Answer: tumbling window

## Detailed Explanation ### Requirements Analysis Based on the scenario, we need a trigger that supports: 1. **Automatic retry capability** when pipeline runs fail due to concurrency or throttling limits 2. **Backfilling support** for existing data in the table 3. **Scheduled execution** every four hours ### Trigger Type Evaluation **D. Tumbling Window Trigger** - **CORRECT** - **Retry Capability**: Tumbling window triggers have built-in retry policies that automatically retry failed pipeline runs due to concurrency limits, throttling limits (status codes 429), and server errors (status codes 500). - **Backfilling Support**: This is a key feature of tumbling window triggers. They support processing historical data by allowing you to trigger pipeline runs for past time windows, which is essential for backfilling existing data. - **Scheduled Execution**: Tumbling window triggers can be configured to run at regular intervals (every four hours in this case) with precise window-based scheduling. **Why Other Options Are Less Suitable:** **A. Event Trigger** - Incorrect - Event triggers respond to storage events (blob creation/deletion) rather than scheduled intervals - No inherent support for backfilling historical data - Limited retry capabilities compared to tumbling window triggers **B. On-demand Trigger** - Incorrect - Manual execution only, no automated scheduling - No automatic retry mechanisms - No support for backfilling without manual intervention **C. Schedule Trigger** - Incorrect - While schedule triggers support regular execution intervals, they lack the advanced backfilling capabilities of tumbling window triggers - Limited retry policy options compared to tumbling window triggers - Cannot easily process historical data windows for backfilling scenarios ### Key Differentiators - **Backfilling Requirement**: This is the critical factor that makes tumbling window trigger the optimal choice. Tumbling window triggers are specifically designed to handle time-series data processing with support for processing past time windows. - **Advanced Retry Policies**: Tumbling window triggers provide more sophisticated retry configurations for handling concurrency and throttling scenarios compared to basic schedule triggers. - **Stateful Execution**: Tumbling window triggers maintain state about which windows have been processed, making them ideal for incremental load scenarios with historical data processing requirements.

Author: LeetQuiz Editorial Team