Release 2025.06

Release versions and dates

Collibra Data Lineage Service

• When ingesting Redshift data sources, Collibra Data Lineage now offers improved handling of INSERT statements when no data models are provided and subsequent queries refer to the inserted tables.
• When ingesting Teradata data sources, Collibra Data Lineage now supports column aliasing in WHERE clauses.
• Fixed an issue where certain error cases caused source code highlights in the DATA_FLOW_COLUMNS table to show incorrectly, instead of showing proper syntax highlighting.
• When ingesting OpenLineage:
  • The “schema” value is now used as the default schema, when it is not explicit.
  • The “database” value is now used for database mapping.

Lineage Harvester (CLI and Edge)

• When integrating Tableau data sources, the lineage harvester no longer includes invalid source configurations in the source ZIP files.
• When the lineage harvester invokes Azure Data Factory endpoints, it now sanitizes the request URLs to optimize harvesting.
• When integrating MicroStrategy data sources, the lineage harvester now properly terminates its session and logs out of the data source upon completing its operations.

Collibra Data Lineage Service

• We fixed a bug that caused the "Browse" tree in the technical lineage viewer to generate twice. As a result, the "Browse" tree now loads faster.
• We optimized the synchronization process to better support large technical lineage graphs.
• When ingesting Microsoft SQL Server data sources, Collibra Data Lineage now correctly handles Common Table Expressions (CTEs) that reference tables with the same name.
• When analyzing Snowflake procedures, Collibra Data Lineage now supports Data Definition Language (DDL) operations in RESULTSET queries.
• When ingesting Snowflake data sources, Collibra Data Lineage now correctly processes SQL queries with trailing commas in FROM clauses.
• When integrating Power BI data sources, metadata processing no longer fails when encountering:
  • A "ResultUnknown" object without a “value” attribute.
  • Dynamic query constructions that use the Power Query M function “Value.NativeQuery”.
• When ingesting PostgreSQL data sources, Collibra Data Lineage now supports SELECT INTO variables.

Security

• We improved the security of Edge and Collibra Cloud sites.
• We improved the security of capabilities via Edge and Collibra Cloud sites.

Collibra Data Lineage Service

• When integrating Power BI, if the Power Query M function “SapHana.Database” is used, the query passed as a parameter is now correctly sent to the SQL extension with “hana” as the dialect. Previously, “mssql” dialect was used, which is not compatible with SAP HANA's syntax.
• When ingesting DataStage data sources:
  • Collibra Data Lineage now retrieves column names from the stage definition to ensure accuracy and prevent incorrect values caused by character casing discrepancies.
  • Column look-up for SQL columns is no longer case-sensitive.
  • Collibra Data Lineage now correctly handles queries where the “AS” keyword is omitted, automatically inferring and inserting it as needed to ensure accurate query parsing and improved processing.
• When ingesting Microsoft SQL Server data sources, Collibra Data Lineage now:
  • Correctly handles WITH statements used in MERGE statements.
  • Generates correct lineage for UPDATE statements that have Common Table Expressions (CTEs) as targets.
• When ingesting Snowflake data sources,Collibra Data Lineage now supports the ARRAY_MIN() and ARRAY_MAX() functions.

Security

• We improved the security of Edge and Collibra Cloud sites.

Metadata integrations

• Integration synchronization errors are now handled more effectively, ensuring that jobs no longer get stuck during synchronization.
• The Databricks Unity Catalog ingestion process no longer creates duplicate Primary Key assets. These duplicates occurred when multiple foreign keys referenced the same primary key in a parent table. Schema names are now accurate, and Primary Key assets are created correctly.

Lineage Harvester (CLI and Edge)

• You can now export technical lineage via the Technical Lineage Admin Edge connection and capability. The result is a ZIP file that contains JSON lineage files and transformations, formatted as described in Custom technical lineage JSON file details (batch definition). You can download the ZIP file from the Activities results on your profile page. This feature is in preview.
  
  Note This command is not dependent on or specific to custom technical lineage. You can export the lineage of any supported data source. The reference to custom technical lineage is only to express the formatting of the exported files.

