#####   ######  #####    ###    #   #  ###  #   #  ######
##  ##  ##      ##  ##  ## ##   #   #   #   #   #  ##
#####   ####    #####   #   #   #   #   #   #   #  ####
##  #   ##      ##      ## ##    # #    #    # #   ##
##   #  ######  ##       ###      #    ###    #    ######

$ curl repovive.com/roadmaps/data-engineer-interview-prep/streaming-and-real-time-data/stream-processing-frameworks

░░░░░░░░░░░░░░░░░░░░░░░░░░░░░░░░░░░░░░░░░░░░░░░░░░░░░░░░░░░░░░░░░░░░░░░░░░░░░░░░░░░░░░░░░░░░░░░░░░░░░░░░░░░░░░░░░░░███████████████████████████████████████████████████████████████████████████████████████████████████████████████████

#####   ######  #####    ###    #   #  ###  #   #  ######
##  ##  ##      ##  ##  ## ##   #   #   #   #   #  ##
#####   ####    #####   #   #   #   #   #   #   #  ####
##  #   ##      ##      ## ##    # #    #    # #   ##
##   #  ######  ##       ###      #    ###    #    ######

$ curl repovive.com/roadmaps/data-engineer-interview-prep/streaming-and-real-time-data/stream-processing-frameworks

Repovive

Stream Processing Frameworks - Streaming and Real-Time Data | Data Engineer Interview Prep | Repovive

Repovive.

Data Engineer Interview Prep16 sections · 140 units8h total

Open in Course

Stream Processing Frameworks

Spark Streaming, Flink, and more

Spark Structured Streaming: Micro-batch approach. Treats streams as unbounded tables. Familiar Spark APIs.

df = spark.readStream.format('kafka').load()
result = df.groupBy(window('timestamp', '5 minutes')).count()

Apache Flink: True streaming. Lower latency than Spark. Complex event processing. Steeper learning curve.

Kafka Streams: Library, not framework. Runs in your application. Good for simple transformations.

For interviews, know Spark Streaming basics. Flink knowledge is a bonus for senior roles. Understand windowing concepts regardless of framework.