Cloudera Docs

What's new in Cloudera Data Visualization 8.1.1

Cloudera Data Visualization 8.1.1, released on March 17, 2026 introduces a number of performance, usability, and security enhancements. This release improves the AI visual, dashboard usability, administration features, and data extract configuration. It also improves performance in snapshot and PDF generation, and resolves several issues related to Knox integration.

New features and improvements

AI visual improvements
VIZ-3539 – Improved AI visual parameter mapping from SQL and vector queries

AI visuals now use a more reliable mapping between SQL query result column names and dataset column aliases when populating parameters. This results in

  • More consistent and readable column labels in the Parameters Sharing modal
  • Better handling of title case and custom dataset aliases
  • Improved compatibility across Solr, Impala, and SQLite connections.

For more information on how the generated parameters can be shared with other visuals when using an AI visual, see Sharing parameters.

VIZ-3664 – Enhanced AI visual Execution Details modal

The Overview section of the AI visual Execution Details modal now includes additional information to simplify troubleshooting and analysis. You can see the completion and embedding model ID, name, and engine, as well as the full prompt and its context.

For more information, see Viewing response details.

VIZ‑3712, VIZ-3898 – Improved reliability of AI visual configurations

AI visual configurations are now automatically validated before queries are executed. and retries generation when SQL syntax errors occur. If a generated configuration produces an invalid SQL query, the system sends the error feedback to the AI model and tries to regenerate the configuration up to three times. If regeneration fails, the original error is returned to the user. This reduces SQL syntax errors when using AI visuals, especially with engines such as Impala and Trino.

VIZ-3837 – Stop button for AI visual responses

You can now cancel in‑progress AI visual chatbot responses using a new Send/Stop button allowing you to cancel long or irrelevant answers. Canceling a response restores your last question to the input field so it can be edited and submitted again.

For more information, see Interacting with the AI visual.

VIZ‑3919 – Improved AI SQL and parameter sharing reliability for Trino data connections

When you generate SQL using the AI visual with Trino, Cloudera Data Visualization now provides more reliable query generation and parameter sharing between visuals. This update

  • Aligns AI prompts and examples with Trino identifier syntax
  • Routes aggregate filters correctly to HAVING instead of WHERE
  • Improves validation and retry behavior
  • Provides more helpful error handling for non-JSON AI visual responses.
Dashboard improvement
VIZ‑3860 – Manually refresh individual visuals on dashboards

When Refresh visuals manually is enabled, you can now refresh visuals individually instead of reloading the entire dashboard. This improves performance for dashboards with many visuals.

For more information, see Refreshing visuals manually in dashboards.

Performance improvement
VIZ‑3868 – More efficient snapshot and PDF generation

The snapshot generation workflow used for downloads and scheduled emails has been optimized. Screenshot and PDF operations are now skipped when they are not required, and PDF generation is triggered only when the PDF output format is requested.

Administration improvements
VIZ‑3490 – Site setting to enforce dashboard titles

A new site setting allows administrators to require a dashboard title before users can save a dashboard. When enabled, you see a Title is required message if you try to save a dashboard without a name. The setting is disabled by default.

For more information, see Managing dashboard site settings.

VIZ‑3685 – Manage Users and Groups page UI improvements

The Manage Users and Groups administration page has been redesigned for a more consistent administration experience when managing users and group assignments.

Data extract improvement
VIZ-3178 – Improved Create Data Extract experience

The Data Extract tab in the Create Data Extract modal now behaves more reliably when configuring extracts. This update includes the following improvements:

  • Validation in the Edit Data Extract modal has been adjusted to prevent misconfigured extracts.
  • Extract creation now uses the correct application/json content type, so creating extracts works as expected when you access Cloudera Data Visualization through Knox.
  • Dimension and partition expressions in extracts are saved using the correct bracketed dataset field alias format (for example, [alias]), and existing extracts that used the previous format are automatically updated.

Fixed issues

VIZ-3921 – Log file download through Knox now works as expected

Fixed an issue where downloading log files from the Logs and Usage page failed when Cloudera Data Visualization was accessed through a Knox proxy. The request now uses the correct application/json content type, so selected log files can be downloaded successfully as a ZIP archive in Knox‑proxied environments.

VIZ‑3867 – Refresh Visuals button layout corrected

Fixed an issue where the Refresh Visuals button overlapped with commenting and parameter icons when Refresh visuals manually was enabled. The button is now positioned correctly.

VIZ-3850 – Fixed memory leak during dashboard snapshot generation

Fixed an issue where resources used for generating dashboard snapshots were not fully released, causing snapshot‑related objects and log file descriptors to accumulate over time. Snapshotting now cleans up these resources correctly, improving memory usage and long‑running stability.

VIZ-3834 – Unnecessary save prompts removed for unchanged AI completion model profiles

Fixed an issue where users were prompted to save AI completion model profiles and site settings even when no changes were made. Unchanged profiles are now correctly detected and users are no longer prompted to save settings when no modifications were made.

VIZ-2955 – Fixed Impala over Knox connection failures with load balancing

Fixed an issue where Cloudera Data Visualization could not connect to Impala through Knox when load balancing was enabled because of a hard‑coded HTTP session cookie name.

Fixed Common Vulnerabilities and Exposures

The following Common Vulnerability and Exposures (CVE) have been fixed in Cloudera Data Visualization 8.1.1.