Unified Batch and Real-Time Stream Processing Using Apache Flink

Unified Batch and Real-Time
Stream Processing Using
Apache Flink
Slim Baltagi
Director of Big Data Engineering
Capital One
September 15, 2015
Washington DC Area
Apache Flink Meetup

2
Agenda
1. What is Apache Flink?
2. Why Apache Flink?
3. How Apache Flink is used at Capital
One?
4. Where to learn more about Apache
Flink?
5. What are some key takeaways?

3
 Apache Flink, like Apache Hadoop and Apache
Spark, is a community-driven open source framework
for distributed Big Data Analytics.
Apache Flink has its origins in a research project
called Stratosphere started in 2009 at the Technische
Universität Berlin in Germany.
In German, Flink means agile or swift.
Flink joined the Apache incubator in April 2014 and
graduated as an Apache Top Level Project (TLP) in
December 2014 (the fastest Apache project to do so!)
DataArtisans (data-artisans.com) is a German start-
up company leading the development of Apache
Flink.

4
What is a typical Big Data Analytics Stack:
Hadoop, Spark, Flink, …?

5
Now, with all the buzz about Apache Spark, where
Apache Flink fits in the Big Data ecosystem and why do
we need Flink!?
Apache Flink is not YABDAF (Yet Another Big Data
Analytics Framework)!
Flink brings many technical innovations and a unique
vision and philosophy that distinguish it from:
 Other multi-purpose Big Data analytics frameworks
such as Apache Hadoop and Apache Spark
 Single-purpose Big Data Analytics frameworks such
as Apache Storm

• Declarativity
• Query optimization
• Efficient parallel in-
memory and out-of-
core algorithms
• Massive scale-out
• User Defined
Functions
• Complex data types
• Schema on read
• Real-Time
Streaming
• Iterations
• Memory
Management
• Advanced
Dataflows
• General APIs
Draws on concepts
from
MPP Database
Technology
Draws on concepts
from
Hadoop MapReduce
Technology
Add
1. What is Apache Flink? hat are the principles on
which Flink is built on?
Apache Flink’s original vision was getting the best from
both worlds: MPP Technology and Hadoop MapReduce
Technologies:

7
What is Apache Flink stack?
Gelly
Table
HadoopM/R
SAMOA
DataSet (Java/Scala/Python)
Batch Processing
DataStream (Java/Scala)
Stream Processing
FlinkML
Local
Single JVM
Embedded
Docker
Cluster
Standalone
YARN, Tez,
Mesos (WIP)
Cloud
Google’s GCE
Amazon’s EC2
IBM Docker Cloud, …
GoogleDataflow
Dataflow(WiP)
MRQL
Table
Cascading
Runtime - Distributed
Streaming Dataflow
Zeppelin
DEPLOYSYSTEMAPIs&LIBRARIESSTORAGE
Files
Local
HDFS
S3, Azure Storage
Tachyon
Databases
MongoDB
HBase
SQL
…
Streams
Flume
Kafka
RabbitMQ
…
Batch Optimizer Stream Builder
Storm

8
The core of Flink is a distributed and scalable
streaming dataflow engine with some unique features:
1. True streaming capabilities: Execute everything
as streams
2. Native iterative execution: Allow some cyclic
dataflows
3. Handling of mutable state
4. Custom memory manager: Operate on managed
memory
5. Cost-Based Optimizer: for both batch and stream
processing

9
1. What is Apache Flink? hat are the
principles on which Flink is built on?
1. Get the best from both worlds: MPP Technology and
Hadoop MapReduce Technologies.
2. All streaming all the time: execute everything as
streams including batch!!
3. Write like a programming language, execute like a
database.
4. Alleviate the user from a lot of the pain of:
manually tuning memory assignment to
intermediate operators
dealing with physical execution concepts (e.g.,
choosing between broadcast and partitioned joins,
reusing partitions)

10
1. What is Apache Flink? n?
5. Little configuration required
 Requires no memory thresholds to configure –
Flink manages its own memory
 Requires no complicated network configurations –
Pipelining engine requires much less memory for
data exchange
 Requires no serializers to be configured – Flink
handles its own type extraction and data
representation
6. Little tuning required: Programs can be adjusted
to data automatically – Flink’s optimizer can choose
execution strategies automatically