Collibra Data Lineage Service

• When ingesting metadata via the OpenLineage Edge integration, Collibra Data Lineage now provides clearer explanations of naming violations, including the specific values that cause the issues.
• We improved memory management in the Power BI integration to avoid issues caused by memory overflow.
• When an Informatica Intelligence Cloud Services (IICS) expression references a variable, Collibra Data Lineage now creates a lineage from the variable to the output.
• When ingesting Power BI reports, you can use the new "createAppReports" filtering parameter to control when an App Report object is created. For example, if you set this parameter to "false", an "[App] Report name" object won't be created for reports that are a part of Apps.
• When you ingest Snowflake queries that use a function as a source, Collibra Data Lineage now analyzes and includes these sources in the technical lineage.
• When processing DataStage metadata, DataStage jobs that don’t use containers no longer cause errors.

Data Quality & Observability

• The DQ Pushdown Edge capability is now upgraded to version 1.48.0 to ensure compatibility with Collibra Platform version 2025.08.

Protect

• Policies created in Snowflake now include the names of the associated standards or rules, making them easier to identify.

Lineage Harvester (CLI and Edge)

• When integrating Tableau data sources, Collibra Data Lineage now handles scenarios where the Tableau API returns an unexpected 400 Bad Request response instead of the expected 401 Unauthorized response when a token expires. This improvement helps to ensure seamless token refresh and an uninterrupted synchronization processes.

Metadata integrations

• We improved the DQ Pushdown Edge capability for a Data Quality & Observability private preview. No action is required.

Infrastructure

• We now support the following versions for our managed Kubernetes clusters:
  • AKS 1.33
  • AWS Fargate using EKS 1.33
  • EKS 1.33
  • GKE 1.33
  • OpenShift 4.19
• New edge sites installed on bundled k3s are now installed on k3s 1.33. No action is required.

Metadata integrations

• Mapping Google Dataplex labels to attributes in Data Catalog now respects each attribute’s max cardinality.
• The Databricks Unity Catalog integration has been updated to ensure all data is ingested from data sources.
• When you synchronize an AWS Bedrock AI capability, the integration now ingests AI models configured in us-west-1 region.
• To allow profiling and classification for Dataplex Catalog integrations, you can now link a Google BigQuery JDBC connection in the "JDBC GCP Connection" field of the Dataplex Catalog capability. If you already ingested assets via a Google BigQuery JDBC connection and later synchronize Dataplex Catalog, the existing assets are reused. (in preview)
  Related topic: Steps overview: Google Dataplex Catalog ingestion
  Note Don't use the "Metadata Synchronization" tab in Database assets ingested via the Dataplex Catalog integration to synchronize the Dataplex Catalog data. The tab will be removed in a future release. Instead, synchronize the data through the capability. For more information, go to Synchronize via the Google Dataplex Catalog ingestion.

Collibra Data Lineage Service

• When ingesting BigQuery data sources, Collibra Data Lineage now supports INSERT ROW statements.
• When ingesting Oracle data sources, Collibra Data Lineage now correctly resolves database links used by synonyms, regardless of the casing used.
• When integrating Power BI with DAX analysis enabled, measures that are used in DAX queries are now shown in the technical lineage.
• When ingesting SQL Server Integration Services (SSIS) data sources, artificial -RES tables no longer break end-to-end lineage.
• When ingesting Informatica Intelligent Cloud Services (IICS) data sources with multiple rules applied in IICS, Collibra Data Lineage now applies all such rules to successfully generate technical lineage.
• When the unsupported dialect 'Trino' is detected, Collibra Data Lineage now automatically uses the 'PostgreSQL' dialect for more successful parsing.

Infrastructure

• You can now add labels and annotations in the edge-cd pod of your Edge site.

Security

• We improved the security of Edge and Collibra Cloud sites.

Lineage Harvester (CLI and Edge)

• If you experience errors when integrating SQL Server Reporting Services (SSRS) data sources, Collibra Data Lineage now provides more detailed error logging, to help resolve issues.
• When ingesting any of the supported data sources via Edge, the “Analyze Only” (deprecated) and “Delete Raw Metadata After Processing” fields are now correctly cleared by default.
• Source configuration is no longer relevant when ingesting Snowflake data sources. As such, the “Source Configuration” field is now deleted from the “Technical Lineage for Snowflake” Edge capability template.
• Resolved an issue during Snowflake metadata harvesting that caused a 'Fail to retrieve row count for first arrow chunk' error.
• When ingesting BigQuery data sources, Collibra Data Lineage:
  • Now correctly harvests stored procedures that do not have parameters.
  • No longer harvests irrelevant objects from JavaScript user-defined functions (UDFs).
• When ingesting dbt Cloud data sources, Collibra Data Lineage now correctly harvests all jobs and projects.

