databricks Platform · Updated Jun 12, 2026

Auto Time-to-Live for Row Deletion

Automate data lifecycle management with configurable row expiration.

What it does

Auto time-to-live (auto-TTL) automatically deletes rows from Unity Catalog managed Delta Lake tables, managed Apache Iceberg tables, and streaming tables based on a specified timestamp column and a defined expiration period in days. The system performs background operations such as DELETE, PURGE, and VACUUM to manage the removal of expired rows and clear them from storage. This feature helps maintain data hygiene and optimize storage costs by ensuring that outdated data is automatically purged.

Why it matters

This feature significantly enhances data management efficiency by automating the deletion of stale data, which can reduce storage costs and improve query performance. Organizations can benefit from streamlined data lifecycle management, ensuring compliance with data retention policies while minimizing manual intervention. Use cases include managing data retention for regulatory compliance and optimizing storage for analytical workloads.

Key capabilities

Configurable Expiration Period

Users can define a specific number of days after which rows are eligible for deletion based on a timestamp column.

Background Deletion Operations

Databricks automatically executes DELETE, PURGE, and VACUUM commands to manage expired rows without user intervention.

Support for Multiple Table Types

Auto-TTL is compatible with Unity Catalog managed Delta Lake tables, Apache Iceberg tables, and streaming tables.

Monitoring and Alerts

Users can monitor auto-TTL operations and set alerts for failures using system tables.

Prerequisites

Predictive optimization must be enabled.

User must have MODIFY permissions on the table.

Databricks Runtime 17.3 or above is required.

Recommended use cases

Automatically delete data older than a specified period to reduce storage costs.Manage data retention for compliance with data governance policies.Remove data marked for deletion by other business processes.

Limitations · remember

!Exact deletion timing is not guaranteed and may vary based on system load.

!Auto time-to-live is not supported on materialized views.

!Modifying auto time-to-live on streaming tables requires updating the pipeline code.

Architecture

Unity Catalog → Delta Lake → Apache Iceberg → Streaming Tables → Predictive Optimization

Get started

Enable Predictive Optimization

Turn on predictive optimization for Unity Catalog managed tables.

Set Auto-TTL Policy

Use SQL commands to define an expiration period and timestamp column for the table.

Monitor Deletion Operations

Query system tables to verify auto-TTL events and monitor performance.

Set Alerts for Failures

Create alerts in Databricks SQL to notify users of any auto-TTL operation failures.

Review Table Properties

Use DESCRIBE TABLE EXTENDED to confirm that auto-TTL is configured correctly.

Key takeaway

Auto time-to-live automates data deletion, optimizing storage and ensuring compliance with retention policies.

2026-06-12-auto-time-to-live-for-automatic-row-deletion-is-generally-available Generated by the Databricks Updates Agent · browse all updates