Table of Contents

Changelog - 5.2.1

Ramya Priya Updated by Ramya Priya

The latest 5.2.1 version brings a powerful suite of updates—key highlights include automatic data deduplication for Redshift to ensure accurate results, shareable conversation threads for better collaboration, and increased flexibility in export and SSO settings. Enhancements like customizable SAML attributes, improved error message handling, and automatic Spark recovery further boost performance and usability. We've also resolved several issues across Vizpads, Search, and Insights to ensure a seamless, reliable experience. Let's dive into these updates.

🚀 New features

Automatic data model deduplication now available for Redshift

With the 5.2 release, automatic data model deduplication is now extended to support Redshift Live Business Views. Tellius logically manages many-to-many relationships to automatically identify and remove duplicate data entries resulting from joins in your data models, preventing issues like fan traps and chasm traps.

Under Data → Business View → Edit, you can enable deduplication at the Business View level. This feature handles complex scenarios involving schemas with multiple interconnected fact and dimension tables. It simplifies data model management by automatically adjusting underlying queries based on the required granularity of analysis.

For example, when combining sales and call data through common dimensions (like customer or territory) in Redshift, deduplication ensures that totals for calls or sales are not inflated due to overlapping data. Tellius smartly deduplicates data at the necessary level (e.g., per employee, per call, per sale) based on the context of your query, ensuring accurate results across different analyses.

📈 Enhancements

Conversations are now shareable

The conversation threads in Kaiya Conversational Search can now be shared, enhancing collaboration and making it easier to continue discussions across the team. Under Kaiya → Previous Chats, click on the three-dot menu and select "Share" to share the required conversation.

Increased bulk selection list

In Vizpads, the bulk selection of filter values has been increased from 2000 to 5000 values, allowing you to work more efficiently with larger datasets.

Enhanced export with row and column limits

You can now specify both row and column limits when exporting Pivot tables. This added flexibility allows for better control over the exported data and helps prevent exceeding data point limitations.

Secure sharing of resources

Restricted sharing is now enabled by default. This means that if a user doesn't have write access to the underlying resources (Vizpads, Insights etc.), they cannot share the top-level resource (Business Views, datasets). This change enhances security by ensuring that only users with the appropriate permissions can share resources.

Automatic recovery for Spark

Implemented Spark probes to automatically detect and handle dead executors by spawning new ones and restarting pods as needed. If executors fail, the automatic recovery ensures the system quickly returns to optimal performance without requiring manual intervention, resulting in minimized downtime and reduced disruption to data processing tasks.

Support for customizable SAML attribute names

Under Settings → Security → Authentication → SAML, you can now configure custom attribute names in SAML claims, making it easier to integrate with your identity provider's specific requirements. This enhancement provides flexibility and compatibility in SAML authentication setups.

Improved display of long messages

Long error and success messages are now more user-friendly. The first 20-30 characters are displayed initially, with a "Show More" option available. Clicking "Show More" opens a popup displaying the full message. This enhancement makes it easier to read and manage lengthy notifications without cluttering the interface.

Update usernames and simplify password settings for SSO integration

You can now update usernames directly under Settings → Users & Groups → Users. This is especially helpful for customers who initially created users with custom usernames and later need to integrate with Single Sign-On (SSO), which requires the username to be an email address for successful login.

In addition, we've introduced a "Set Password" toggle in the user creation form. Setting a password is now optional—if you disable this toggle, the password fields will be hidden, allowing you to create users without a password. If you don’t set a password, then sign-in will be allowed through SSO only. This enhancement simplifies user management and enhances security by allowing accounts that rely solely on SSO for authentication.

🛠️ Minor fixes

  1. In Vizpad tables, the issue with the display of blue dot indicating comments being present even after removing the associated comment has been resolved.
  2. In Vizpad tables, resolved the issue with the comments added during the View mode not being reflected in the Edit mode.
  3. In Vizpad tables, fixed the issue with replicating the comments after cloning the selected tables.
  4. In Search, addressed the issue with applying incorrect filters for Insights triggered from charts in Search.
  5. In Vizpads, resolved the issue with the overlapping and removal of applied filter pills in the explainable AI charts.
  6. Values alignment feature will now be available for explainable AI charts (in addition to tables and detail tables).
  7. Provided relevant error message to scenarios when users are trying to share the resources for which they don't have write access.
  8. Under Data → Prepare, fixed the issue with downloading the CSV format of datasets before transformation.
  9. Fixed the issue with using Search after a long idle time in the browser.
  10. Fixed the issue with incorrect selection of Include/Exclude columns for creating Insights from Search.
  11. In Trend Insights, resolved an issue where selecting a weekend date on a weekly date-column chart caused errors related to current and previous time filters.
  12. In Vizpads, fixed the issue with seamless display of explainable AI charts.
  13. In Search and Vizpad charts, fixed an issue where the prior time period overlapped with the current time period when the current time was included in analyses. 
  14. In the embedded mode, fixed an issue where the Vizpad filter section was not properly aligned. 
  15. In Feed, fixed the issue with hourly ARIMA forecasts, improper handling where no insights are discovered, and the absence of associated background jobs due to incorrect start time formatting.
  16. In Insights, added support for analyzing market share aggregations using calculated columns.
  17.  In Vizpads, updated error messages for pivot table exports that exceed 20,000 data points. The message now clearly states: "Data points exceed 20k. Please adjust row or column limits to reduce data points". 
  18. Under Data → Connect → Create New, resolved an issue where users were unable to edit the advanced URL settings in the Spark SQL connector.
  19. In Vizpads, resolved an issue where long dimension values with exponents were not properly handled in Table comments.
  20. Under Data → Connect → Create New, the maximum file size supported for upload is displayed. Additionally, an error message will now be shown if the uploaded file exceeds the 2 GB limit.
  21. In point-and-click model/AutoML, fixed the issue with incorrect display of feature transformations for all columns, ensuring One-Hot Encoding selections are now shown accurately.

Did we help you?

Changelog - 5.2.2

Release 5.2

Contact