Providing Open Architecture High Availability Solutions

5.0 System Capabilities – Configuration Management

5.1 Introduction

Configuration management involves knowing what types of hardware, firmware, and software

components are actually in a system. It also tracks the intended configuration of the system (the

system model), which may or may not match the actual system configuration. Finally,

configuration management contains the capability to modify the configuration of the individual

components that comprise a system.

In order to provide accurate fault management, one must know the intended make-up of the

system. The actions taken by a fault management system typically require changes to the system

configuration, and this requires corresponding capabilities in the configuration management

strategy. This is attained by maintaining an inventory of the system’s components (hardware,

firmware, and software) and by accurately maintaining information on the health of each

component.

Although the title of this section is “System Capabilities – Configuration Management”, the

capabilities discussed here are only the sub-set of configuration management activities that are

necessary in support of high availability. Details of fault management are outlined in Section 6.0.

5.2 Characteristics of System Components

System components are the hardware, software and data that comprise the system. A component is

any part of a system that is individually managed. In infrastructure equipment, components

(managed or not) typically include:

• Hardware – Disk drives, boards, fans, switches, cables, chassis, lights, and power supplies

• Software – Operating systems, drivers, servers, content, and applications

• Logical Entities – Databases, network topologies, routes, clusters and geographical regions

To lay the foundation for the sections that follow, there are six characteristics of system

components that should be addressed in modeling and controlling a system:

• Heterogeneity. System components include hardware (boards, drives, fans, LEDs, etc.) and

software (O/S, drivers, applications, middleware, etc.)

• Transience. System components are available or not available, according to either their health

or their presence in the system due to a service event or upgrade

• Dynamics. The state of system components is typically dynamic and includes attributes such

as health, operational state, and administrative state

• Controllability. Certain system components can be controlled from other parts of the system

or through external system interfaces

• Dependency. System components typically have dependencies upon one another. For

example, an application depends on the O/S, which in turn depends on the hardware.

• Redundancy. Some critical system components are redundant and are represented as logical

groups that provide a single service within the system. For example, a system may have

redundant fans, CPU cards and power supplies so that the system can continue to operate in

the event that an individual component fails.