This post also appears as a guest post at the Confluent blog.
Stream processing is becoming very popular with open source projects like Apache Kafka, Apache Samza, Apache Storm, Apache Spark’s Streaming module, as well as cloud services like Google Cloud Dataflow.
In this post, we introduce Apache Flink™ as a streaming system. Flink fills what we perceive as an open gap in the space of streaming systems, by providing a combination of
- a proper foundation with a low-latency stream processor,
- expressive APIs that enable programmers to quickly develop streaming data applications,
- flexible operator state and streaming windows, and
- efficient fault tolerance including streams and state.