am … partition partition Stream (low latency) Stream (high latency)
9. © 2017 MapR Technologies @tgrall Streaming and Batch 2016-3-1
 12:00 am 2016-3-1
 1:00 am 2016-3-1
 2:00 am 2016-3-11
 11:00pm 2016-3-12
 12:00am 2016-3-12
 1:00am 2016-3-11
 10:00pm 2016-3-12
 2:00am 2016-3-12
 3:00am … partition partition Stream (low latency) Batch (bounded stream) Stream (high latency)
10. © 2017 MapR Technologies @tgrall Processing • Request / Response
11. © 2017 MapR Technologies @tgrall Processing • Request / Response • Batch
12. © 2017 MapR Technologies @tgrall Processing • Request / Response • Batch • Stream Processing
13. © 2017 MapR Technologies @tgrall Processing • Request / Response • Batch • Stream Processing • Real-time reaction to events • Continuous applications • Process both real-time and historical data
14. © 2017 MapR Technologies @tgrall
15. © 2017 MapR Technologies @tgrall Flink Architecture
16. © 2017 MapR Technologies @tgrall Flink Architecture Deployment Local Cluster Cloud Single JVM Standalone, YARN, Mesos AWS, Google
17. © 2017 MapR Technologies @tgrall Flink Architecture Deployment Local Cluster Cloud Single JVM Standalone, YARN, Mesos AWS, Google Core Runtime Distributed Streaming Dataflow
18. © 2017 MapR Technologies @tgrall Flink Architecture Deployment Local Cluster Cloud Single JVM Standalone, YARN, Mesos AWS, Google Core Runtime Distributed Streaming Dataflow DataSet API Batch Processing API & Libraries
19. © 2017 MapR Technologies @tgrall Flink Architecture Deployment Local Cluster Cloud Single JVM Standalone, YARN, Mesos AWS, Google Core Runtime Distributed Streaming Dataflow DataSet API Batch Processing API & Libraries FlinkML Machine Learning Gelly Graph Processing Table Relational
20. © 2017 MapR Technologies @tgrall Flink Architecture Deployment Local Cluster Cloud Single JVM Standalone, YARN, Mesos AWS, Google Core Runtime Distributed Streaming Dataflow DataSet API Batch Processing DataStream API Stream Processing API & Libraries FlinkML Machine Learning Gelly Graph Processing Table Relational
21. © 2017 MapR Technologies @tgrall Flink Architecture Deployment Local Cluster Cloud Single JVM Standalone, YARN, Mesos AWS, Google Core Runtime Distributed Streaming Dataflow DataSet API Batch Processing DataStream API Stream Processing API & Libraries FlinkML Machine Learning Gelly Graph Processing Table Relational CEP Event Processing Table Relational
22. © 2017 MapR Technologies @tgrall Demonstration Flink Basics
23. © 2017 MapR Technologies @tgrall Batch & Stream case class Word (word: String, frequency: Int) // DataSet API - Batch val lines: DataSet[String] = env.readTextFile(…) lines.flatMap {line => line.split(“ ”).map(word => Word(word,1))} .groupBy("word").sum("frequency") .print() // DataStream API - Streaming val lines: DataSream[String] = env.fromSocketStream(...) lines.flatMap {line => line.split(“ ”).map(word => Word(word,1))} .keyBy("word”).window(Time.of(5,SECONDS)) .every(Time.of(1,SECONDS)).sum(”frequency") .print()
24. © 2017 MapR Technologies @tgrall Steam Processing Source Filter /
 Transform Sink
25. © 2017 MapR Technologies @tgrall Flink Ecosystem Source Sink Apache Kafka MapR Streams AWS Kinesis RabbitMQ Twitter Apache Bahir … Apache Kafka MapR Streams AWS Kinesis RabbitMQ Elasticsearch HDFS/MapR-FS …
26. © 2017 MapR Technologies @tgrall Stateful Steam Processing Source Filter /
 Transform State
 read/write Sink
27. © 2017 MapR Technologies @tgrall Is Flink used?
