DE Mastery GitHub

Apache Flink · FLK-02

DataStream API

From your world. a dbt staging model (filter + enrich + route), but as a continuous DataStream job

Practice quiz with instant feedback. For a scored run that counts toward your local progress, use uv run de quiz FLK-02.

1Which DataStream operator hash-partitions the stream, so that all records for a key go to the same subtask? You need it before keyed aggregations, state and windows.
2Good records must flow downstream. Invalid records must go to a dead-letter stream. This must happen in a single pass. Which mechanism does this?
3Why must an OutputTag be created as an anonymous subclass, e.g. new OutputTag("rejects"){} ?
4In Flink 2.x, how is a Kafka source added to a job?
5You must merge two streams of DIFFERENT types and process them together with shared state. Which operator do you use?
6Which transformation can emit zero, one or many output records for each input record?
7When should you prefer Flink SQL over the DataStream API?
8What does stream.sinkTo(sink) attach in Flink 2.x?