Release Notes#


Warning

Support for Python versions below 3.10 has been dropped as of the Streamsets SDK (7.0.1) release.

7.0.1 (April 2026)#

New Features

  • Python Code Generator - Added streamsets.sdk.codegen.generator.PythonGenerator to convert existing pipelines to Python SDK code. Supports pipeline definitions from .zip archives, .json files, and streamsets.sdk.sch_models.Pipeline objects. See build/code_generation:Code Generation for details.

  • HTTP Retry Management - Introduced comprehensive HTTP retry mechanism with exponential backoff and jitter. The new streamsets.sdk.retry module provides configurable retry policies for different HTTP status codes, with special handling for rate limiting (429). See manage/http_retry_management:HTTP Retry Management for configuration details.

  • Engine Type Standardization - Centralized all engine type references into streamsets.sdk.utils.EngineType enumeration. Engine types are now consistently referenced as COLLECTOR, TRANSFORMER, and SNOWPARK across the SDK. See usage/engine_types:A Note on Engine Types for migration guidance.

Breaking Changes

  • Dropped Python 3.8 and 3.9 Support - Python versions below 3.10 are no longer supported. Please upgrade to Python 3.10 or higher before upgrading to SDK 7.0.1.

  • Removed Deprecated Features - This major release removes all features that were previously marked as deprecated. Please update your code to use the replacement features before upgrading.

Removed API Methods (sch_api.py):

  • Removed get_all_registered_executor_labels() - Use get_all_registered_engine_labels() instead

  • Removed get_all_registered_executor_versions() - Use get_all_registered_engine_versions() instead

  • Removed get_all_executor_stats() - Use get_all_engine_stats() instead

  • Removed get_executor_cpu_usage_time_series() - Use get_engine_cpu_usage_time_series() instead

  • Removed get_executor_memory_usage_time_series() - Use get_engine_memory_usage_time_series() instead

Removed Model Properties (sch_models.py):

  • Removed User.roles - Use User.organization_roles instead

  • Removed Group.group_id - Use Group.id instead

  • Removed Group.roles - Use Group.organization_roles instead

  • Removed Pipeline.pipeline_id - Use Pipeline.id instead

  • Removed Pipeline.executor_type - Use Pipeline.engine_type instead

  • Removed Pipeline.sdc_id - Use Pipeline.engine_id instead

  • Removed Job.job_id - Use Job.id instead

  • Removed Job.executor_type - Use Job.engine_type instead

  • Removed Topology.topology_id - Use Topology.id instead

  • Removed Environment.environment_id - Use Environment.id instead

  • Removed Deployment.deployment_id - Use Deployment.id instead

  • Removed Deployment.engine_instances - Use Deployment.desired_instances instead

Removed ControlHub Methods (sch.py):

  • Removed ControlHub.publish_scheduled_task() - Use ControlHub.add_scheduled_task() instead

Migration Guide:

All removed features have direct replacements. Simply update your code to use the new method/property names as indicated above. The functionality remains identical, only the names have changed to provide better consistency across the SDK.

History#

6.6.2 (August 2025)#

  • Fixed race conditions when retrieving objects

  • Bug fixes and improvements

6.6.1 (June 2025)#

6.6.0 (June 2025)#

6.5.0 (November 2024)#

  • The streamsets.sdk.sch_models.Step class now supports Finish Conditions and has a new attribute name which will allow retrieving and setting the name of a step.

  • Added support for running an arbitrary step using the The streamsets.sdk.ControlHub.run_job_sequence() method.

  • Extended AQL functionality to now support searching on streamsets.sdk.sch_models.JobSequence objects.

  • Updated API calls to be more efficient by changing the default page size to 250 instead of 50.

  • Added SDK support for creating connections without an authoring engine specified (engineless connections).

  • Added SDK support for stopping a Deployment through the streamsets.sdk.ControlHub.delete_deployment() method.

  • Added the ability to specify the install_type for a self-managed deployment when retrieving the install script.

  • Bug fixes and improvements.

6.4.0 (July 2024)#

6.3.0 (May 2024)#

6.2.0 (February 2024)#

6.1.0 (November 2023)#

6.0.1 (September 2023)#

6.0.0 (August 2023)#

5.2.1 (May 2023)#

  • Fixed a TypeError Exception when filtering jobs by status in ControlHub.

5.2.0 (April 2023)#

5.1.0 (December 2022)#

  • Support for StreamSets Advanced Query Language has been added for Pipelines, Fragments, and Jobs. Refer to the StreamSets SDK Search Documentation for further details.

  • The streamsets.sdk.ControlHub.engines.get_all() method now supports filtering by the id parameter.

  • Bug fixes and improvements.

5.0.0 (August 2022)#

4.3.0 (August 2022)#

4.2.1 (July 2022)#

4.2.0 (May 2022)#

  • Programmatic User creation and management has been added

  • Pagination and “lazy” loading improvements have been made to various classes

  • The Group class has been refactored slightly to better match the experience seen in the UI

Note

When filtering the streamsets.sdk.sch_models.Groups objects in StreamSets Platform, the id argument has been replaced by group_id to match the streamsets.sdk.sch_models.Group class’s representation. Please refer to the documentation for the correct, updated usage.

Warning

The deployment_type and environment_type arguments are deprecated and will be removed in a future release. Please refer to the documentation for the correct, updated usage.

Warning

The id argument has been deprecated and will be removed in a future release. Please refer to the documentation for the correct, updated usage.

4.1.0 (March 2022)#

  • Modified error handling to return all errors returned by an API call to StreamSets Platform

  • Transformer for Snowflake support

  • Support for nightly builds of execution engines

4.0.0 (January 2022)#

  • Activation key is no longer required

  • DataCollector and Transformer classes are no longer public because these are headless engines in StreamSets Platform

  • Authentication is now handled using API Credentials

  • The usage and syntax for PipelineBuilder has been updated

  • Support for environments and deployments