Providing Open Architecture High Availability Solutions

1.0 Executive Summary

This document describes the best known methods and capabilities needed for systems that are

required to have almost no down-time, frequently referred to as high availability (HA) systems. It

is intended as a guide to a common vocabulary and possible HA functions. The system designer

must select the appropriate functions for each system based on HA requirements, design

complexity and cost.

The document was written by the HA Forum, an industry group with the goal of increasing the

number and capability of open architecture high availability systems by standardizing the

interfaces and capabilities of building blocks for HA systems. Where blocks are missing the HA

Forum will look for solutions to fill the blocks. In addition, the Forum intends to promote solid

development models and best known methods for providing high availability systems.

By providing standards and guidelines for open architecture HA systems, development of these

systems will become easier. Additionally, systems will be better because developers will be able to

focus on their products instead of on low level services required to support high availability.

Three major sections comprise this document: Introductory Material, Functional Capabilities, and

Capabilities of Major Building Blocks. A brief summary of each section is given below:

Introductory Material

The introductory material consists of a document introduction, a review of the concepts and

principals of developing and maintaining an HA system, and an overview of typical customer

requirements for high availability systems.

Concepts and principles of HA are discussed in Section 3.0. In this section, the fundamental

principles of engineering HA systems and the issues involved in the design, development and

deployment of these systems are discussed. Included are some industry best practices to mitigate

the risk and increase the success of deploying HA systems. Some modeling techniques are

reviewed to build a quantitative understanding of HA measurement.

Customer Requirements for HA systems are summarized in Section 4.0. This section identifies the

application areas under consideration by the Forum. It reviews at a high level the requirements,

expectations, and desirable features that typical applications may have of an HA system.

Functional Capabilities

Functional capabilities for HA are separated into those needed to configure and operate the system

during normal operation and those that are used to detect and handle faults.

Managing the system configuration is discussed in Section 5.0. This involves knowing what

hardware, firmware, and software components are intended to be in the system (the system model)

and which of these are actually in a system at any given time. Configuration Management also

allows modification of the configuration of the individual components that comprise a system.

Fault management, as discussed in Section 6.0, is typically a five-stage process:

1. Detection – The fault is found

2. Diagnosis – The cause of the fault is determined

3. Isolation – The rest of the system is protected from the fault