11
21. What is Apache Flink? n. What are the
principles on which Flink is built on?
7. Support for many file systems:
 Flink is File System agnostic. BYOS: Bring Your
Own Storage
8. Support for many deployment options:
Flink is agnostic to the underlying cluster
infrastructure.. BYOC: Bring Your Own Cluster
9. Be a good citizen of the Hadoop ecosystem
Good integration with YARN and Tez
10. Preserve your investment in your legacy Big Data
applications: Run your legacy code on Flink’s powerful
engine using Hadoop and Storm compatibilities layers
and Cascading adapter.

12
1. What is Apache Flink? n?
11. Native Support of many use cases:
 Batch, real-time streaming, machine learning,
graph processing, relational queries on top of the
same streaming engine.
Support building complex data pipelines leveraging
native libraries without the need to combine and
manage external ones.

13
Agenda
One?
Flink?

14
Apache Flink is uniquely positioned at the
forefront of the following major trends in the
Big Data Analytics frameworks:
1. Unification of Batch and Stream Processing
2. Multi-purpose Big Data analytics
frameworks
Apache Flink is leading the movement of
stream processing-first in the open source.
Apache Flink can be considered the 4G of the
Big Data Analytics Frameworks.

15
2. Why Apache Flink? - The 4G of Big Data
Analytics Frameworks
Big Data Analytics engines evolved?
 Batch  Batch
 Interactive
 Hybrid
(Streaming
+Batch)
 Interactive
 Near-Real Time
Streaming
 Iterative
processing
 In-Memory
 Hybrid
(Streaming
+Batch)
 Interactive
 Real-Time
Streaming
 Native Iterative
processing
 In-Memory
MapReduce Direct Acyclic
Graphs (DAG)
Dataflows
RDD: Resilient
Distributed
Datasets
Cyclic Dataflows
1G 2G 3G 4G

16
2. Why Apache Flink? - The 4G of Stream
Processing Tools
engineeolved?
 Single-
purpose
 Runs in a
separate
non-
Hadoop
cluster
 Single-
purpose
 Runs in the
same Hadoop
cluster via
YARN
 Hybrid
(Streaming
+Batch)
 Built for
batch
 Models
streams as
micro-
batches
 Hybrid
(Streaming
+Batch)
 Built for
streaming
 Models
batches as
finite data
streams
1G 2G 3G 4G

17
2. Why Apache Flink? – Good integration
with the Hadoop ecosystem
 Flink integrates well with other open source tools for
data input and output as well as deployment.
 Hadoop integration out of the box:
HDFS to read and write. Secure HDFS support
Deploy inside of Hadoop via YARN
Reuse data types (that implement Writables
interface)
 YARN Setup http://ci.apache.org/projects/flink/flink-docs-
master/setup/yarn_setup.html
 YARN Configuration
http://ci.apache.org/projects/flink/flink-docs-master/setup/config.html#yarn

18
Hadoop Compatibility in Flink by Fabian Hüske -
November 18, 2014 http://flink.apache.org/news/2014/11/18/hadoop-
compatibility.html
Hadoop integration with a thin wrapper (Hadoop
Compatibility layer) to run legacy Hadoop MapReduce
jobs, reuse Hadoop input and output formats and
reuse functions like Map and Reduce.
https://ci.apache.org/projects/flink/flink-docs-
master/apis/hadoop_compatibility.html
Flink is compatible with Apache Storm interfaces and
therefore allows reusing code that was implemented for
Storm.
https://ci.apache.org/projects/flink/flink-docs-master/apis/storm_compatibility.html

19
Service Open Source Tool
Storage/Servi
ng Layer
Data Formats
Data
Ingestion
Services
Resource
Management

20
Apache Bigtop (Work-In-Progress) http://bigtop.apache.org
Here are some examples of how to read/write data
from/to HBase:
 https://github.com/apache/flink/tree/master/flink-staging/flink-
hbase/src/test/java/org/apache/flink/addons/hbase/example
Using Kafka with Flink: https://ci.apache.org/projects/flink/flink-docs-
master/apis/ streaming_guide.html#apache-kafka
Using MongoDB with Flink:
http://flink.apache.org/news/2014/01/28/querying_mongodb.html
Amazon S3, Microsoft Azure Storage

