Overview

This article is designed to be a high-level overview of the components required for an on-prem deployment within a production environment. We provide recommended sizing guidance, but it is important to understand that these components can be scaled based on your workloads and needs. The following will assist with preparing your on-prem environment:

• 1. Application Server
• 2. SQL Server Database Engine
• 3. Tabular Analysis Services

Preparing your on-prem environment

The below details each resource required along with recommended sizing. Resources can and should be adjusted to support the performance objectives of the business.

1. Application Server

The application server will host the Jet Data Manager application. This application is the data modeling tool used by developers to build the underlying data structures that will support the business.

Requirements for application server:

• Microsoft Supported Windows Server Operating System
• Recommend 16 GB of Memory
• .NET Framework 4.6.2 or higher
• Core XML Services (MSXML) 6.0
• Microsoft SQL Server Native Client
• Microsoft SQL Server System CRL Types
  • Install version that corresponds to the version of SQL Server
  • Not necessary for SQL 2017 editions and above
• Microsoft SQL Server Management Objects (SMO)
  • Install version that corresponds to the version of SQL Server
  • Not necessary for SQL 2017 editions and above
• Microsoft Analysis Management Objects (AMO, ADOMD, and AS_OLEDB)

2. SQL Server Database Engine

The relation database engine will host our data warehouse. This database will hold our report ready data. 

Requirements for SQL Server Instance:

• SQL Server version should be supported by Microsoft
• SQL Server should be licensed
• SQL Server should have the latest service packs / cumulative updates applied 
• SQL Server should have enough disk space to store data and allow for future growth
• SQL Server should have enough memory to allow for efficient operation 
  • Recommend starting at 64-128 GB and scaling based on benchmarks  

3. Tabular Analysis Services

The tabular analysis services engine will host our semantic models. 

Requirements for SQL Server Analysis Services:

• Tabular instance (not multidimensional) 
• Recommend starting memory of 64-128 GB and scaling based on benchmarks