Knowledge Base Docs
8 members
7 discussions
Specific guides written by Sisense experts and trusted partners.
- Discussion
Eiji Ozawa
AdminIntroduction: Sisense supports row-level security (RLS) rules on data models, allowing administrators to control data access at a granular level per user and user group. However, in some environments, RLS rules may be applied inconsistently — data that should be restricted is sometimes exposed, particularly when toggling field visibility in a widget. This behavior is not caused by rule conflicts, but by insufficient system resources — specifically, an undersized DataContextRulesCacheSize and limited memory allocation — which cause the security rule engine to behave unpredictably. This article explains how to resolve this issue by adjusting the relevant settings in the Configuration Manager. Step-by-Step Guide: Check the number of existing data security rules. Self-hosted environment Run the following command: kubectl -n sisense exec -it `kubectl -n sisense get pods | grep sisense-mongodb | head -1 | awk '{print $1}' ` -- mongosh prismWebDB --eval 'db.dataContext.find().count()' Sisense Cloud environment Connect to Repository from ElastiCube data model using MongoDB connector and execute query. Use the following connection settings in the MongoDB connector. Location: Single-Node: sisense-mongodb.sisense Multi-Node: sisense-mongodb-headless.sisense Port: 27017 Default Database: prismWebDB Authentication Mechanism: No Authentication Define the following table query: SELECT Count(*) AS NumberOfDataSecurityRegistrations FROM "prismWebDB"."dataContext" Preview data of the query and check the returned value Check the DataContextRulesCacheSize parameter value Log in to Sisense as a user with Admin privileges. Open Configuration Manager : Admin > Server & Hardware > System Management > Configuration Click the Logo in top left five times to open the left pane, and select Query in the left pane Check the DataContextRulesCacheSize.value parameter, the default value is 1000 Update the DataContextRulesCacheSize parameter value greater than the value obtained in Step 1 Update MemoryLimit parameter value (default: 1024m) to 2G or greater value If you want to increase the MemoryLimit parameter more than 2G, you have to update the memory limit of the query deployment, which should be greater than Memory Limit + 1G. Save the changes and restart the relevant services. Verify that data security rules are now applied consistently across all widgets. Conclusion: Once these configuration changes are applied, the RLS rules should be enforced reliably. Increasing the DataContextRulesCacheSize ensures the system can cache a sufficient number of data context rules without dropping or misapplying them under load. Add a disclaimer for custom solutions . Disclaimer: This post outlines a potential custom workaround for a specific use case or provides instructions regarding a specific task. The solution may not work in all scenarios or Sisense versions, so we strongly recommend testing it in your environment before deployment. If you need further assistance with this, please let us know.
- Discussion
Ihor Yokhym
AdminIntroduction In multi-tenant Sisense, a Data Designer may hit intermittent 3404 "Connection is Missing" build errors that reference the wrong tenant's connection, while an admin builds successfully. This guide explains why connection scoping causes it and how to fix it. Step-by-Step Guide Why this happens In multi-tenant Sisense, connections are isolated per tenant and per owner. The 3404 "Connection is Missing" error at build time means the build service cannot resolve the connection OID within the scope of the user running the build. When a Data Designer does not own or cannot see the connection the cube references, resolution falls back and intermittently binds to the wrong tenant's connection - which is why the reported missing connection changes between build attempts with no changes to the cube. An admin succeeds because admin visibility spans all connections. Step 1 - Share the connection with the Data Designer As an admin, go to Admin → Data → Connections. Open the connection the cube uses and share it with the Data Designer user. Grant access with edit rights so builds can resolve it. Repeat this for every connection the cube references. Step 2 - Verify tenant scoping while sharing Confirm the Data Designer is operating in the same tenant the connections belong to. Also confirm that no same-named connection exists in another tenant, since duplicate names across tenants can cause the build service to bind to the wrong one. Step 3 - Refresh the connection binding (if sharing does not clear it) If the error persists after sharing, re-bind the connection at the table level. For each affected table: Change Table → re-parse the SQL expression → Done. This forces Sisense to re-resolve the connection binding for that table. Conclusion Cross-tenant 3404 “Connection is Missing” during builds are almost always a connection-scoping and permissions issue rather than a broken cube. Sharing every referenced connection with the building user, verifying correct tenant scope, and re-binding table connections resolves the error. The key takeaway: a non-admin build user must have visibility of every connection the cube depends on. References/Related Content Connecting to Data / Connection Manager Sisense Multitenancy Building ElastiCubes
- Discussion
Ihor Yokhym
