Version 24.3

1. Overview

Dundas BI is a web-based server application that can run on Windows or Linux or in containers, using SQL Server/Azure SQL or PostgreSQL databases for storage. Users can then log on from any device using a supported web browser and connect to a variety of other databases and data sources, which may require additional drivers to be installed.

2. Client

One of the following is required for client-side devices, and a recent version is recommended:

Desktop:

• Chrome
• Firefox
• Safari
• Microsoft Edge

Mobile:

• iOS
• Android - Android Chrome browser only

3. Server

Below are the hardware and software requirements for installing Dundas BI on your own server, computer, or cloud resource.

Processor:

• 1.5 GHz - minimum for evaluation or testing
  
• 2.0 GHz - minimum for production use
• x64 architecture (e.g., AMD64 or Intel 64)

At least two logical cores are required for in-memory storage.

RAM:

• 4 GB - minimum for evaluation or testing
  
• 8 GB - minimum for production use
• 16 GB - minimum for production use with in-memory analytics

Hard disk space:

• 10 GB free - minimum for evaluation or testing
  
• 50 GB free - minimum for production use
  

Operating systems:

The following 64-bit operating system versions are supported:

• Windows 10 or higher - evaluation or gateway use only
• Windows Server 2012 R2 or higher
• Red Hat Enterprise Linux (RHEL) 8 - upgrade to version 24.2 or higher for Python features
  
• Ubuntu 20.04 LTS
• Ubuntu 22.04 LTS
• Debian 11

You can also deploy Dundas BI on other distributions using Docker and Kubernetes.

There are certain known limitations when running on Linux, including Docker and Kubernetes.

.NET:

• .NET 8 - installed automatically for a Dundas BI instance
  

Web server:

Dundas BI instances are hosted on one of the following:

• For Windows: IIS 8.5 or higher - installed automatically as a prerequisite
• For Linux: a reverse proxy server such as NGINX - can be installed automatically

Application database server:

One of the following is used by an application instance to store its own data:

• Microsoft SQL Server 2017 or higher
• Microsoft SQL Server Express Editions - supported but not recommended for production use due to limitations; Microsoft SQL Server Express LocalDB is not supported
• Microsoft Azure SQL
• PostgreSQL version 9.6 or higher

This can be on the same server or separate from the web server (the database server must be configured to accept remote connections if separate).

4. Data sources

You can connect to the following types of data sources, each of which may have additional (driver) requirements for the server or gateway as indicated below. Kubernetes and Docker deployments rely on Linux support.

The ODBC and JDBC data sources listed below allow you to connect to various other data sources that have ODBC/JDBC drivers. The Generic Web data provider can connect to various online data sources.

Additional details about installing and using drivers for certain data sources are available in the How-To section of the documentation on connecting to data.

5. Installation

The following requirements apply when installing a Dundas BI instance directly on your own servers.

5.1. For Windows

The Dundas BI deployment application checks if the following prerequisites are installed on the server and can automatically download and install any missing prerequisites (this requires an active internet connection).

If you need to install them manually, use the links below to access them. In the Prerequisites screen of the Deployment application, you can also click on the prerequisites and press F8 to download a list of URLs.

• IIS (on Windows Server, add the Web Server role)
• ASP.NET Core Runtime 6.0 - Windows - x64
• ASP.NET Core Runtime 6.0 - Windows - Hosting Bundle
• .NET Desktop Runtime 6.0
• (Optional) Microsoft Access Database Engine 2010 Redistributable x64 for Microsoft Access and certain Microsoft Excel legacy support such as for XLS files
• (Optional) Python 3.9 for the metric set Assistant and Python transforms. The Assistant also requires the following packages (please see the Deployment application's prerequisites list for version numbers):
  • Python nltk module
  • Python pandas module
  • Python requests module
• Visual C++ 2015, 2017, 2019, and 2022 Redistributable - x64

5.2. For Linux

Certain requirements are installed automatically when running the Dundas BI deployment wizard (this requires an active internet connection). You can install the following ones manually using the provided links:

• ASP.NET Core Runtime 6.0
  • For RHEL
  • For Ubuntu
  • For Debian

5.3. Additional requirements

During installation, you may be asked for the following (unless you choose an Express installation option):

• The user ID and password of an account that has db_create permissions on your SQL Server instance, or a usesuper account on PostgreSQL.
  Note

  After the installation, the SQL Server administrator will require only db_owner permissions. A PostgreSQL administrator will require the CanSelect, CanUpdate, CanInsert, and CanDelete table privileges, as well as the CanCreate database privilege.

• When using a PostgreSQL database for application and warehouse storage, a PostgreSQL instance with the uuid-ossp extension installed (you will be notified if it is missing). This is part of the postgresql-contrib package.
• For Windows installations, the user ID and password of an account to use for the IIS application pool.
• A web server port for accessing the Dundas BI website (the installer may choose a default) or an existing website for installing Dundas BI as an IIS virtual directory (this can be the IIS 'Default Website').

6. See also

• Installing Dundas BI
• Installing Dundas BI on Linux
• Using Dundas BI Docker containers
• Orchestrating Dundas BI with Kubernetes and Helm charts
• Using a Dundas BI gateway
• Dundas BI - Release Notes
• Connect to Data: How-To