Implementing High Availability in the Processing Tier

Objectives

After completing this lesson, you will be able to:

  • Implement high availability in the processing tier
  • Use System Configuration Wizard to implement high availability

Servers for High Availability

High availability can be achieved in the processing tier by including more servers in your deployment.

To achieve high availability in the processing tier, add one or more of the following servers to your organization’s deployment:

  • Adaptive Processing Servers of the type of service(s) needed for your implementation

  • Cache Servers (Crystal Reports)

  • Crystal Reports related Processing Servers

  • Web Intelligence Processing Server

Adaptive Processing Servers (APSs)

Note
Selecting a deployment template in the wizard or manually creating more APSs doesn't replace system sizing. For more information about BI sizing:

Sizing Guide

System Configuration Wizard

The wizard helps you with the basic configurations of your BI platform system, including configuring APSs according to set deployment templates. The APS configuration given by the wizard is a good starting point; however, system sizing must still be performed.

The wizard is available from the Central Management Console (CMC).

For a detailed description of the deployment template, click the deployment template link on the Capacity page of the wizard to open the DeploymentTemplates.pdf file.

Features of the System Configuration Wizard

The following are features of the System Configuration Wizard:

  • Simple workflow to do initial server configuration of the BI system

  • Command-line facility to enable preset configuration

  • Installation of more APSs that can be configured to meet requirements

Uses for APSs

The installation program installs one APS per host system. Depending on the features that you've installed, this APS can host many services, such as the Monitoring Service, Promotion Management Service, Multi-Dimensional Analysis Service (MDAS), Publishing Service, and others.

A generic server hosts a service responsible for processing requests from various sources.

For production or test systems, the best practice is to create more APSs, and then configures the APSs to meet your organization’s business requirements.

Deployment Template Overview

Reasons for Changing from the Default Deployment Size

The following are considerations when deploying the BI platform:

  • The default small deployment is good only for a demo environment on limited hardware.

  • You can choose a preset deployment template to match your hardware and use cases.

  • Although the template helps with the initial configuration, it doesn't replace system sizing and configuration efforts.

The default installation of the BI platform configures a small deployment that is good for a demo environment on limited system hardware. To match your hardware and wanted use case (for example, preparing a test system or production system), choose one of the preset deployment templates from the Capacity view. These templates are meant to help the BI platform system up and running quickly, and to shorten the initial deployment time.

Choosing a proper deployment template helps with the initial configuration. But, it is not a replacement for system sizing and configuration, which must be performed. For better performance, size your system by referring to a sizing guide:

Sizing guide

Considerations When Choosing a Template

  • The request handling capacity of your system affects the size of the deployment template that you select.

  • Choosing a large deployment doesn't guarantee better performance.

  • The deployment template matches your business needs and your available hardware resources.

  • Larger deployment templates give a good grouping of services.

Choosing a proper deployment template is important for several reasons:

  • The request handling capacity of your system affects the deployment template that you select. A larger deployment gives capacity to handle more requests or more complex requests. However, a larger deployment needs more system resources.

  • Choosing a large deployment doesn't guarantee better performance, especially when you don't have enough hardware resources.

  • The deployment template matches your business needs and your available hardware resources. If you choose a deployment template that is too small for your business needs, or too large for the available hardware resources, the system will suffer reduced capacity and performance.

  • Larger deployment templates give a good grouping of services. Failures in one product are unlikely to affect other products. Choose a template that balances resource (RAM) usage and performance. For example, if a large amount of RAM is available, you want to pick the deployment template that your RAM allows. This template will result in a good system grouping of services.

Types of Deployment Template

The following types of deployment templates are available:

  • XS: This extra small deployment is used as a demo system. The XS template is an extra small deployment. This deployment needs 6–8 GB of RAM, 8 GB of Host RAM, and uses one APS.

  • S: This small deployment is used as a development system. The S template is a small deployment template. This deployment needs 12–16 GB of RAM, 16 GB of Host RAM, and uses four APSs.

  • M: The medium deployment is used as a test or small production system. The M template is a medium deployment template. This deployment needs 15–25 GB of RAM, 16–32 GB of Host RAM, and uses 7 APSs.

  • L: This large deployment is used as a production system. The L template is a large deployment template. This deployment needs 30–45 GB of RAM, 32–48 GB of Host RAM, and uses 9 APSs.

  • XL: This extra large deployment is used as a large production system. The XL template is an extra large deployment template. This deployment needs 40–60 GB of RAM, 48–64+ GB of Host RAM, and uses 11 APSs.

Configure the Processing Tier for High Availability

Log in to track your progress & complete quizzes