Managing Configuration files

• To integrate Kyvos with host applications, configure the User Interface (UI) elements for chat sessions:

  1. On the Manage Configuration page, download the ‘integration.config.json’ file.

  2. Open the downloaded file and add the necessary configurations, which include: ‘clearChat’, ‘showMetadata’, ‘showSelectedSemanticModel’, and ‘allowSemanticModelChange’.

  3. Set the value of these configurations to true if you want them to be displayed. By setting these configurations to true, the corresponding features will be displayed in the integration mode.

  4. Once you have made the changes, upload the file back on the Manage Configuration page.

• In the Kyvos integration mode with host applications, to add a new chat session as a worksheet from a URL, the host application must send the following URL to Kyvos:

To configure Elastic Search Service, you must first download the synonym.xls file. Ensure that you must not change the name of the synonym file while saving the file.

1. Download the file. In the file, add the actual value and its synonym you want to use for elastic search.

   Open

   

2. Save the file without changing its name.

3. Upload the file.

1. On the Manage Configuration page, expand Kyvos >Commons>kyvos/commons/config/kyvossearch.

2. Upload the updated synonym file.

3. Click Save.

Open

You can launch dedicated compute servers within YARN containers, enhancing compatibility with YARN-based on-premises deployments. This enhancement enables users to run dedicated compute servers specifically for semantic model processing, ensuring improved resource isolation, scalability, and performance.

To set up a dedicated compute server for Yarn deployments, do the following:

1. On the navigation pane, click Utilities > Manage Configuration Files.

2. On the Manage Configuration Files page, click the Kyvos folder to expand, and then click Query Engine > /kyvos/queryengine/bin/ > start-yarn-cs.sh

3. Download the start-yarn-cs.sh script, and the following parameters (other parameters in this file are automated) can be modified with their default values. To modify these defaults, update the values of the following parameters:

   1. Change Application Master Name,

      1. APPLICATION_NAME: Ensure that the application name (Dedicated compute) is unique across Kyvos clusters running on the same Hadoop cluster.

      2. APP_MASTER_LOG_PATH

   2. Update containers default values in which Kyvos Compute Server will be launched as Java application,

      1. export CS_CONTAINER_VCORES=8

      2. export CS_CONTAINER_MEMORY=16384

      3. export CS_COUNT=1

   3. Update Kyvos Compute Server configurations,

      1. CS_HEAP_MEMORY="4096M"

      2. CS_OFF_HEAP_MEMORY="2048M"

      3. CS_ANALYTICAL_STORE_BUILD_MEMORY="6GB"

      4. CS_ANALYTICAL_STORE_MAX_TASK_COUNT=2

      5. AUTO_CONFIGURE_PROCESSING_TASK_COUNT="AUTO"

4. Upload the start-yarn-cs.sh script.

5. Click the Save button from the top-right of the screen.

6. Restart the Analytical Server, then trigger a sanity semantic model job. By default, Kyvos will launch 1 Compute Server with the following specifications:

   • 16 GB Memory

   • 4 GB Heap

   • 2 GB Off-heap

   • 6 GB Analytical Store Build Memory

   • 2 Tasks

When exporting a worksheet to CSV and PDF formats, you can include metadata in the report. A separate text file, following the csvMetadataTemplate.txt template, is also exported and can be modified as needed.

To modify the CSV metadata template, perform the following steps.

1. On the navigation pane, click Utilities > Manage Configuration Files.

2. On the Manage Configuration Files page, click the Kyvos folder to expand, and then click BI Server> /kyvos/olapengine/conf/dataExport/en_US/ > csvMetadataTemplate.txt

3. Download the template file. In the file, you can view the following:
   
   

   Open

   

4. Make the changes in the template file as needed and save it.

5. Upload the template file.

6. Click Save. The template file will be updated with your changes.

Open

From Kyvos 2026.3 onwards, you can manage custom PPT templates through Kyvos Manager. You can upload, download, and delete custom PPT templates in the custom\ppt folder under the client\templates directory.

To add a custom template file, perform the following steps.

1. On the navigation pane, click Utilities > Manage Configuration Files.

2. On the Manage Configuration Files page, click the Kyvos folder to expand, and navigate to Web Portal > kyvos/jakarta/webapps/kyvos/client/templates/custom/ppt/.

3. Click Add File next to thecustom\pptfolder.

Open

4. In the dialog box, click Choose File, select the file you want to upload from your system and then click Add. A “File uploaded successfully” message will be displayed.

Open

6. Click Save and enter your Kyvos Manager password.

7. Refresh the web page to view the newly added file in the list.

After a file has been added to the custom\ppt folder, you can use the following actions by clicking the corresponding links displayed next to the file:

• Upload: Replace the file with a new version.

• Download: Download the file to your local system.

• Delete: Delete the file from the folder.

You can customize the commands executed by the system monitoring framework by modifying the systemconfig.json file.

To configure tier-based monitoring commands, perform the following steps.

1. Log in to Kyvos Manager.

2. Navigate to Utilities > Manage Configuration Files.
   
   

   Open

   

3. Expand Kyvos > BI Server > kyvos/olapengine/conf/.

4. Locate the systemconfig.json file.

5. Click Download to download the configuration file.

6. Open the downloaded systemconfig.json file in a text editor.
   
   

   Open

   

7. Locate the system_monitor section. The file contains monitoring commands organized into different tiers (for example, tier0, tier1, tier2, tier3, and tier4).

