A company has deployed a data pipeline that streams customer transaction data from an on- premises PostgreSQL database to Snowflake for real-time analytics. A Data Engineer needs to configure a…

Question

Options

• AConfigure the Snowflake Connector for Kafka to continuously stream data from the PostgreSQL
• BDeploy the Snowflake JDBC driver to connect the PostgreSQL database directly to Snowflake,
• CInstall the Snowflake Connector for Python on the PostgreSQL server to stream data directly to
• DConfigure the Snowflake Connector for Kafka on a Kafka Connect cluster to stream data from the

• A
  19% (6)
• B
  3% (1)
• C
  9% (3)
• D
  69% (22)

Explanation

Option D is correct because deploying the Snowflake Connector for Kafka on a Kafka Connect cluster provides the complete, production-grade streaming architecture. Kafka Connect manages distributed processing, fault tolerance, offset tracking, and automatic restarts - all essential for a continuous, reliable pipeline with minimal latency into Snowflake.

Option A is a trap: it mentions the Snowflake Kafka Connector but omits the critical "Kafka Connect cluster" component. Without the cluster infrastructure, you lose scalability, fault tolerance, and proper connector lifecycle management - making it unreliable for production streaming.

Option B (JDBC driver) is a batch/query mechanism, not a streaming solution. It introduces significant latency and puts load directly on the PostgreSQL server, making it unsuitable for real-time analytics requirements.

Option C (Python connector on the PostgreSQL server) conflates the database server with the integration layer. Running custom Python scripts on the source DB server is resource-intensive, not fault-tolerant, and not designed for scalable, continuous streaming pipelines.

Memory tip: The key phrase is "Kafka Connect cluster" - the connector is just a plugin; the cluster is what provides reliability and scalability. On exams, watch for answers that name the right tool but strip away the infrastructure that makes it production-ready (A vs. D).

No community discussion yet for this question.