Flink is a streaming dataflow execution engine that handles data distribution, communication, and fault tolerance for distributed stream processing. It also provides higher-level APIs for writing distributed tasks.
Open source compatibility
EMR Flink is fully compatible with open source Flink. Key community resources:
Scenarios
Flink supports a wide range of real-time big data scenarios, spanning both technical and enterprise applications.
-
Technical areas
Flink covers the following technical scenarios:
-
Real-time ETL and data streams
Real-time ETL and data streams deliver data between systems with inline cleansing and integration. Examples: real-time search engine indexing and data warehouse ETL pipelines.
-
Real-time data analytics
Real-time data analytics extracts actionable insights from raw data as it arrives. Examples: daily top-10 product rankings, warehouse turnover time, click rates, and push open rates. Results typically appear on real-time dashboards and reports.
-
Event-driven applications
Event-driven applications process subscribed events using internal state. Examples: fraud detection, risk control, and O&M anomaly detection. When a risk control rule is triggered, the system analyzes current and historical behavior to decide whether action is needed.
-
-
Enterprise applications
From an enterprise perspective, common Flink applications include:
-
Business: real-time risk control, real-time recommendations, and search engine indexing.
-
Data: real-time data warehouses, real-time reports, and real-time dashboards.
-
O&M: real-time monitoring, real-time anomaly detection and alerting, and end-to-end debugging.
-