21
 Apache Flink + Apache SAMOA for Machine Learning
on streams http://samoa.incubator.apache.org/
 Flink Integrates with Zeppelin
http://zeppelin.incubator.apache.org/
 Flink on Apache Tez
http://tez.apache.org/
 Flink + Apache MRQL http://mrql.incubator.apache.org
 Flink + Tachyon
http://tachyon-project.org/
Running Apache Flink on Tachyon http://tachyon-project.org/Running-
Flink-on-Tachyon.html
 Flink + XtreemFS http://www.xtreemfs.org/

22
2. Why Apache Flink? - Unification of
Batch & Streaming
Many big data sources represent series of events that
are continuously produced. Example: tweets, web logs,
user transactions, system logs, sensor networks, …
Batch processing: These events are collected together
for a certain period of time (a day for example) and
stored somewhere to be processed as a finite data set.
What’s the problem with ‘process-after-store’ model:
Unnecessary latencies between data generation and
analysis & actions on the data.
Implicit assumption that the data is complete after a
given period of time and can be used to make
accurate predictions.

23
Batch & Streaming
Many applications must continuously receive large
streams of live data, process them and provide results
in real-time. Real-Time means business time!
 A typical design pattern in streaming architecture
http://www.kdnuggets.com/2015/08/apache-flink-stream-processing.html
 The 8 Requirements of Real-Time Stream Processing,
Stonebraker et al. 2005 http://blog.acolyer.org/2014/12/03/the-8-
requirements-of-real-time-stream-processing/

24
2. Why Apache Flink? - Unification of Batch & Streaming
case class Word (word: String, frequency: Int)
val env = StreamExecutionEnvironment.getExecutionEnvironment()
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()
env.execute()
val env = ExecutionEnvironment.getExecutionEnvironment()
val lines: DataSet[String] = env.readTextFile(...)
lines.flatMap {line => line.split(" ")
.map(word => Word(word,1))}
.groupBy("word”).sum("frequency")
.print()
env.execute()
DataSet API (batch): WordCount
DataStream API (streaming): Window WordCount

25
Batch & Streaming
 Google Cloud Dataflow (GA on August 12, 2015) is a
fully-managed cloud service and a unified
programming model for batch and streaming big data
processing.
https://cloud.google.com/dataflow/ (Try it FREE)
http://goo.gl/2aYsl0
Flink-Dataflow is a Google Cloud Dataflow SDK
Runner for Apache Flink. It enables you to run
Dataflow programs with Flink as an execution engine.
The integration is done with the open APIs provided
by Google Data Flow.
Support for Flink DataStream API is Work in Progress

26
Batch & Streaming
Unification of Batch and Stream Processing:
In Lambda Architecture: Two separate execution
engines for batch and streaming as in the Hadoop
ecosystem (MapReduce + Apache Storm) or Google
Dataflow (FlumeJava + MillWheel) …
In Kappa Architecture: a single hybrid engine (Real-
Time stream processing + Batch processing) where
every workload is executed as streams including
batch!
Flink implements the Kappa Architecture: run
batch programs on a streaming system.

27
Batch & Streaming
References about the Kappa Architecture:
Batch is a special case of streaming- Apache Flink
and the Kappa Architecture - Kostas Tzoumas,
September 2015.http://data-artisans.com/batch-is-a-special-case-of-
streaming/
Questioning the Lambda Architecture - Jay Kreps ,
July 2nd, 2014 http://radar.oreilly.com/2014/07/questioning-the-lambda-
architecture.html
Turning the database inside out with Apache Samza -
Martin Kleppmann, March 4th, 2015
o http://www.youtube.com/watch?v=fU9hR3kiOK0 (VIDEO)
o http://martin.kleppmann.com/2015/03/04/turning-the-database-inside-
out.html(TRANSCRIPT)
o http://blog.confluent.io/2015/03/04/turning-the-database-inside-out-with-
apache-samza/ (BLOG)

28
Flink is the only hybrid (Real-Time Streaming +
Batch) open source distributed data processing
engine natively supporting many use cases:
Real-Time stream processing Machine Learning at scale
Graph AnalysisBatch Processing

29
2. Why Flink? - Alternative to MapReduce
1. Flink offers cyclic dataflows compared to the two-
stage, disk-based MapReduce paradigm.
2. The Application Programming Interface (API) for
Flink is easier to use than programming for
Hadoop’s MapReduce.
3. Flink is easier to test compared to MapReduce.
4. Flink can leverage in-memory processing, data
streaming and iteration operators for faster data
processing speed.
5. Flink can work on file systems other than Hadoop.

