Deprecations and migrations¶
Deprecation notices are included within the Cloud announcement pages. The pages within this section provide more detailed guidance related to deprecated features. When applicable, a guide that explains the process for migrating your workflows to support the new functionality is provided.
Important update on credential management¶
To ensure that DataRobot is providing a secure and reliable platform, we are enhancing security measures so that they are in compliance with the latest security standards.
Specifically, all users must review credentials stored in the credential manager, ensuring that they meet the following requirements:
-
Any password must encode to at least 14 bytes (112 bits), meaning it should be at least 14 characters long.
-
Any user-supplied private keys must be FIPS compliant. Compliance requires a FIPS-approved cipher and a salt length of at least 16 bytes (128 bits).
Impacted credential types include, but are not limited to, basic credentials and Snowflake (keypair) credentials. All necessary changes must be applied by July 1, 2025.
Deprecations and retirements¶
See the following pages for more information. For GenAI functionality, see the section on LLM availability.
Deprecation/migration | Description |
---|---|
MLOps Library Java version requirement | As of March 2025 (and DataRobot v11.0), the MLOps monitoring library requires Java 11 or higher. This requirement also applies when using the MLOps library to instrument monitoring with Scoring Code. |
Capability matrix | A comparison of the capabilities available in DataRobot Classic and NextGen; maintenance ended January 2025. |
End of support calendar | A list of end-of-support dates for the Self-Managed AI Platform. |
Prediction server migration | Guidance for migrating models from dedicated to serverless prediction environments |
Custom model training data assignment update | Guidance for migrating models from the "per model" training data assignment method to the "per version" method. |
Model Registry workflow update | Guidance for adapting to the Model Registry updates introduced to the DataRobot Classic UI between the October and November 2023 AI Platform releases (and in the DataRobot v9.2 release). |
In-app documentation | Information related to the removal of the in-app documentation, directing all users to the public documentation site. |
Python 2 deprecation/migration guide | Information, action-items, and important dates related to the deprecation of Python 2 and migration to Python 3. |
Batch scoring script deprecation | Guidance for moving from Batch Scoring Script use to Batch Prediction Script use. |
Open source and USER models deprecation | As of November 2022, DataRobot disabled all models containing User/Open source (“user”) tasks. |
Pricing 5.0 (Classic MLOps) | Information related to Pricing 5.0, the original DataRobot MLOps pricing plan, introducing several MLOps capabilities. |
Alteryx deprecation in DataRobot | Guidance for the use of Alteryx predictive tools outside DataRobot. |
Accuracy over time data storage | Information related to the migration of "accuracy over time" data storage from MongoDB to S3. |