Introduction to Apache Flink

For our first meetup we will start with the basics. Fokko Driesprong will give us an introduction to Apache Flink by creating a simple Flink DAG which ingests an endless stream of data. We will apply Flink's stream processing capability to extract knowledge from the underlying data.

After Zhenhao, will give us a live demo which will dive into Flink's event time handling. The demo is based on Jamie Grier's demo at OSCON and his code on github, https://github.com/dataArtisans/oscon. During the demo he'll build a toy realtime analytics application. We'll demonstrate why event time is such an important concept in robust stateful stream processing and talk about key features of Flink that is crucial to run in production, such as fault tolerance and easy continuous integration.

About Apache Flink:

Apache Flink® is an open source platform for distributed stream and batch data processing. It contrasts itself from Apache Spark by considering Streaming First. Providing high throughput and low latency stream processing with exactly-once guarantees. Moreover, batch processing applications run efficiently as special cases of stream processing applications.

About Fokko:

Fokko Driesprong studied Distributed Systems at the University of Groningen, and working with GoDataDriven as a data engineer.

About Zhenhao:

Zhenhao is a data scientist and big data solution architect at Accenture NL. He has helped major clients building real time big data analytics systems, and successfully used Flink in production. He gave a talk not long ago at London Flink meetup based on his work for a big UK client. Zhenhao previously worked as a PhD candidate in mathematical logic at the University of Amsterdam.