Let's talk about "Stream Processing with Apache Flink"!


Vasia and Fabian are the authors of the recently published book "Stream Processing with Apache Flink". Meet them at our office talking about their new book and the upcoming roadmap.


18.00 Welcome

18.30 Apache Flink - Latest Features & Roadmap

Apache Flink is a distributed, stateful stream processor and evolving at rapid speed. The community aims to release a new version every four months. The latest release Flink 1.8 was published end of April and brought features like state migration and significantly improved the support for state TTL.
While you are reading this, there is a lot of work going on for the next release.
Recently, the community published a roadmap to share its vision on the future of Flink.
In this talk, I'll discuss the most important features of recent Flink releases and the direction in which Flink is evolving.

19.30 Writing "Stream Processing with Apache Flink"

In mid 2016, we - Vasia and Fabian - decided to write a book about Apache Flink for O'Reilly.
Not sure if we would have committed to this project if we had known that it would take us eight Flink releases (or 2.5 years) to finish it.
In this talk, we take a look back at the journey of writing the book, share the experience that we gained on the way, and tell you some fun stories.

20.30 Teaching stream processing with Apache Flink at ETH Zurich (Vasia Kalvari)

In this talk, I will share my experiences developing and teaching a brand new Master-level stream processing course at ETH Zurich (https://www.systems.ethz.ch/courses/spring2019/dspa). The course was offered to ~40 CS and Data Science students, most of whom were entirely new to stream processing. During the semester, the students used Apache Flink or Timely dataflow to design and develop the backend of a hypothetical online social network. I will discuss some of their surprising design choices, interesting architectures they came up with, favored connectors, as well as their most common challenges and frustrations.

Vasia Kalavri (@vkalavri)
Vasia is a Postdoc at the Systems Group at ETH Zurich (@SystemsGroupETH). She holds a PhD from the KTH Stockholm (@KTHUniversity) and the UC Louvain (@UCLouvain_be). Besides that she is a PMC member of Apache Flink (@ApacheFlink) and the mother of Apache Flinks graph library Gelly.
She is broadly interested in distributed stream processing and large-scale graph analytics.

Fabian Hueske (@fhueske)
Fabian is a PMC Member of Apache Flink (@ApacheFlink) and a software engineer at Ververica (@VervericaData), the former DataArtisans. He is also a co-founder of the DataArtisans.
The focus of his work is on Flink's relational APIs, the Table API and SQL, which are unified APIs for batch and stream processing. he is interested in distributed data processing and query optimization.