Explanation:
In a multiplexed architecture, a single Bronze table typically consolidates various event types to simplify the ingestion process. To ensure downstream Silver or Gold layer processing is efficient, filtering by specific event type columns allows the query optimizer to prune irrelevant data. This reduces unnecessary I/O and focuses computational resources on the specific subset of data required for a given transformation.
Ultimate access to all questions.
No comments yet.
When streaming from a multiplexed Bronze table containing multiple event types, which of the following architectural best practices ensures the most efficient downstream processing and reduced I/O overhead?
A
Configuring a fixed batch size for all streaming micro-batches to stabilize throughput.
B
Utilizing event type columns to filter data subsets before executing downstream logic.
C
Disabling schema evolution in the Bronze table to prevent metadata overhead.
D
Refactoring the ingestion pipeline to write each event type into its own isolated table.