Quick Answer

The correct answer is to increase maxFailedItems to a higher value, such as 100. This configuration directly controls Azure AI Search indexer failure tolerance: when the indexer encounters documents with errors, it tracks the cumulative count of failed items, and once that count exceeds the maxFailedItems threshold, the entire indexing job stops. In the given scenario, the indexer fails after processing 6 documents with errors because the default or configured maxFailedItems is set to 5, meaning the fifth failure triggers a halt. On the Microsoft Azure AI Engineer Associate AI-102 exam, this concept tests your understanding of indexer resilience and error handling, often appearing in scenario-based questions where you must distinguish between tuning failure thresholds versus adjusting batch size or scheduling. A common trap is confusing maxFailedItems with batch size—increasing batch size can actually introduce more failures per batch, while decreasing it may reduce errors but does not raise the failure tolerance ceiling. Memory tip: think of maxFailedItems as a “patience counter”—the indexer will keep working until it runs out of patience, so raise the number to let it push through minor errors.