28. © 2017 MapR Technologies @tgrall Powered by Flink
29. © 2017 MapR Technologies @tgrall 10 Billion events/day 2Tb of data/day 30 Applications 2Pb of storage and growing Source Bouyges Telecom : http://berlin.flink-forward.org/wp-content/uploads/2016/07/Thomas-Lamirault_Mohamed-Amine-Abdessemed-A-brief-history-of-time-with-Apache-Flink.pdf
30. © 2017 MapR Technologies @tgrall Stream Processing Windowing
31. © 2017 MapR Technologies @tgrall Stream Windows
32. © 2017 MapR Technologies @tgrall Stream Windows
33. © 2017 MapR Technologies @tgrall Stream Windows
34. © 2017 MapR Technologies @tgrall Stream Windows
35. © 2017 MapR Technologies @tgrall Stream Windows
36. © 2017 MapR Technologies @tgrall Demonstration Flink Windowing
37. © 2017 MapR Technologies @tgrall What about it ? What about it ? Time
38. © 2017 MapR Technologies @tgrall Time in Flink • Multiple notion of “Time” in Flink • Event Time • Ingestion Time • Processing Time
39. © 2017 MapR Technologies @tgrall What Is Event-Time Processing 1977 1980 1983 1999 2002 2005 2015 Processing Time Episode
 IV Episode
 V Episode
 VI Episode
 I Episode
 II Episode
 III Episode
 VII Event Time
40. © 2017 MapR Technologies @tgrall Time in Flink
41. © 2017 MapR Technologies @tgrall Complex Event Processing
42. © 2017 MapR Technologies @tgrall Complex Event Processing • Analyzing a stream of events and drawing conclusions • “if A and then B ! infer event C” • Demanding requirements on stream processor • Low latency! • Exactly-once semantics & event-time support
43. © 2017 MapR Technologies @tgrall Use Case
44. © 2017 MapR Technologies @tgrall Order Events Process is reflected in a stream of order events Order(orderId, tStamp, “received”) Shipment(orderId, tStamp, “shipped”) Delivery(orderId, tStamp, “delivered”) orderId: Identifies the order tStamp: Time at which the event happened
45. © 2017 MapR Technologies @tgrall Real-time Warnings
46. © 2017 MapR Technologies @tgrall CEP to the Rescue Define processing and delivery intervals (SLAs) ProcessSucc(orderId, tStamp, duration) ProcessWarn(orderId, tStamp) DeliverySucc(orderId, tStamp, duration) DeliveryWarn(orderId, tStamp) orderId: Identifies the order tStamp: Time when the event happened duration: Duration of the processing/delivery
47. © 2017 MapR Technologies @tgrall CEP Example
48. © 2017 MapR Technologies @tgrall Processing: Order ! Shipment
49. © 2017 MapR Technologies @tgrall Processing: Order ! Shipment val processingPattern = Pattern .begin[Event]("received").subtype(classOf[Order]) .followedBy("shipped").where(_.status == "shipped") .within(Time.hours(1))
50. © 2017 MapR Technologies @tgrall val processingPattern = Pattern .begin[Event]("received").subtype(classOf[Order]) .followedBy("shipped").where(_.status == "shipped") .within(Time.hours(1)) val processingPatternStream = CEP.pattern( input.keyBy("orderId"), processingPattern) Processing: Order ! Shipment
51. © 2017 MapR Technologies @tgrall val processingPattern = Pattern .begin[Event]("received").subtype(classOf[Order]) .followedBy("shipped").where(_.status == "shipped") .within(Time.hours(1)) val processingPatternStream = CEP.pattern( input.keyBy("orderId"), processingPattern) val procResult: DataStream[Either[ProcessWarn, ProcessSucc]] = processingPatternStream.select { (pP, timestamp) => // Timeout handler ProcessWarn(pP("received").orderId, timestamp) } { fP => // Select function ProcessSucc( fP("received").orderId, fP("shipped").tStamp, fP("shipped").tStamp – fP("received").tStamp) } Processing: Order ! Shipment
52. © 2017 MapR Technologies @tgrall Count Delayed Shipments
53. © 2017 MapR Technologies @tgrall Compute Avg Processing Time
54. © 2017 MapR Technologies @tgrall Demonstration Streaming Analytics
55. © 2017 MapR Technologies @tgrall Demonstration • https://github.com/mapr-demos/mapr-streams-flink-demo • https://github.com/mapr-demos/wifi-sensor-demo • http://tgrall.github.io/blog/2016/10/12/getting-started-with- apache-flink-and-kafka/ • http://tgrall.github.io/blog/2016/10/17/getting-started-with- apache-flink-and-mapr-streams/ • more soon….
56. © 2017 MapR Technologies @tgrall Kostas Tzoumas Stephan Ewen Fabian Hueske Till Rohrmann Jamie Grier Thanks to
57. © 2017 MapR Technologies @tgrall Streaming Architecture http://mapr.com/ebooks/ Free ebooks & Online training http://mapr.com/training/
58. © 2017 MapR Technologies MapR Confidential 58 Stream Processing with Apache Flink Tugdual Grall @tgrall