> For the complete documentation index, see [llms.txt](https://docs.collibra.com/llms.txt). Markdown versions of documentation pages are available by appending `.md` to page URLs; this page is available as [Markdown](https://docs.collibra.com/data-quality-self-hosted/release-notes/data-quality-observability-self-hosted-release-archive/release-2025-11.md).

# Release 2025.11

* [New and improved](#new-and-improved)
* [Fixes](#fixes)
* [Patch release](#patch-release)

## Release Information

* Release date of Data Quality & Observability (self-hosted) 2025.11: November 24, 2025
* Release notes publication date: October 31, 2025

## New and improved

### Platform

{% hint style="danger" %}
In the upcoming Data Quality & Observability (self-hosted) version 2026.02, Spark 4 is required. Depending on your installation type, additional action may be required.

* **Kubernetes installations**: The upgrade to 2026.02 is seamless and managed automatically within the application container.
* **Spark Standalone installations**: You must manually upgrade your underlying Spark environment to Spark 4 before upgrading to version 2026.02. This is a mandatory prerequisite. To learn how to upgrade Spark, go to [Upgrade Spark](/data-quality-self-hosted/installation/agent/upgrade-spark.md).
  {% endhint %}

{% hint style="info" %}

* The CVE-2025-66516 vulnerability is detected in Apache Tika, which Data Quality & Observability (self-hosted) only uses for content-type detection and the detection workflow processes. For more information on this issue, go to the official [Collibra Support article](https://support.collibra.com/hc/en-us/articles/36894729038615-DQ-Classic-App-critical-vulnerability-identified-in-Apache-Tika-XML-External-Entity-XXE-CVE-2025-66516).

* The CVE-2025-59419 vulnerability is detected only in the DQ Livy containers. However, Data Quality & Observability (self-hosted) does not use the vulnerable module netty-codec-smtp. This module is included due to a transitive dependency, but it is not in use, and there is no direct or indirect way to exploit it because DQ Livy does not expose or use any SMTP protocol or functionality.

* The release images may show CVE-2025-59250 as a vulnerability, even though it has been addressed in the packages. This detection is a false positive.
  {% endhint %}

* To enhance security, authentication tokens now require a minimum secret key length of 256 bits (32 characters) for the `security.jwt.token.secret-key`. If you use environment variables, update the `SECURITY_JWT_TOKEN_SECRET-KEY` configuration to a securely generated value of at least this length. The existing default value will be removed in January 2026. We strongly recommend providing a custom value if you haven't already.

* You can now use the OAuth2 authentication type for [Snowflake connections](/data-quality-self-hosted/connections/supported-connections/supported-jdbc-data-sources/connecting-to-snowflake.md). This enhancement provides a more secure and flexible authentication option for connecting to Snowflake.

* The following drivers have been upgraded:
  * The [Snowflake](/data-quality-self-hosted/connections/supported-connections/supported-jdbc-data-sources/connecting-to-snowflake.md) JDBC driver is now version 3.26.1, to support OAuth-based authentication.
  * The [Databricks](/data-quality-self-hosted/connections/supported-connections/supported-jdbc-data-sources/connecting-to-databricks.md) JDBC driver is now version 2.7.5.
  * The [Sybase](/data-quality-self-hosted/connections/supported-connections/supported-jdbc-data-sources/connecting-to-sybase.md) JDBC driver is now version 16.

* Out-of-the-box [FIPS-compliant JDBC drivers](/data-quality-self-hosted/installation/installing-data-quality-observability-self-hosted-on-self-hosted-spark-standalone/configure-fips-compliant-algorithms.md) are now included in Data Quality & Observability (self-hosted) installation packages. You can test connections by setting "DQ\_APP\_FIPS\_ENABLED" to true or false in the owl-env.sh file or DQ Web ConfigMap.
  * In FIPS mode, "SAML\_KEYSTORE\_FILE" is no longer effective. Instead, DQ uses "SAML\_KEYSTORE\_ALIAS" at the "KEYSTORE\_FILEPATH" with the "SAML\_KEYSTORE\_PASS."

* You can now [manage admin user creation](/data-quality-self-hosted/admin-configuration/more-admin-settings/multi-tenancy-and-tenant-manager/using-tenant-manager.md) during tenant creation, supporting SSO-only user access and meeting security policy requirements.

* You can now use the new "rulelimit" [admin configuration](/data-quality-self-hosted/admin-configuration/admin-console/configuration-settings/admin-limits.md) option to limit the number of rule break records stored in the metastore for Pushdown jobs. This helps prevent out-of-memory issues when a rule returns a large number of break records. The default value is 1000. This option does not affect Pullup or Pushdown jobs with archive break records enabled.

### Jobs

* When [archiving break records](/data-quality-self-hosted/core-concepts/findings/break-records-and-link-ids/archiving-break-records-from-pullup-jobs.md) for Pullup jobs, [link IDs](/data-quality-self-hosted/core-concepts/findings/break-records-and-link-ids.md) are now optional. This change provides greater flexibility in storing and accessing break records, offering a more scalable solution for remediating records outside of Data Quality & Observability (self-hosted). The new format includes the full rule result.
  * Additionally, archiving break records for Pullup jobs now supports multiple runs in a single day. The newest archive file remains "ruleBreaks.csv," and a "ruleBreaks\[timestamp].csv" file is created for each run.
* The pagination controls in Explorer are now at the top of the table list, making navigation easier.
* When you access the [Dataset Overview](/data-quality-self-hosted/core-concepts/navigation/dataset-overview.md) from Explorer or the Rule Workbench, column headers now stay locked in place, consistent with other areas of the application.