Collibra Data Lineage Service

• When creating a custom technical lineage via the batch definition method, if you use the “props “property in your lineage files to specify the full names and domain IDs of assets, Collibra Data Lineage now checks the validity of the “domain_id” value, to mitigate synchronization failure.
• Collibra Data Lineage now also uses the '$XX' notation for columns in files when creating technical lineage for Azure Data Factory data sources.

Security

• We improved the security of capabilities via Edge and Collibra Cloud sites.

Metadata integrations

• You can again classify data and get samples via the cdata.jdbc.oracleoci.OracleOciDriver driver on Edge.
• You can now download the input metadata from the Databricks Unity Catalog integration "Synchronization Results" dialog box.
• You can now use "Workload Identity Federation (WIF) authentication with GKE" for Google Dataplex integrations.
• You can now download the input metadata from the Databricks Unity Catalog integration "Synchronization Results" dialog box.
• You can now download the input metadata from the GCS "Synchronization Results" dialog box.

Protect

• You can now mask data for roles that aren't part of standards and rules, using the new "Default masking for everyone" option in the Protect for Snowflake Edge capability. This can help protect data by default and reduce the risk of unauthorized access.

Collibra Data Lineage Service

• When creating technical lineage for Snowflake, Collibra Data Lineage now processes SQL queries from the access history without failing due to parsing errors caused by the "query_tag."
• When ingesting DataStage data sources, ingestion and synchronization processes no longer fail if “stage.outputpins”, meaning output links from a specific stage in a DataStage job, are empty.
• Microsoft recently announced that “datasets” have been renamed to “semantic models” in Power BI. This change is now reflected in the Collibra Data Lineage user interface.
• When integrating MicroStrategy data sources, Collibra Data Lineage now creates the relevant relations between metrics in the business lineage, as well as in the technical lineage.
• Collibra Data Lineage can now create lineage for the UNPIVOT clause in Snowflake.

Metadata integrations

• Profiling with random sampling on Databricks driver version 2.7.3 no longer requires manual configuration. The EnableNativeParameterizedQuery=0 property is now automatically set at runtime when this driver version is detected.

Protect

• Protect for Snowflake now excludes columns from shared databases.
• The correct row filter is now applied in Snowflake when Protect rules result in a combination of "Show Everything" and "Show Some" row filters on the same table.

Collibra Data Lineage Service

• When integrating MicroStrategy data sources, Collibra Data Lineage now creates lineage between metrics.
• When ingesting Microsoft SQL Server data sources, Collibra Data Lineage now supports ALTER TABLE … DROP CONSTRAINT constructs.
• When ingesting OpenLineage sources, Collibra Data Lineage automatically recognizes the data source type when not provided in the "Source Configuration" field in the technical lineage for OpenLineage capability,
• Technical lineage for Azure Data Factory now derives dataset columns directly from the source "SELECT * FROM" SQL query, when query is available. This enhancement improves end-to-end column-level lineage.

Security

• We improved the security of capabilities via Edge and Collibra Cloud sites.

Metadata integrations

• We implemented a code change to fix a persistent failure of “refresh enhance schema" jobs when using the “SAP Datasphere Catalog synchronization” capability. The failing jobs were UI refresh jobs, not data synchronization or ingestion jobs. Now, when ingesting SAP Datasphere metadata, Collibra no longer triggers “refresh enhance schema" jobs.
• You can now use the ADLS connection to integrate with the government-specific Microsoft Entra ID, formerly Azure Active Directory. The new "Azure US Government Cloud Host" checkbox allows you to specify that authentication must go through the government-specific Microsoft Entra authentication endpoint.
• You can now use the Azure Entra ID service principal authentication method to connect to Databricks when integrating Databricks Unity Catalog.
• When some tables are not ingested during Databricks integration, you can now use the JVM argument field of the Databricks Unity Catalog synchronization capability to ensure all tables are ingested. Use this field only if recommended by Collibra support.

Protect

• Protect for Snowflake now supports creating masking policies and row filtering on the same columns.