30
6. Flink lets users work in a unified framework allowing
to build a single data workflow that leverages,
streaming, batch, sql and machine learning for
example.
7. Flink can analyze real-time streaming data.
8. Flink can process graphs using its own Gelly library.
9. Flink can use Machine Learning algorithms from its
own FlinkML library.
10. Flink supports interactive queries and iterative
algorithms, not well served by Hadoop MapReduce.

31
11. Flink extends MapReduce model with new operators:
join, cross, union, iterate, iterate delta, cogroup, …
Input Map Reduce Output
DataSet DataSet
DataSet
Red Join
DataSet Map DataSet
OutputS
Input

32
2. Why Flink? - Alternative to Storm
1. Higher Level and easier to use API
2. Lower latency
Thanks to pipelined engine
3. Exactly-once processing guarantees
Variation of Chandy-Lamport
4. Higher throughput
Controllable checkpointing overhead
5. Flink Separates application logic from
recovery
Checkpointing interval is just a configuration
parameter

33
6. More light-weight fault tolerance strategy
7. Stateful operators
8. Native support for iterative stream
processing.
9. Flink does also support batch processing
10. Flink offers Storm compatibility
Flink is compatible with Apache Storm interfaces and
therefore allows reusing code that was implemented for
Storm.
https://ci.apache.org/projects/flink/flink-docs-
master/apis/storm_compatibility.html

34
‘Twitter Heron: Stream Processing at Scale’ by Twitter
or “Why Storm Sucks by Twitter themselves”!!
http://dl.acm.org/citation.cfm?id=2742788
 Recap of the paper: ‘Twitter Heron: Stream
Processing at Scale’ - June 15th , 2015
http://blog.acolyer.org/2015/06/15/twitter-heron-stream-processing-at-
scale/
High-throughput, low-latency, and exactly-once stream
processing with Apache Flink. The evolution of fault-
tolerant streaming architectures and their performance
– Kostas Tzoumas, August 5th 2015
http://data-artisans.com/high-throughput-low-latency-and-exactly-once-
stream-processing-with-apache-flink/

35
Clocking Flink to a throughputs of millions of
records per second per core
Latencies well below 50 milliseconds going to
the 1 millisecond range
References from Data Artisans:
 http://data-artisans.com/real-time-stream-processing-the-next-
step-for-apache-flink/
 http://data-artisans.com/high-throughput-low-latency-and-
exactly-once-stream-processing-with-apache-flink/
 http://data-artisans.com/how-flink-handles-backpressure/
 http://data-artisans.com/flink-at-bouygues-html/

36
2. Why Flink? - Alternative to Spark
1. True Low latency streaming engine
Spark’s micro-batches aren’t good enough!
Unified batch and real-time streaming in a single
engine
2. Native closed-loop iteration operators
Make graph and machine learning applications run
much faster
3. Custom memory manager
No more frequent Out Of Memory errors!
Flink’s own type extraction component
Flink’s own serialization component

37
2. Why Flink? - Alternative to Spark
4. Automatic Cost Based Optimizer
little re-configuration and little maintenance when the
cluster characteristics change and the data evolves
over time
6. Little tuning required
7. Flink has better performance

38
1. True low latency streaming engine
 Many time-critical applications need to process large
streams of live data and provide results in real-time.
For example:
Financial Fraud detection
Financial Stock monitoring
Anomaly detection
Traffic management applications
Patient monitoring
Online recommenders
 Some claim that 95% of streaming use cases can
be handled with micro-batches!? Really!!!