### Rules

* [SQL Assistant for Data Quality](/data-quality-self-hosted/core-concepts/navigation/dataset-overview/sql-assistant-for-data-quality/about-sql-assistant-for-data-quality.md) now directly connects to your Collibra Platform instance to make LLM-router calls, eliminating the need for a Kong endpoint configuration. This improves efficiency and expedites requests to the Collibra AI proxy.
* When you add or edit a [rule template](/data-quality-self-hosted/core-concepts/dq-rules/templates.md), the same naming rules as creating a rule are now applied. You can no longer use special characters in rule template names.
* The "Add Rule Template" modal now requires you to complete the "Rule Query" fields before submitting, ensuring all necessary information is provided when creating a rule template.
* The checkbox in the "Settings" modal now accurately shows the rule's status. It is checked when the rule is active and clear when the rule is inactive.
* Users without sufficient permissions can no longer see break record data in the "Archived Breaks" modal, which is available from the Rules tab of the Findings page. This enhancement improves data security by ensuring that only authorized users can access sensitive information.

### Alerts

* You can now [configure webhooks](/data-quality-self-hosted/admin-configuration/admin-console/alerts/alert-webhook-configuration/add-webhooks.md) to send Global Failure alerts, enabling faster identification and resolution of issues.
* You can now include JSON templates with replacement variables in the definition of a webhook. These variables are included in a custom JSON payload sent to the webhook server, enabling more flexible and tailored integrations for alerting.

### Reports

* When you open the Dataset Findings report on the Reports page, you now see the total number of rows in the dataset.

### APIs

* The API v3/jobs/findings now populates the same valid values for datashapes as the API v2/getdatashapes.
* The API v3/datasetdef now includes time zone information for job schedules.

## Fixes

### Platform

* You can now enter credentials for a remote connection when creating a database, restoring the previous functionality.
* When you add role mappings without first specifying a role in Admin Settings > Role Management, the role name and mapping now show correctly. The success message accurately reflects the completed action.

### Jobs

* Jobs that run in EMR instances no longer get stuck in an "Unknown" status.
* When you use the file lookback parameter "-fllb" with the [added date column](/data-quality-self-hosted/core-concepts/findings/data-quality-monitors/job-tab/job-command-line-parameters/advanced/add-date-column.md) "-adddc", the "OWL\_RUN\_ID" column is now correctly included in historical and current dataframes. This prevents job failures during outlier calculation.
* When working with SQL Server Pushdown jobs, the "ORDER BY" clause in a source query now requires the "TOP" command due to a SQL Server limitation. For queries that return an entire table, "ORDER BY" cannot be used because Pushdown relies on the source system for query processing. Additionally, API calls for rule breaks on SQL Server Pushdown jobs require a limit to ensure repeatable ordering for queries using offsets to page the rule preview. If no limit is applied, the API call will return a 404 error.
* The search string in the profile now refreshes correctly when you change the column. Search results are displayed accurately without needing to manually collapse and re-expand to update the values.
* The integration status button in the Metadata Bar now correctly reflects the status when manual integration fails. This behavior is aligned with the accurate functionality of the Dataset Manager.
* Explorer now remembers its last state when you return from the job configuration.
* When creating a job, the "Include Views" option under the Mapping step now functions correctly.

### Findings

* The data preview of Length\_N shapes now displays correctly in the Shapes tab on the Findings page.

### Rules

* You no longer encounter access errors when using the "Run Result Preview" function in Rule Workbench. The function now checks if you are the dataset owner, ensuring proper access permissions.
* When you add or edit a rule template, the same naming rules as when creating a rule are now applied. You can no longer use special characters in rule template names.
* The validation message "Validation is not supported for stat rules" is now shown again during syntax validation, both as a warning banner and in the dialog box, restoring the legacy behavior.
* When "snowflake\_key\_pair" is used as a secondary connection under a rule in a dataset, the command line now adds a Spark configuration in the Kubernetes environment.
* The "Pretty Print" function on the Rule Workbench now preserves spaces inside double or single quotes during formatting, preventing unintended changes to column names.

### Alerts

* Alerts for non-authenticated webhooks now send correctly at runtime.
* Multi-condition alerts now trigger correctly and emails are sent successfully.
* The search feature now works correctly on the Alert Notifications page.

### Collibra Platform integration

* Large aggregation requests no longer cause out-of-memory errors. Results are now fetched in pages and partially stored on the filesystem, ensuring memory usage stays within safe limits while still returning results in the required stats format.
* After running an integration, Collibra Platform now shows correct values for custom rule attributes when a rule in Data Quality & Observability (self-hosted) contains a filter. The Loaded Rows, Rows Failed, and Rows Passed attributes now reflect the rule row count, which is based on the subset of the dataset defined by the rule filter.

### APIs

* The "/v2/getdataasset" and "/v2/getcatalogbydataset" endpoints now correctly return the name of the Data Category assigned to the dataset.

## Patch release

### 2025.11.1

* We improved the security of our application.

### 2025.11.2

* Rules now work with "REGEXP" in the same way as "RLIKE," including scenarios with parentheses. This ensures consistent behavior across both operators.