March 2025 changelog

Here you'll find the March 2025 changelog for the Data Productivity Cloud. Just want to read about new features? Read our New Features blog.

27th March

DesignerNew features 🎉

Schedules:

Added a new Allow concurrent schedule runs option when creating a schedule that lets you choose whether a scheduled pipeline run is skipped if the previous scheduled run is still in progress.

DesignerNew features 🎉

Connectors:

StreamingImprovements 🔧

Updated the Streaming agent version from 2.110.4 to 2.111.6. This is the minimum agent version required for the new snapshot feature.
New snapshot feature introduces recoverability and flexibility. This replaces initial and on-demand snapshot options.
New dashboard layout makes it easier to navigate to relevant information.
Upgrade of the IBMi DB2 JDBC driver in the Agent.
Library updates and security fixes.

DesignerNew features 🎉

Connectors:

DesignerNew features 🎉

AI:

Added an ML Classification component, which uses the Snowflake Classification machine learning (ML) function to sort data into different classes using patterns detected in training data. This component is available in public preview.

DesignerNew features 🎉

New and updated roles and permissions for projects and environments have been introduced: Owner, Contributor, Viewer, and None, replacing Admin, User, Read Only, and No Access, respectively. Additionally, a new environment role, Runner, has been introduced. All roles and permissions are documented in the following.

New connector:

X Ads, which lets you query the X Ads (formerly Twitter Ads) API to retrieve data and load it into a table. You can then use transformation components to enrich and manage the data in permanent tables.

Copilot:

Copilot can now write your Git Commit messages for you using the Conventional Commits format.

DesignerNew features 🎉

AI components:

Add a Cortex Multi Prompt component, which uses Snowflake Cortex to receive a prompt and then generate multiple responses (completions) using your chosen supported language model. This component is available in public preview.

CDCImprovements 🔧

Updates to the CDC agent version.
- From 2.108.1 to 2.110.3.
- From 2.110.3 to 2.110.4.
Reverted the Snowflake JDBC driver to 3.22.0 to mitigate an issue where the first Snowflake request on an agent could fail.
When using a storage target or Snowflake with the "Source Database & Schema" prefix, setting the database value is now Root where previously the Db2 system name would have been used.
Library updates and security fixes.

DesignerImprovements 🔧

System variables and post-processing are now generally available:
- System Variables can now be used in component parameters, Python, Bash, and SQL scripts, in addition to post-processing.
- Users can access shared and public scalar variables from a child pipeline using the post-processing Update Scalar property in Run Transformation, Run Orchestration, and Run Shared Pipeline components.
- System variables are read-only and cannot be modified manually. Instead, system variable values are updated automatically during component executions.
- System variables can be directly used in Python Pushdown, manually assigned in Bash Pushdown, and used inline in SQL script (orchestration) and SQL (transformation) components, with mapping required for external SQL.
- New child pipeline system variable syntax is ${sysvar.childPipeline.vars.<varname>} and allows referencing variables from child pipelines, e.g., ${sysvar.childPipeline.vars.var1}.
Data lineage now includes orchestration pipelines, providing a comprehensive view of data origins. This expansion introduces support for additional connectors and offers key benefits such as audit and compliance, impact analysis, and faster debugging.