databricks Databricks Release Notes ·

Databricks Introduces Session Restore for Serverless Jobs in Public Preview

datadatabrickspreviewengineer
feature announcement

Databricks now offers session restore for serverless compute job runs, allowing users to capture and load Python variables and Spark sessions into new interactive notebooks. This capability helps engineers and data scientists debug failures or explore results without the need to re-execute the entire job. Snapshots are automatically captured for eligible runs, and the feature is currently available in Public Preview.

Features (1)
  • Session Restore for Serverless Job Debugging

    Databricks has released a new capability to restore Python variables and Spark sessions from serverless compute job runs. This allows users to import the state into a new interactive notebook for debugging failures or exploring results, eliminating the need to rerun the full job.

Read the original announcement →

https://docs.databricks.com/aws/en/release-notes/product/2026/september#session-restore-for-serverless-jobs-is-in-public-preview

Related releases