Databricks tables

Databricks supports multiple table types and storage formats to meet different data management needs. For an overview of table types, storage formats, and Unity Catalog integration, see Databricks tables concepts.

Table types

Explore different table types and their capabilities for various data management scenarios.

• • Unity Catalog managed tables in Databricks for Delta Lake and Apache Iceberg
  • Databricks manages metadata and data files for new tables that require optimized performance.
• • Temporary tables
  • Session-scoped Unity Catalog managed tables for intermediate data. SQL warehouses only.
• • Work with external tables
  • Data stored in external systems. Unity Catalog manages metadata only.
• • Work with foreign tables
  • Read-only access to data in external systems connected through Lakehouse Federation.

Storage formats

Work with open table formats that support advanced data management capabilities.

• • Delta Lake
  • Default storage format with ACID transactions, time travel, and schema enforcement for managed and external tables.
• • Apache Iceberg
  • Open table format for integration with the Iceberg ecosystem, supporting advanced metadata management.

Table management

Configure and optimize table behavior, structure, and performance.

• • Table constraints
  • Define and enforce data quality rules with check constraints and not null constraints.
• • Schema enforcement
  • Control how Databricks handles schema changes and data type enforcement during writes.
• • Table partitioning
  • Organize data by partition keys to improve query performance and data management.
• • Table size monitoring
  • Monitor and analyze table storage usage and growth patterns.
• • Convert external to managed
  • Migrate external tables to managed tables for improved performance and management.
• • External partition discovery
  • Automatically discover and register partitions in external tables stored in cloud storage.