This repository provides a training for Flink's SQL API.
In this training you will learn to:
- run SQL queries on streams.
- use Flink's SQL CLI client.
- perform window aggregations and window joins with SQL queries.
- specify a continuous SQL query that maintain a dynamic result table.
- write the result of streaming SQL queries to Kafka and ElasticSearch.
Please find the training instructions in the Wiki of this repository.
The training is based on Flink's SQL CLI client and uses Docker Compose to setup the training environment.
You only need Docker to run this training.
You don't need Java, Scala, or an IDE.
Apache Flink is a framework and distributed processing engine for stateful computations over unbounded and bounded data streams. Flink has been designed to run in all common cluster environments, perform computations at in-memory speed and at any scale.
Flink features multiple APIs with different levels of abstraction. SQL is supported by Flink as a unified API for batch and stream processing, i.e., queries are executed with the same semantics on unbounded, real-time streams or bounded, recorded streams and produce the same results. SQL on Flink is commonly used to ease the definition of data analytics, data pipelining, and ETL applications.
The following example shows a SQL query that computes the number of departing taxi rides per hour.
SELECT
TUMBLE_START(rowTime, INTERVAL '1' HOUR) AS t,
COUNT(*) AS cnt
FROM TaxiRides
WHERE
isStart
GROUP BY
TUMBLE(rowTime, INTERVAL '1' HOUR)Apache Flink, Flink®, Apache®, the squirrel logo, and the Apache feather logo are either registered trademarks or trademarks of The Apache Software Foundation.