AdminIntroduction Resolve the Snowflake JDBC driver "HTTP request: null" communication error in Sisense (Cloud and On-Premises). This guide outlines how to upgrade the Snowflake JDBC driver to version 3.24.2 or later to bypass a known NullPointerException bug. Step-by-Step Guide Symptom When attempting to connect to Snowflake, the connection fails, and the logs display the following error message: Code Type: Error Message JDBC driver encountered communication error. Message: Exception encountered for HTTP request: null. This behavior stems from a known issue within the Snowflake JDBC driver itself (SNOW-1958601: NullPointerException for cacheFile in net.snowflake:snowflake-jdbc:3.23.0 ). The issue is resolved in connector version 3.23.2 and later or upgrade to Sisense version 2026.2.1+ . To completely safeguard your environment, follow the steps below to upgrade to version 3.24.2. Step 1: Download the Updated JDBC Driver Download the stable snowflake-jdbc-3.24.2.jar file or later directly from the official Snowflake repository or Maven Central. Maven repository: download driver Snowflake documentation Step 2: Upload the Driver to File Management Navigate to the File Management section on your Sisense server and upload the newly downloaded .jar file to the target Snowflake connector directory. /connectors/framework/SnowflakeJDBC/ [Screenshot: Sisense File Management interface showing the SnowflakeJDBC directory] Description: The image shows the Sisense File Management browser with the snowflake-jdbc-3.24.2.jar file successfully uploaded into the connectors framework directory. Step 3: Update the Connector Configuration Open the Sisense Admin portal. Navigate to Service Configuration → Manage Connectors → SnowflakeJDBC. Update the configuration properties to point directly to the new driver version ( 3.24.2 ). [Screenshot: Service Configuration UI for Manage Connectors] Description: This screenshot highlights the driver version field within the SnowflakeJDBC configuration menu where the version number needs to be updated. Step 4: Restart the Query Service To apply the changes and force Sisense to initialize the new JDBC driver, restart the Query service. Conclusion A known NullPointerException bug in the 3.23.0 Snowflake JDBC driver causes connection handshakes to fail with a generic null HTTP request error. Upgrading your Sisense environment to use the 3.24.2 driver patch (or later) via Service Configuration and executing a Query service restart successfully resolves the communication roadblock. References/Related Content Maven repository: download driver Snowflake documentation GitHub Issue #2105: snowflakedb/snowflake-jdbc NullPointerException
- Discussion
Ihor Yokhym
AdminIntroduction: Learn how to successfully connect Sisense to Amazon DocumentDB using a generic JDBC connection. This guide applies to Sisense Linux editions (both Cloud and On-Premises deployment models) and resolves common TLS-related connection reset errors. Step-by-Step Guide: Step 1: Whitelist the Sisense Instance Before modifying any configurations in Sisense, verify that your Sisense instance IP address or network range is properly whitelisted in your AWS DocumentDB cluster's security groups. The default connection port is 27017 . Step 2: Download the driver Get the JDBC Driver for Amazon DocumentDB from the Amazon website and Github: AWS DocumentDB: Documentation AWS DocumentDB: Driver Github Step 3: Register the connector in Sisense 2.1 Go to Admin > Server & Hardware > System Management > Configuration . 2.2 Scroll to the bottom and click Manage Connectors . 2.3 In the left pane, click +JDBC Connector and fill in: Connector ID: [YOUR_NAME] Display Name: [YOUR_NAME] Class Name: software.amazon .documentdb.jdbc.DocumentDbDriver Upload JAR File: upload jar file Click Create . Step 4: Use the Generic JDBC Connection String Because of architectural differences, connecting via the native Sisense MongoDB connector can result in errors. Instead, connect using the generic AWS DocumentDB JDBC driver syntax. JDBC Connection String Template jdbc:documentdb://<hostname>:<port>/<database-name>?<options> Step 5: Configure TLS Settings and Resolve Socket Exceptions According to official AWS DocumentDB JDBC Driver documentation, the driver’s default behavior is tls=true. If your DocumentDB configuration expects plain text, a protocol mismatch occurs: the JDBC driver attempts an encrypted SSL handshake with a database node expecting unencrypted packets. DocumentDB will abruptly drop the network socket, resulting in the following error: java.net.SocketException: Connection reset To fix this, modify your connection string based on your database environment: For Non-TLS Environments: Explicitly turn off TLS encryption by adding tls=false and completely removing the tlsAllowInvalidHostnames flag. For TLS Environments: Ensure both tls=true and tlsAllowInvalidHostnames=true are active. Code Type: JDBC Connection String Example (TLS Disabled) jdbc:documentdb://<a target="_blank" rel="noopener noreferrer nofollow" href="http://my-cluster.cluster-xyz.us-east-1.docdb.amazonaws.com:27017/my_database?tls=false"><u>my-cluster.cluster-xyz.us-east-1.docdb.amazonaws.com:27017/my_database?tls=false</u></a> [Screenshot: Generic JDBC connection setup window in the Sisense Data tab] Description: The image above illustrates where to input the customized DocumentDB JDBC connection string and credential details within the Sisense Data Connector UI. Conclusion: Successfully connecting Sisense to Amazon DocumentDB requires leveraging the generic JDBC driver framework rather than the native MongoDB connector. The most common pitfall is a TLS mismatch; explicitly declaring tls=false in your connection parameters will bypass the java.net .SocketException when working with unencrypted database nodes. References/Related Content Sisense Documentation: Deploy Custom JDBC connector AWS DocumentDB: Documentation AWS DocumentDB: Driver Github