39
Spark’s micro-batching isn’t good enough!
Ted Dunning, Chief Applications Architect at MapR,
talk at the Bay Area Apache Flink Meetup on August
27, 2015
http://www.meetup.com/Bay-Area-Apache-Flink-
Meetup/events/224189524/
Ted described several use cases where batch and micro
batch processing is not appropriate and described
why.
He also described what a true streaming solution needs
to provide for solving these problems.
These use cases were taken from real industrial
situations, but the descriptions drove down to technical
details as well.

40
 “I would consider stream data analysis to be a major
unique selling proposition for Flink. Due to its
pipelined architecture, Flink is a perfect match for big
data stream processing in the Apache stack.” – Volker
Markl
Ref.: On Apache Flink. Interview with Volker Markl, June 24th 2015
http://www.odbms.org/blog/2015/06/on-apache-flink-interview-with-volker-markl/
 Apache Flink uses streams for all workloads:
streaming, SQL, micro-batch and batch. Batch is just
treated as a finite set of streamed data. This makes
Flink the most sophisticated distributed open source
Big Data processing engine (not the most mature one
yet!).

41
2. Iteration Operators
Why Iterations? Many Machine Learning and Graph
processing algorithms need iterations! For example:
 Machine Learning Algorithms
Clustering (K-Means, Canopy, …)
Gradient descent (Logistic Regression, Matrix
Factorization)
 Graph Processing Algorithms
Page-Rank, Line-Rank
Path algorithms on graphs (shortest paths,
centralities, …)
Graph communities / dense sub-components
Inference (Belief propagation)

42
 Flink's API offers two dedicated iteration operations:
Iterate and Delta Iterate.
 Flink executes programs with iterations as cyclic
data flows: a data flow program (and all its operators)
is scheduled just once.
 In each iteration, the step function consumes the
entire input (the result of the previous iteration, or the
initial data set), and computes the next version of the
partial solution

43
 Delta iterations run only on parts of the data that is
changing and can significantly speed up many
machine learning and graph algorithms because the
work in each iteration decreases as the number of
iterations goes on.
 Documentation on iterations with Apache Flink
http://ci.apache.org/projects/flink/flink-docs-master/apis/iterations.html

44
Step
Step
Step Step Step
Client
for (int i = 0; i < maxIterations; i++) {
// Execute MapReduce job
}
Non-native iterations in Hadoop and Spark are
implemented as regular for-loops outside the system.

45
 Although Spark caches data across iterations, it still
needs to schedule and execute a new set of tasks for
each iteration.
 Spinning Fast Iterative Data Flows - Ewen et al. 2012 :
http://vldb.org/pvldb/vol5/p1268_stephanewen_vldb2012.pdf The
Apache Flink model for incremental iterative dataflow
processing. Academic paper.
 Recap of the paper, June 18,
2015http://blog.acolyer.org/2015/06/18/spinning-fast-iterative-dataflows/
Documentation on iterations with Apache
Flinkhttp://ci.apache.org/projects/flink/flink-docs-
master/apis/iterations.html

46
3. Custom Memory Manager
Features:
 C++ style memory management inside the JVM
 User data stored in serialized byte arrays in JVM
 Memory is allocated, de-allocated, and used strictly
using an internal buffer pool implementation.
Advantages:
1. Flink will not throw an OOM exception on you.
2. Reduction of Garbage Collection (GC)
3. Very efficient disk spilling and network transfers
4. No Need for runtime tuning
5. More reliable and stable performance

47
public class WC {
public String word;
public int count;
}
empty
page
Pool of Memory Pages
Sorting,
hashing,
caching
Shuffles/
broadcasts
User code
objects
ManagedUnmanagedFlink contains its own memory management stack.
To do that, Flink contains its own type extraction
and serialization components.
JVM Heap
Network
Buffers

48
Peeking into Apache Flink's Engine Room - by Fabian
Hüske, March 13, 2015 http://flink.apache.org/news/2015/03/13/peeking-
into-Apache-Flinks-Engine-Room.html
Juggling with Bits and Bytes - by Fabian Hüske, May
11,2015
https://flink.apache.org/news/2015/05/11/Juggling-with-Bits-and-Bytes.html
Memory Management (Batch API) by Stephan Ewen-
May 16,
2015https://cwiki.apache.org/confluence/pages/viewpage.action?pageId
=53741525
Flink added an Off-Heap option for its memory
management component in Flink 0.10:
https://issues.apache.org/jira/browse/FLINK-1320

