April 2018

Releases are staged. Your Databricks account may not be updated until a week after the initial release date.

Note

We are now providing Databricks Runtime deprecation notices in Databricks Runtime release notes versions and compatibility.

AWS account updates

April 24 - May 1, 2018: Version 2.70

You can now change the AWS account associated with your Databricks account.

See Configure your AWS cross-account IAM role (legacy).

Spark error tips

April 24 - May 1, 2018: Version 2.70

Databricks now provides tips to help you interpret and troubleshoot many of the errors you might see when you run Spark commands. And we’ll keep adding more.

Spark error tips

Databricks CLI 0.7.0

April 24, 2018

Databricks CLI 0.7.0 includes bug fixes.

See Databricks CLI (legacy).

Increase init script output truncation limit

April 24 - May 1, 2018: Version 2.70

We have increased the output truncation limit for init scripts to 500,000 characters.

See What are init scripts?.

Clusters API: added UPSIZE_COMPLETED event type

April 24 - May 1, 2018: Version 2.70

The new UPSIZE_COMPLETED cluster event type indicates that nodes have finished being added to a cluster.

See Clusters API in the Clusters API reference.

Command autocomplete

April 10 - 17, 2018: Version 2.69

Databricks now supports two types of autocomplete in your notebooks: local and server. Local autocomplete completes words that exist in the notebook. Server autocomplete is more powerful because it accesses the cluster for defined types, classes, and objects, as well as SQL database and table names. To activate server autocomplete you must attach your notebook to a running cluster and run all cells that define completable objects.

Notebook autocomplete

Serverless pools upgraded to Databricks Runtime 4.0

April 10, 2018

The serverless pools runtime version has been upgraded from Databricks Runtime 3.5 (which includes Apache Spark 2.2.1) to Databricks Runtime 4.0 (which includes Apache Spark 2.3.0). You must restart your clusters to pick up this change.

The upgrade represents a minor Apache Spark version update and is backwards compatible.

See Compute configuration reference.