Lineage Harvester (CLI and Edge)

• When you create technical lineage for Databricks Unity Catalog, synchronization no longer fails if the username and password fields in the Collibra Console are empty, provided the Technical Lineage Admin connection is used.
• When the lineage harvester is harvesting Azure Data Factory metadata, the process no longer gets stuck in a loop if an incorrect client ID or client secret is provided.
• When ingesting dbt Cloud data sources, Collibra Data Lineage now correctly harvests all jobs and projects.

Collibra Data Lineage Service

• When ingesting Oracle data sources, Collibra Data Lineage now ensures that synonyms use the correct schema when accessed via DB links.
• When ingesting Microsoft SQL Server data sources:
  • SQL queries that take a long time to process no longer result in out-of-memory errors.
  • SQL queries with a large number of constant values are now processed more quickly and generate accurate highlights.
• When creating technical lineage for dbt data sources, nested CREATE OR REPLACE statements no longer cause parsing errors.
• Collibra Data Lineage now correctly generates lineage for Microsoft SQL Server and SQL Server Integration Services (SSIS) data sources that use stored procedures with SELECT statements. The lineage is created with accurate name casing, ensuring correct representation of object names.
• When ingesting Microsoft SQL Server data sources, Collibra Data Lineage now supports the parameter "OPTIMIZE_FOR_SEQUENTIAL_KEY" in CREATE statements.

Infrastructure

• When you update an Edge site, the OTEL pods now have more time to startup. Previously, the OTEL pods were unable to startup in time, resulting in unhealthy Edge sites.
• We fixed an issue where Edge sites with forward proxies that used special characters became unhealthy when they were updated.

Security

• We improved the security of capabilities via Edge and Collibra Cloud sites.

Metadata integrations

• You can now integrate with Azure AI Foundry on Edge sites that have been installed with a Man-in-the-Middle (MITM) proxy.
• When you add custom AI metric mappings for an Azure ML synchronization, the Metric field is now a dropdown list of available and supported Azure ML metrics.
• When you add custom AI metric mappings for an Vertex AI synchronization, the Metric field is now a dropdown list of available and supported Vertex AI metrics.
• We expanded the list of custom metrics that can be ingested when you synchronize an Azure AI Foundry integration.

Protect

• Protect now supports Databricks Unity Catalog metadata integration.

Lineage Harvester (CLI and Edge)

• When creating technical lineage for Dataplex, Collibra Data Lineage now supports the "Processing Level" capability parameter.
• You can now use the Azure Entra ID service principal authentication method to connect to Databricks when creating technical lineage for Databricks Unity Catalog.

Collibra Data Lineage Service

• Technical lineage for OpenLineage, including Apache Airflow and AWS Glue via OpenLineage, is now generally available.
• When ingesting Snowflake data sources, Collibra Data Lineage now correctly tags external stages that are loaded from dependent metadata as schema-less tables, which allows for correct downstream consumption.
• When ingesting Microsoft SQL Server, Oracle, Snowflake, or Teradata data sources, Collibra Data Lineage now accurately generates lineage for SQL queries that include INSERT INTO combined with a SELECT statement that contains a wildcard (*).
• When integrating Power BI, Collibra Data Lineage now supports the Power Query M function "Table.Group".
• When integrating one of the supported BI tools, synchronization no longer fails if you change the specified target domain in Collibra.

Metadata integrations

• The JDBC database synchronization now correctly uses the first synchronization rule to create Schema assets in the target domain.
• The performance issue during the Databricks JDBC synchronization of empty schemas has been resolved.

Infrastructure

• We improved the installation checks for Edge sites with custom certificates. This ensures a smoother Edge site installation process when using Man-in-the-Middle proxies.
• You can now use one of the following custom Docker Helm registries to mirror the Collibra JFrog image registry for your Edge site:
  • Amazon Elastic Container Registry
  • Azure Container Registry
  • Google Container Registry

Protect

• When adding the Protect for BigQuery capability, you can now grant table-level permissions using the new "Grant access to tables" checkbox. When enabled, Protect assigns the BigQuery Data Viewer role to tables instead of assigning the Fine-Grained Reader role to columns. This feature applies only to rules that contain row filters and have "Grant Access to Data Linked to Selected Assets" enabled. To use this feature, ensure the GCP service account configured for Protect has the "bigquery.tables.getIamPolicy" and "bigquery.tables.setIamPolicy" permissions.

