Скачать или смотреть PyFlink: Building Flink applications with Python language

The Apache Flink community has started to add Python language support (PyFlink) since two years ago to ease the lives of Python users. In the last few releases, a lot of features were added in PyFlink. It has provided comprehensive support for Python Table API & SQL and enabled users to use all kinds of Python user-defined functions such as Python UDF, Python UDTF, Python UDAF, Pandas UDF, Pandas UDAF, etc in Table API & SQL programs. Besides, it has also supported Python DataStream API which gives users lower-level control over the core building blocks (e.g. state, time, etc) of Flink to build more complex stream processing use cases in Python language. In this talk, we will firstly give a basic introduction of PyFlink. Then we will go through the new features and show what is possible to achieve with PyFlink. Finally, we will share some performance data of PyFlink.
0:00 Pyflink Overview
4:10 Pyflink Internals
16:23 Demo
32:46 Future of Pyflink