8. Add, remove, or modify commands under the required monitoring tier based on your diagnostic requirements.

9. Save the updated systemconfig.json file.

10. Return to the Manage Configuration Files page in Kyvos Manager.

11. Click Upload next to the systemconfig.json file and upload the modified file.

12. Click Save to apply the changes.

Kyvos supports integration with an Application Performance Monitoring (APM) platform to monitor application performance and health. When enabled, telemetry provides end-to-end distributed tracing across Kyvos components and helps you troubleshoot application and communication issues.

Telemetry supports the following capabilities:

• End-to-End Distributed Tracing: Tracks requests across the Web Client, Analytical Server, Query Engine, Compute Master, and databases by maintaining the parent-child trace hierarchy.

• TCP Communication Tracing: Creates child spans for proprietary TCP communication and associates them with the corresponding parent spans.

• Span Details: Captures relevant information, such as the host name and port, for TCP communication spans.

Before enabling telemetry, obtain the APM connection string for the target deployment.

For Azure Application Insights:

1. Open the required Application Insights resource in the Azure portal.

2. On the Overview page, locate Connection string under Essentials.
   
   

   Open

   

3. Copy the connection string for use in the telemetry configuration.

You must update the applicationinsights.json file with the required APM connection string.

To configure the APM connection, perform the following steps:

1. Go to Utilities > Manage Configuration Files.

2. Under Commons, expand the following path: kyvos/commons/config/opentelemetry/

3. The applicationinsights.json file is available under the opentelemetry configuration folder, as shown below.
   
   

   Open

   

4. Download the applicationinsights.json file.
   JSON format: The following is a sample JSON file. For more details about advance configuration, see Microsoft Azure Documentation.

   { "connectionString":"<YOUR_CONNECTION_STRING>", "role": { "name": "KyvosClient" }, "sampling": { "percentage": 100 }, "preview": { "captureHttpServer4xxAsError": false } }

5. In the file, add the APM configuration string that you have copied from Azure Portal.

6. Save the file.

7. Upload the updated file.

8. After updating the telemetry configuration, save the configuration changes in Kyvos Manager and restart Kyvos services for them to take effect.

You can enable or disable telemetry, use the ENABLE_TELEMETRY property to control telemetry:

• Set the ENABLE_TELEMETRY property to yes to enable telemetry. You can upload the OpenTelemetry JAR files.

• Set it to no to disable telemetry.

• The default value is no.

You can also use Kyvos' built-in configuration file to implement RLS, without uploading a custom JAR.

To enable built-in configuration-based RLS Security, configure the RLSConfiguration.json file, located at /kyvos/olapengine/conf/RLSConfiguration.json.

The RLSConfiguration.json file defines the actual row-level filtering behavior for each semantic model — including which users are allowed access when they are missing from the rules table (missingUserBehaviour), the filter expression used to generate the RLS filter (rlsRuleQuery), and the wildcard character that grants full column access (wildCardCharacter).

To enable built-in configuration-based RLS Security, perform the following steps.

1. Navigate to Utilities > Manage Configuration Files.

2. Go to the following location: /kyvos/olapengine/conf/RLSConfiguration.json
   
   

   Open

   

    

3. Download the RLSConfiguration.json file.

4. Update it with the required configuration and upload it back to apply the changes.

5. Save the file.

6. Upload the updated file.

The following is a sample JSON file.

The following table describes the key configuration properties used in the RLSConfiguration.jsonfile.

Use RLSConfiguration.json to define Row-Level Security (RLS) for semantic models. The configuration supports:

 The rlsRuleQuery property defines the SQL condition used for applying Row-Level Security (RLS). It can be any valid SQL condition, and it can reference:

• Semantic model column names or aliases defined through column mapping. 

• Column names from the RLS rule table. 

For example, if region is the semantic model column and rule_region is the corresponding rule-table column, configure:

At runtime, the value of rule_region is substituted into the query. 

• Use parentheses for conditions: When combining conditions with AND or OR, enclose each condition in parentheses.

  "rlsRuleQuery": "(region IN (rule_region)) AND (country IN (rule_country))"

  For mixed conditions, use additional parentheses to define the required logical grouping.

  "rlsRuleQuery": "((region IN (rule_region)) AND (country IN (rule_country))) OR (state IN (rule_state))"

• Enclose string values in single quotation marks: When using multiple string values with the IN operator, format the values as: ('US', 'UK')
  You can generate these values in either of the following ways:

  • Use SQL functions such as CONCAT in fetchRulesQuery to construct the single-quoted, comma-separated values.

  • Store the values directly in the rule table in the following format: 'US', 'UK'

• Use different names for semantic model and rule-table columns: Rule-table column names are replaced with their corresponding values during RLS processing. Using the same name for both columns can result in unexpected substitution behavior. For example:

Use columnMappings when the dimension name used in rlsRuleQuery doesn't match the field name Kyvos actually sends to the provider.

Example: The Kyvos semantic model field is Sales Region (with a space). After normalization (spaces become underscores), the key seen by the provider is Sales_Region. If the rule template also uses Sales_Region, no mapping is needed in most cases.

But if the semantic model field were named something entirely different — say Geo — add a mapping:

This renames Sales_Region to Geo in the template at load time, so the final expression uses the name that actually matches the semantic model’s field.

A single RLSConfiguration.json file can hold configurations for multiple semantic models at once. Each model can define its own rule query, fetch query, and column mappings, while sharing the same database connection settings: