Version 26.3.0

Date: 7th March 2026

• Millicore-Level CPU Allocation in Resource Configuration: ADOC now supports configuring Kubernetes CPU requests and limits at the millicore level for Spark drivers and executors while configuring data planes, enabling finer-grained resource control and better optimization in shared environments. For more information, see Resource Recommendations and Auto-Sizing.
• New Snowflake Authentication Methods: ADOC now supports Snowflake Key Pair authentication and External OAuth (Microsoft Entra ID) for Snowflake connections, providing secure alternatives to username/password authentication. For more information, see Snowflake Authentication Methods.
• Templated Persistence Path Configuration: ADOC now allows you to define template-based storage paths for Good, Bad, and Metadata records at both tenant and policy levels. You can organize execution outputs by team, datasource, or time using predefined variables, enabling standardized and flexible storage management across policies. For more information, see Persistence Configuration.
• Native Trino Integration with OpenLineage: ADOC now supports native integration with Trino using OpenLineage, enabling automatic pipeline discovery and query-level lineage for Trino workloads. Each query pattern is modeled as a pipeline, and every execution is tracked as a pipeline run, providing end-to-end observability and enhanced impact analysis for Trino-based analytics workflows. For more information, see Trino Integration with OpenLineage.
• Enhanced Job Filtering Across Job Pages: ADOC now provides additional filtering options across Profile, Data Quality, Reconciliation, Freshness, and Cadence job pages. Users can now filter jobs by Datasource and Asset, along with new filters for Profile Status and Profile Type on the Profile Jobs page, making it easier to locate and monitor specific job executions in large environments. For more information, see Monitoring and Managing Jobs.
• Consistent Column Ordering in Good/Bad Records: ADOC now preserves the column order defined in the SQL SELECT clause for policies created on SQL Views and SQL-based DQ/Reconciliation policies. This ensures consistent column ordering across UI results and downloaded violations for both manual and scheduled executions.
• Multi Asset Selection Enhancement in Asset Export: ADOC now provides a persistent selection list during asset export, allowing users to search, add, and remove multiple assets without losing previously selected items. This enhancement improves usability and ensures a more consistent and efficient bulk export experience.
• Bulk Enable or Disable Policies: ADOC now allows you to enable or disable multiple policies at once, excluding archived policies.

• Resolved an issue where profiling failed in GBQ Pushdown mode for assets in GCP projects with mixed alphanumeric suffixes in the Project ID. Profiling now runs successfully regardless of the project ID naming pattern.
• Resolved an issue affecting anomaly detection timing in certain Data Freshness scenarios, which could result in alerts being generated earlier than expected. Anomaly detection now begins after the configured training window period, ensuring more stable baseline calculation across Change in Size, Absolute Size, and related metrics.
• Resolved an issue where certain policy execution results were not consistently syncing to Alation. Users can now view execution details for all supported policies in the Alation Health dashboard across mapped assets.
• Resolved an issue where the Databricks crawler repeatedly retried tables without the required permissions, leading to prolonged execution times. The crawler now skips inaccessible tables and handles permission errors gracefully to prevent unnecessary retries and extended runtimes.
• Resolved an issue where users with viewing permissions could encounter access denied errors when opening the Reliability Policy Summary page. The page now correctly respects view-level permissions without requiring additional policy creation privileges.
• Resolved an issue where DBT pipelines with a large number of job and asset nodes could intermittently fail to load. Pipelines with complex structures now load more reliably in the ADOC interface.
• Resolved an issue where the Groups section on the Discover Assets page had failed to load in some environments due to query parameter limits. The page now loads correctly and displays groups as expected.
• Resolved an issue where the Data Quality policy name could appear as blank in the UI for certain assets. The policy name is now displayed correctly in the interface.
• Resolved an issue where certain pipeline run APIs could return data across tenants. API responses are now correctly restricted to the requesting tenant’s data, ensuring proper tenant-level isolation.
• Resolved an issue where anomaly detection for Data Freshness policies produced unrealistic expected ranges under certain conditions. Improvements have been made to ensure more accurate baseline calculations, resulting in more reliable anomaly detection.
• Resolved an issue where reconciliation policies could not be edited after being archived due to changes in underlying SQL views. Users can now update or remove invalid rule items and successfully unarchive the policy.
• Resolved an issue where labels were not returned in the response when retrieving reconciliation policies using the API. Labels are now correctly included in the API response when present.
• Resolved an issue where the Evaluation Summary in Reliability alerts showed an incorrect column label for the Data Quality dimension. The column label and values are now displayed correctly in the alert overview.
• Resolved an issue where Kafka crawler runs could fail when encountering malformed or corrupt messages in a topic. The crawler now handles such records more reliably to prevent unexpected failures during crawling.
• Resolved an issue where archived assets created during recrawling could persist and cause the Retrieve Asset Metadata API to return outdated Asset IDs for SAP HANA datasources. Asset resolution now correctly returns the current asset ID, preventing conflicts during API-based operations such as reconciliation policy creation.
• Resolved an issue where tags added to columns were not consistently displayed on the Profile Results page. The system now correctly retrieves and displays column tags without duplicate entries in the child-tags API response.
• Resolved an issue where assets imported using the new Asset Export/Import Manager did not appear in Reliability Explorer reports filtered by tags. Tagged assets now correctly appear in tag-filtered reports after import.
• Resolved an issue where the Analyze Query option under Compute Recommendations for Snowflake long-running queries could open a new window without loading the query analysis details due to timeout conditions. The query analysis page now loads correctly and displays the expected query information.
• Resolved an issue where log streaming in the Kubernetes Dashboard (Skooner) UI could fail in certain Data Plane versions, while log downloads continued to work. Log streaming now functions correctly in the UI, restoring the ability to view pod logs directly from the Data Plane Observability interface.