Comprehensive and Detailed Explanation From Exact Extract of Databricks Data Engineer Documents:
When writing data to Delta Lake tables, the option mergeSchema = true allows automatic evolution of the table schema by merging new columns or data types introduced in the incoming data with the existing table definition. This ensures pipelines remain resilient to upstream schema changes, especially in append-only or streaming ingestion scenarios. The Databricks documentation specifies that this option should be enabled for schema-on-write flexibility, while overwriteSchema replaces the existing schema entirely and ignoreChanges applies only to CDC operations. validateSchema ensures consistency and does not handle evolution. Therefore, B (mergeSchema = true) is the correct configuration for handling automatic schema evolution safely.