Apache Flink@ Strata & Hadoop World London

Stephan Ewen

Flink committer

co-founder @ data Artisans

@StephanEwen

ApacheFlink

1 year of Flink - code

April 2014 April 2015

What is Flink

3

Gelly

Table

ML

SA

MO

A

DataSet (Java/Scala/Python) DataStream (Java/Scala)

Hadoop M

/R

Local Remote Yarn Tez Embedded

Data

flow

Data

flow

(W

iP)

MR

QL

Table

Ca

sca

din

g (

WiP

)

Streaming dataflow runtime

Native workload support

5

Flink

Streaming

topologies

Long batch

pipelinesMachine Learning at scale

How can an engine natively support all these workloads?

And what does "native" mean?

Graph Analysis

E.g.: Non-native iterations

6

Step Step Step Step Step

Client

for (int i = 0; i < maxIterations; i++) {// Execute MapReduce job

}

E.g.: Non-native streaming

7

stream

discretizer

Job Job Job Jobwhile (true) {// get next few records// issue batch job

}

Native workload support

8

Flink

Streaming

topologies

Heavy

batch jobsMachine Learning at scale

How can an engine natively support all these workloads?

And what does native mean?

Flink Engine

1. Execute everything as streams

2. Allow some iterative (cyclic) dataflows

3. Allow some mutable state

4. Operate on managed memory

9

Program compilation

10

case class Path (from: Long, to:Long)val tc = edges.iterate(10) {

paths: DataSet[Path] =>val next = paths

.join(edges)

.where("to")

.equalTo("from") {(path, edge) =>

Path(path.from, edge.to)}.union(paths).distinct()

next}

Optimizer

Type extraction

stack

Task

scheduling

Dataflow

metadata

Pre-flight (Client)

MasterWorkers

Data

Sourceorders.tbl

Filter

MapDataSourc

elineitem.tbl

JoinHybrid Hash

build

HTprobe

hash-part [0] hash-part [0]

GroupRed

sort

forward

Program

Dataflow

Graph

deploy

operators

track

intermediate

results

Flink by Use Case

11

Data Streaming Analysis

streaming dataflows

12

3 Parts of a Streaming Infrastructure

13

Gathering Broker Analysis

Sensors

Transaction

logs …

Server Logs

3 Parts of a Streaming Infrastructure

14

Gathering Broker Analysis

Sensors

Transaction

logs …

Server Logs

Result may be fed back to the broker

Cornerstones of Flink Streaming

Pipelined stream processor (low latency)

Expressive APIs

Flexible operator state, streaming windows

Efficient fault tolerance for streams and state.

15

Pipelined stream processor

16

StreamingShuffle!

Expressive APIs

17

case class Word (word: String, frequency: Int)

val lines: DataStream[String] = env.fromSocketStream(...)

lines.flatMap {line => line.split(" ")

.map(word => Word(word,1))}

.window(Time.of(5,SECONDS)).every(Time.of(1,SECONDS))

.groupBy("word").sum("frequency")

.print()

val lines: DataSet[String] = env.readTextFile(...)

lines.flatMap {line => line.split(" ")

.map(word => Word(word,1))}

.groupBy("word").sum("frequency")

.print()

DataSet API (batch):

DataStream API (streaming):

Windows

18More at: http://flink.apache.org/news/2015/02/09/streaming-example.html

Checkpointing / Recovery

19

Chandy-Lamport Algorithm for consistent asynchronous distributed snapshots

Pushes checkpoint barriersthrough the data flow

Operator checkpointstarting

Checkpoint done

Data Stream

barrier

Before barrier =part of the snapshot

After barrier =Not in snapshot

Checkpoint done

checkpoint in progress

(backup till next snapshot)

Long batch pipelines

Batch on Streaming

20

Batch Pipelines

21

Batch on Streaming

Batch programs are a special kind of streaming program

22

Infinite Streams Finite Streams

Stream Windows Global View

PipelinedData Exchange

Pipelined or Blocking Exchange

Streaming Programs Batch Programs

Batch Pipelines

23

Data exchange (shuffle / broadcast)is mostly streamed

Some operators block (e.g. sorts / hash tables)

Operators Execution Overlaps

24

Memory Management

25

Memory Management

26

Smooth out-of-core performance

27More at: http://flink.apache.org/news/2015/03/13/peeking-into-Apache-Flinks-Engine-Room.html

Blue bars are in-memory, orange bars (partially) out-of-core

Table API

28

val customers = envreadCsvFile(…).as('id, 'mktSegment).filter("mktSegment = AUTOMOBILE")

val orders = env.readCsvFile(…).filter( o => dateFormat.parse(o.orderDate).before(date) ).as("orderId, custId, orderDate, shipPrio")

val items = orders.join(customers).where("custId = id").join(lineitems).where("orderId = id").select("orderId, orderDate, shipPrio,

extdPrice * (Literal(1.0f) – discount) as revenue")

val result = items.groupBy("orderId, orderDate, shipPrio").select('orderId, revenue.sum, orderDate, shipPrio")

Machine Learning Algorithms

Iterative data flows

29

Iterate by looping

for/while loop in client submits one job per iteration step

Data reuse by caching in memory and/or disk

Step Step Step Step Step

Client

30

Iterate in the Dataflow

31

Example: Matrix Factorization

32

Factorizing a matrix with28 billion ratings forrecommendations

More at: http://data-artisans.com/computing-recommendations-with-flink.html

Graph Analysis

Stateful Iterations

33

Iterate natively with state/deltas

34

Effect of delta iterations…

0

5000000

10000000

15000000

20000000

25000000

30000000

35000000

40000000

45000000

1 6 11 16 21 26 31 36 41 46 51 56 61

# o

f e

lem

en

ts u

pd

ate

d

iteration

… fast graph analysis

36More at: http://data-artisans.com/data-analysis-with-flink.html

Closing

37

Flink Roadmap for 2015

Some highlights that we are working on

More flexible state and state backends in streaming

Master Failover

Improved monitoring

Integration with other Apache projects

• SAMOA, Zeppelin, Ignite

More additions to the libraries

38

39

flink.apache.org

@ApacheFlink