Databricks Certified Data Engineer - Associate

Explanation

Option E is the correct answer because it demonstrates a streaming hop from raw data to a Bronze table. Here's why:

Key Characteristics of a Streaming Hop:

1. Reads from raw data source - Uses spark.readStream.load(rawSalesLocation) which reads streaming data from raw files
2. Writes to a table - Uses .table("newSales") to write to a Delta table (Bronze layer)
3. Streaming processing - Uses readStream and writeStream for continuous data ingestion
4. Checkpointing - Includes .option("checkpointLocation", checkpointPath) for fault tolerance
5. Append mode - Uses .outputMode("append") appropriate for streaming ingestion

Why Other Options Are Incorrect:

• Option A, B, C: These read from an existing table (spark.table("sales")) rather than raw data, so they're not performing the initial ingestion from raw to Bronze.
• Option D: This uses batch processing (spark.read.load) instead of streaming, so it's not a streaming hop.

Bronze Table Context:

In Databricks Medallion Architecture:

• Bronze tables store raw, unprocessed data as-is from source systems
• Streaming hops are used for continuous ingestion from raw data sources
• The Bronze layer serves as the landing zone for all incoming data before any transformations

Option E correctly implements this pattern by streaming data from raw files directly into a Bronze table.