What's new in Cloudera Data Visualization 8.0.8

Administrators can now define maximum token limits for each AI model profile in the embedding and completion model configurations. AI visuals and annotations automatically respect these limits, ensuring greater control, more predictable behavior, and reducing unexpected errors.

Cloudera Data Visualization provides improved management and new auto-discovery for AI Inference profiles within the same environment. Profiles can be disabled and are now clearly marked on the interface. The updated UI includes Type and Status columns, consolidated action buttons, and improved sorting. In Cloudera AI environments, models can be synced directly from Cloudera AI. Synced profiles are mostly read-only, allowing users to view all relevant values while preventing edits to settings that are managed by Cloudera AI and would not be meaningful to change locally.

For more information about configuring AI models, see Managing AI settings.

The Annotations modal now includes a new Advanced tab, visible when the AI summary option is enabled. This tab displays detailed dashboard and response data, allowing you to see the exact dashboard or visual data sent to the AI service, along with the AI’s full response. You can gain clearer visibility into the AI data exchange ith this improvement.

For more information, see Creating an AI-powered annotation.

OpenAI GPT-5 and o3 models can now be used as a completion model for AI-powered features in Cloudera Data Visualization like AI visual and annotations.

Anthropic’s Claude models can now be used as a completion model for AI-powered features in Cloudera Data Visualization like AI visual and annotations.

For more information on managing AI settings for completion models, see Completion.

You can now copy a direct link to any comment or reply in a comment thread for dashboards and visuals using the new Copy comment link or Copy reply link option from the comment’s dropdown menu. Pasting the link in a browser opens the relevant comments panel, highlights the specific comment or reply, and scrolls it into view. This improvement makes it easier to share or navigate to specific comments and replies, improving collaboration.

Comment threads now show only the most recent reply by default, keeping discussions cleaner and easier to read. If additional replies exist, a View X replies option appears, allowing you to view all replies with a single click. This enhancement makes it easier to focus on the latest activity while still providing quick access to the full conversation when needed.

For more information, see Commenting on visuals and dashboards.

Cloudera Data Visualization 8.0.8 introduces support for Ubuntu 22 as a parcel deployment platform in Cloudera Base on premises environments. This enhancement enables new installation and upgrade scenarios for customers running Ubuntu 22.

For installation instructions, see Installing Cloudera Data Visualization in Cloudera Base on premises.

The icon for Cloudera Data Visualization on the Knox application list has been updated to the official Cloudera Data Visualization replacing the previous generic icon. This provides a more consistent product branding experience.

You can now download dashboard sheets and visuals in PowerPoint (PPTX) format and send them in email as PPTX attachments. This improvement makes it easier to share insights and present visual content outside Cloudera Data Visualization.

You can now create visuals directly from selected fields on the Dataset Profiling tab. This improvement simplifies dashboard creation, making it easier to move from data profiling to visualization. You can quickly generate visuals and either add them to existing dashboards or create new ones.

For more information, see Creating a visual from the Dataset Profiling tab.

Import and export operations now generate detailed audit log entries for each visual, indicating whether it was created or updated and by which user. Previously, when importing dashboards or visuals from a JSON file, the logs did not specify which visuals were created or replaced. This enhancement allows cluster administrators to easily track which dashboards and visuals were created or updated during export and import operations.

For more information, see Logging manual export/import.

A new Never Expire option is now available for dataset and data connection caches. This allows cached results to be retained indefinitely, which is especially useful for static or infrequently changing data sources. The cache settings interface has also been updated for clearer visibility and easier configuration. You can configure caches to never expire, reducing unnecessary data refreshes and improving performance for stable datasets.

For more information, see Result cache.

Cloudera Data Visualization now supports specifying a custom Kerberos host when configuring Kerberos-secured connections. A new Kerberos Host field (attribute: KrbHost) has been added to the data connection modal, allowing connections to clusters that use non-default Kerberos hosts.

A new API endpoint (/adminapi/connection/export/<id>) allows exporting data connection definitions in JSON format. You can now automate the migration, backup, and replication of data connections, reducing manual effort, minimizing errors, and supporting scripting for automated workflows.