Azure Event Hubs is Azure's managed big-data ingestion service for event streams. Producers send events; consumers read them with consumer groups. It is the Azure analogue people compare to Kafka / Kinesis.
Apps / IoT / CDC
|
v
Event Hubs namespace / hub
|
+--> Stream Analytics
+--> Spark / Databricks / Flink
+--> Capture to ADLS (optional)Key ideas
- Partitions spread throughput (Kafka-like)
- Consumer groups let multiple apps read independently
- Event Hubs Capture can land raw events to ADLS/Blob on a schedule
- Kafka protocol support exists so some Kafka clients can talk to Event Hubs
Vs Kafka (short)
Event Hubs is fully managed Azure PaaS with partition + consumer group concepts. Kafka is the open ecosystem; Event Hubs is the Azure-native front door (or you run Kafka on Azure separately).
Interview tip: "Event Hubs = Azure managed event ingestion with partitions and consumer groups." Mention Capture-to-lake as a classic DE pattern.