Here is the list of features and updates included in the Kyvos 2024.11.x release.
Kyvos 2024.11 release brings the following new features, enhancements, and bug fixes.
This release comes with the following Conversational Analytics enhancements:
Seamless Visualization Control with Pinboard: Kyvos now introduces pinboard, allowing users to easily pin or unpin visualizations. Pinned visualizations appear as customizable widgets that can be resized, dragged, and organized. Changes to pinned visualizations are fully synchronized between chat session and pinboard modes, ensuring seamless updates across both views.
Export Chat Sessions as PDF: Users can now export their chat sessions as PDF files. Kyvos offers flexibility, allowing users to export specific parts of a chat session, such as summaries, card information, or only answered conversations.
From the GenAI Configuration page in Kyvos Manager, users can now:
Generate a summary, title, or key insight for the answer tailored to their business requirements.
Set a default connection for MDX calculations, MDX queries, and/or Conversational AI.
Fine-tune OpenAI models to enhance the accuracy and relevance of answers.
This release comes with the following no-Spark enhancements:
JRE Version Modification: For no-Spark-based clusters, you can now switch between JRE versions 8 and 17 (JRE 8 or JRE 17). However, switching the JRE is not applicable for Spark-based deployments, and only JRE version 8 is used.
Conditional Column Masking (Beta): Prior to this release (Kyvos 2024.10), Kyvos supported conditional masking on dimensions. With this release, Kyvos has added support for conditional masking on measures.
Automated Concurrent Processing Task Count for Compute Engines: Kyvos now support automatic calculation of concurrent processing task count for compute engines to optimize Kyvos Analytical Store memory usage. The Auto mode is set by default. Additionally, users can also set this configuration manually by setting the Manual mode. This applies to automatic and wizard-based deployments.
Support of changing default compute cluster type for on-premises deployments.
Users can now view calculated members created in Kyvos when connected through the SQL interface in BI tools such as Tableau, Power BI, or MicroStrategy.
NOTE: To enable calculated members via the SQL interface, set the value of the advanced semantic model property kyvos.query.sql.calculatedMember.enabled to true.
Enhancements in the semantic model now enable SQL users to interact with it in a tabular format. Users can use any column as a dimension or measure, regardless of its definition in the semantic model design. This feature requires enabling Raw Data Querying on the semantic model.
Users can now use hidden base measures of a semantic model when creating calculated measures under Non-Empty Behavior by setting the value of the allowHiddenMeasuresForNonEmptyBehavior property to true in the portal.properties file.
Users can now view a column marked as "Modified" when changes are made to it in a dataset. Additionally, they can view the details of those changes. This feature helps users quickly identify modified columns and review their specifics.
With this release, Helix Task Manager will be enabled by default for fresh Spark and No-Spark deployments by setting the value of the ENABLE_HELIX_TASK_MANAGER property to yes.
Kyvos has upgraded the Apache Tomcat version for Kyvos Manager and Kyvos to 10.1.33.
This release comes with the following Kyvos Enterprise enhancements (AWS, Azure, GCP, and on-premises):
Automatic Memory Configuration for Wizard-Based Deployments: Kyvos now supports memory configuration in Auto mode for all Kyvos services, including the BI Server, Query Engine, and Compute Server. The default mode is Auto, which was previously supported for automated deployment, but it can be switched to Manual if needed.
Single-click rollback and patch deployment: Users can now perform single-click rollback and patch deployment for Kyvos and Kyvos Manager. This means both can be easily reverted or updated with just a single click.
This release comes with the following cloud enhancements (AWS, Azure, and GCP) enhancements:
S3 Reading for Quering on Kyvos Analytical Store: With this release, Kyvos is certified for S3 reading for querying on AWS using the Kyvos Analytical Store.
Configure Application Gateway from Different Resource Group: Users can now configure Application Gateway from Different resource group during node addition or deletion.
Use Existing GKE Cluster: Kyvos now supports using an existing GKE Cluster using Terraform-based deployment for new deployments, or existing deployments after upgrade.
With this release, users can now use Kyvos base and calculated measures together in SQL queries to create a derived SQL expression.
This release comes with the following Kyvos Visualization enhancements:
Highlight Declining Trends with Downward Triangle Icon in Reports Measures: Kyvos now supports highlighting a decline in report measure values by showing a decreasing trend using a downward or inverted triangle icon.
Persisting Column Widths in Crosstab and Table Data Cards: Users can save changes to column widths in crosstabs and tables within data cards while in design mode. Additionally, column widths can be reset to match those in the worksheet. If a column's width has not been modified in the data card, it will inherit the width set in the worksheet.
This release comes with the following API enhancements:
The following new API has been added.
Execute Natural Language Query: To execute a natural language query.
New parameters have been added to the following existing REST API:
Create semantic model: To create a single flexible semantic model.
smodelType: To create either a multidimensional or single table.
Default: ["MULTIDIMENSIONAL"]
Enum: "MULTIDIMENSIONAL" "SINGLE_TABLE"
smodelViewType: If a semantic model type is 'SINGLE_TABLE', then the semantic model view type could be Flat or Smart view.
Default: ["FLAT_VIEW"]
Enum: "FLAT_VIEW" "SMART_VIEW"
smodelName: The name the user wants to use to create the semantic model. If a semantic model with the same name already exists, a unique name will be assigned to the semantic model. This name will be used to create the underlying entities of the semantic model.
folderName: The folder name where the user wants to create the semantic model. A new folder with the specified name will be created if the folder does not exist. All underlying entities will also be created in the folder with the same name.
connectionName: The name of the connection that must be used in a dataset
dbName: The database name that must be used in the dataset.
tableName: The name of the table that must be used in the dataset.
schemaName: The schema name that must be used in the dataset.
For further details, please see the REST API guides.
Kyvos Resources Deployment for Azure using Template Specs: With this release, Kyvos resources for Azure will be deployed exclusively using Template Specs. Azure Resource Manager templates for deploying Kyvos resources on Azure are no longer supported.
Kyvos Resources Deployment for GCP using Terraform Script: With this release, Kyvos resources for GCP will be deployed exclusively using Terraform scripts. Deployment Manager scripts for deploying Kyvos resources on GCP are no longer supported.
Manual Steps for Presto Connection Creation Instead of Kyvos: With the Kyvos 2024.11 release, users will not be able to create a Presto connection from the Kyvos User Interface. However, users can still establish a Presto connection by performing certain manual steps.
Updated default value of the MDX_SESSION_TIMEOUT property to 60 mins from the previous value of 1440 minutes.
Note This document covers generic release notes for all Kyvos editions. Depending upon your deployment type, some of these may not be relevant to your environment. |
Read more