49
Compared to Flink, Spark is still behind in custom
memory management but is catching up with its
project Tungsten for Memory Management and
Binary Processing: manage memory explicitly and
eliminate the overhead of JVM object model and
garbage collection. April 28,
2014https://databricks.com/blog/2015/04/28/project-tungsten-bringing-
spark-closer-to-bare-metal.html
It seems that Spark is adopting something similar to
Flink and the initial Tungsten announcement read
almost like Flink documentation!!

50
4. Built-in Cost-Based Optimizer
 Apache Flink comes with an optimizer that is
independent of the actual programming interface.
 It chooses a fitting execution strategy depending on
the inputs and operations.
 Example: the "Join" operator will choose between
partitioning and broadcasting the data, as well as
between running a sort-merge-join or a hybrid hash
join algorithm.
 This helps you focus on your application logic
rather than parallel execution.
 Quick introduction to the Optimizer: section 6 of the
paper: ‘The Stratosphere platform for big data
analytics’http://stratosphere.eu/assets/papers/2014-
VLDBJ_Stratosphere_Overview.pdf

51
Run locally on a data
sample
on the laptop
Run a month later
after the data evolved
Hash vs. Sort
Partition vs. Broadcast
Caching
Reusing partition/sort
Execution
Plan A
Execution
Plan B
Run on large files
on the cluster
Execution
Plan C
What is Automatic Optimization? The system's built-in
optimizer takes care of finding the best way to
execute the program in any environment.

52
In contrast to Flink’s built-in automatic optimization,
Spark jobs have to be manually optimized and
adapted to specific datasets because you need to
manually control partitioning and caching if you
want to get it right.
Spark SQL uses the Catalyst optimizer that
supports both rule-based and cost-based
optimization. References:
Spark SQL: Relational Data Processing in
Sparkhttp://people.csail.mit.edu/matei/papers/2015/sigmod_spark_sql.pdf
Deep Dive into Spark SQL’s Catalyst Optimizer
https://databricks.com/blog/2015/04/13/deep-dive-into-spark-sqls-catalyst-
optimizer.html

53
 Flink requires no memory thresholds to
configure
 Flink manages its own memory
 Flink requires no complicated network
configurations
 Pipelining engine requires much less
memory for data exchange
 Flink requires no serializers to be configured
Flink handles its own type extraction and
data representation

54
6. Little tuning required
Flink programs can be adjusted to data
automatically
Flink’s optimizer can choose execution
strategies automatically
According to Mike Olsen, Chief Strategy
Officer of Cloudera Inc. “Spark is too knobby
— it has too many tuning parameters, and they need
constant adjustment as workloads, data volumes, user
counts change.”
Reference: http://vision.cloudera.com/one-platform/

55
7. Flink has better performance
Why Flink provides a better performance?
Custom memory manager
Native closed-loop iteration operators make graph
and machine learning applications run much faster.
Role of the built-in automatic optimizer. For
example: more efficient join processing.
Pipelining data to the next operator in Flink is more
efficient than in Spark.
See benchmarking results against Flink here:
http://www.slideshare.net/sbaltagi/why-apache-flink-is-the-4g-of-big-
data-analytics-frameworks/87

56
Agenda
One?
Flink?

57
3. How Apache Flink is used at Capital One?
We started our journey with Apache Flink at Capital
One while researching and contrasting stream
processing tools in the Hadoop ecosystem with a
particular interest in the ones providing real-time
stream processing capabilities and not just micro-
batching as in Apache Spark.
 While learning more about Apache Flink, we
discovered some unique capabilities of Flink which
differentiate it from other Big Data analytics tools not
only for Real-Time streaming but also for Batch
processing.
We are currently evaluating Apache Flink capabilities
in a POC.