Lineage Harvester (CLI and Edge)

• When creating technical lineage for Databricks Unity Catalog, Collibra Data Lineage now supports the "Processing Level" capability parameter.
• You can now include external locations in technical lineage for Databricks Unity Catalog. This allows you to visualize the lineage between a Databricks table and the external location it was created from, such as an S3 file.
• When creating a technical lineage via Edge, you can now use a new Edge Cloud Connection to access SQL and data source files in an AWS S3 bucket, a Google Cloud Storage bucket, or a Azure Data Lake Storage container. This new connection option is available for all data sources for which a Shared Storage connection is available, meaning:
  • All supported JDBC data sources
  • Apache Airflow (via OpenLineage)
  • AWS Glue (via OpenLineage)
  • Custom technical lineage
  • DataStage
  • dbt Core
  • Informatica PowerCenter
  • OpenLineage
  • SSIS

  New Edge capabilities, per data source, are required and are now available when you choose to connect via a Cloud Connection. For guidance on how to use the new connection option, go to Create a technical lineage via Edge, select the relevant data source, and then click “Cloud Storage connection”.

Collibra Data Lineage Service

• When creating technical lineage for Azure Data Factory (ADF), Collibra Data Lineage now identifies and avoids potential duplicates caused by invoking certain ADF pipelines multiple times.
• When ingesting Microsoft SQL Server data sources, Collibra Data Lineage now supports SQL queries with large WHERE clauses containing thousands of repetitive OR/AND conditions, for example: OR Counter_Replaced_Indicator='X' AND measuring_point='...'.
• When integrating Power BI with DAX analysis enabled, synchronization no longer fails if the input token count exceeds the maximum token allowed limit. Instead, an analysis error is generated and processing continues.
• When integrating Power BI, Collibra Data Lineage now processes workspaces in shared capacities.

Infrastructure

• We fixed an issue which resulted in wildcards not being accepted in the no_proxy or no_host properties of the proxy.proprerties.
• We fixed an issue where connections and capabilities couldn't be created and jobs weren't able to run on healthy Edge sites.

Security

• We improved the security of capabilities via Edge and Collibra Cloud sites.

Metadata integrations

• When you synchronize Databricks Unity Catalog, new schemas are now ingested under new domain.
• When you ingest MLflow AI via Edge, the MLflow model tags are now ingested as key value pairs under the characteristic "Source Tags".
• When you ingest Azure AI Foundry via Edge, the agent description is now ingested and mapped to the Azure AI Foundry agent asset description.
• You can now use the ADLS connection to integrate with the Azure Government-specific Azure AD endpoint for authentication.

Lineage Harvester (CLI and Edge)

• You can now specify a range of days to collect data when creating technical lineage for Databricks Unity Catalog.
• When ingesting Snowflake data sources, Collibra Data Lineage now provides improved lineage for:
  • The "COPY INTO TABLE FROM <snowflake_stage>" operation. Lineage shows the flow from the underlying cloud storage to the stage and then to the table.
  • The “COPY INTO <snowflake_stage> FROM <table>” operation. Lineage shows the flow from the table to the stage to the underlying cloud storage.

  In the technical lineage Browse tab pane, the files are shown in accordance with the following hierarchies:

  • For Azure: “AZURE” > Account > Container > Directory > Directory (n levels) > File
  • For GCS: “GCS” > Bucket > Directory > Directory (n levels) > File
  • For S3: “S3" > Bucket > Directory > Directory (n levels) > File

Collibra Data Lineage Service

• When ingesting SAP HANA data sources with calculation views, Collibra Data Lineage now creates lineage between a synonym in one schema and a CUBE in another schema, based on the ID of the calculation tag.
• When ingesting Teradata data sources:
  • Collibra Data Lineage now supports TD_UNPIVOT clauses.
  • The FOR column in an UNPIVOT clause is now treated as an output column.

Collibra Data Lineage Service

• When ingesting Azure Data Factory data sources, Collibra Data Lineage now
  • Supports the Until activity type.
  • Separates "Pipeline Runs" and "Pipeline Triggers". Additionally, individual objects are separated based on date, ID, or name.
• When ingesting Informatica Intelligent Cloud Services data sources, Collibra Data Lineage now supports recursive parameter resolution.
• When ingesting BigQuery data sources, Collibra Data Lineage now supports SQL queries that include the “GROUP BY ALL” clause.