Flink
Distributed stream processing engine for computing over continuous, unbounded data streams with exactly-once correctness guarantees.
- Actively maintained
- Foundation-backed
- Well documented
- Permissive license
Overview
Apache Flink processes data as continuous streams rather than fixed batches, computing aggregations, joins, and pattern detection over events as they arrive with low latency and exactly-once processing guarantees even after failures. It also supports batch processing as a special case of streaming.
Flink is a fit for teams building real-time data pipelines, fraud detection, or event-driven analytics that need to process data as it arrives rather than in scheduled batches, commonly paired with Apache Kafka (also in this catalog) as the event source.
Why it's in this catalog
Catalog rating: ★★★★☆ (4/5)
Leading stream-processing engine, a strong representative pick for real-time analytics.
- Foundational
- Foundation-governed
- Categories
- Analytics
- Languages
- Java
- License
- Apache-2.0
Related Projects
Spark
Apache-2.0Unified analytics engine for large-scale batch and stream data processing, one of the most widely adopted big data frameworks.
Actively maintained
- Analytics
ClickHouse
Apache-2.0Column-oriented database built for real-time analytical queries over massive datasets, generating aggregated reports in milliseconds rather than seconds or minutes.
Actively maintained
- Analytics
Beam
Apache-2.0Unified programming model for defining batch and stream data processing pipelines that can run portably across multiple execution engines.
Actively maintained
- Analytics
Druid
Apache-2.0Real-time analytics database built for sub-second queries over continuously streaming and historical event data at scale.
Actively maintained
- Analytics
Spotted an error? Suggest an edit on GitHub.