58
Where are we in our Flink journey?
Successful installation of Apache Flink 0.9 in
testing Zone of our Pre-Production cluster running
on CDH 5.4 with security and High Availability
enabled.
Successful installation of Apache Flink 0.9 in a 10
nodes R&D cluster running HDP.
We are currently working on a POC using Flink for a
real-time stream processing. The POC will prove
that costly Splunk capabilities can be replaced by a
combination of tools: Apache Kafka, Apache Flink
and Elasticsearch (Kibana, Watcher).

59
What are the opportunities for using Apache
Flink at Capital One?
1. Real-Time stream analytics after
successful conduction of our streaming
POC
2. Cascading on Flink
3. Flink’s MapReduce Compatibility Layer
4. Flink’s Storm Compatibility Layer
5. Other Flink libraries (Machine Learning
and Graph processing) once they come
out of beta.

60
Cascading on Flink:
 First release of Cascading on Flink is being announced
soon by Data Artisans and Concurrent. It will be
supported in upcoming Cascading 3.1.
 Capital One will be the first company to verify this
release on real-world Cascading data flows with a
simple configuration switch and no code re-work
needed!
 This is a good example of doing analytics on bounded
data sets (Cascading) using a stream processor (Flink)
 Expected advantages of performance boost and less
resource consumption.
 Future work is to support ‘Driven’ from Concurrent Inc.
to provide performance management for Cascading
data flows running on Flink.

61
 Flink’s DataStream API 0.10 will be released soon and
Flink 1.0 GA will be at the end of 2015 / beginning of
2016.
Flink’s compatibility layer for Storm:
We can execute existing Storm topologies using
Flink as the underlying engine.
We can reuse our application code (bolts and
spouts) inside Flink programs.
 Flink’s libraries (FlinkML for Machine Learning and
Gelly for Large scale graph processing) can be used
along Flink’s DataStream API and DataSet API for our
end to end big data analytics needs.

62
Agenda
One?
Flink?

63
4. Where to learn more about Flink?
To get an Overview of Apache Flink:
http://www.slideshare.net/sbaltagi/overview-of-
apacheflinkbyslimbaltagi
To get started with your first Flink project:
Apache Flink Crash Course
http://www.slideshare.net/sbaltagi/apache-
flinkcrashcoursebyslimbaltagiandsrinipalthepu
Free Flink Training from Data Artisans
http://dataartisans.github.io/flink-training/

64
Flink at the Apache Software Foundation: flink.apache.org/
data-artisans.com
@ApacheFlink, #ApacheFlink, #Flink
apache-flink.meetup.com
github.com/apache/flink
user@flink.apache.org dev@flink.apache.org
Flink Knowledge Base (One-Stop for all Flink
resources) http://sparkbigdata.com/component/tags/tag/27-flink

65
50% off Discount Code: FlinkMeetupWashington50
Consider attending the first dedicated Apache Flink
conference on October 12-13, 2015 in Berlin,
Germany! http://flink-forward.org/
Two parallel tracks:
Talks: Presentations and use cases
Trainings: 2 days of hands on training workshops
by the Flink committers

66
Agenda
One?
Flink?

67
1. Although most of the current buzz is about Spark,
Flink offers the only hybrid (Real-Time Streaming +
Batch) open source distributed data processing
engine natively supporting many use cases.
2. I foresee more maturity of Apache Flink and more
adoption especially in use cases with Real-Time
stream processing and also fast iterative machine
learning or graph processing.
3. I foresee Flink embedded in major Hadoop
distributions and supported!
4. Apache Spark and Apache Flink will both have their
sweet spots despite their “Me Too Syndrome”!

68
Thanks!
To all of you for attending and/or reading the
slides of my talk!
To Capital One for hosting and sponsoring
the first Apache Flink Meetup in the DC Area.
http://www.meetup.com/Washington-DC-Area-Apache-Flink-Meetup/
Capital One is hiring in Northern Virginia and
other locations!
Please check jobs.capitalone.com and
search on #ilovedata

Unified Batch and Real-Time Stream Processing Using Apache Flink

Recommended

More Related Content

What's hot (20)

Viewers also liked (20)

Similar to Unified Batch and Real-Time Stream Processing Using Apache Flink (20)

More from Slim Baltagi (6)

Recently uploaded (20)

Unified Batch and Real-Time Stream Processing Using Apache Flink