Databricks Certified Data Engineer - Associate

Explanation

Option D is the correct answer because:

1. Streaming Source: It uses spark.readStream.load(rawSalesLocation) which indicates reading from a streaming source (raw data).
2. Streaming Sink: It uses .writeStream to write to a table, making it a true streaming pipeline.
3. Bronze Table Ingestion: This pattern represents the typical first hop in a medallion architecture where raw data is ingested into a Bronze table.
4. Checkpointing: It includes checkpointing for fault tolerance, which is essential for production streaming pipelines.

Why other options are incorrect:

• Option A: Starts with spark.table("sales") - this reads from an existing table, not raw data. It's processing an already ingested table.
• Option B: Uses spark.read.load() (batch read) and .write (batch write) - no streaming involved at all.
• Option C: Similar to Option A, starts with spark.table("sales") - processing existing table data, not raw streaming data.

Key Concept: A "streaming hop from raw data to a Bronze table" specifically requires reading from a streaming source (like files arriving in cloud storage, Kafka, etc.) and writing to a table using streaming APIs. This is the foundation of incremental data processing in Databricks.