ABSTRACT
The phrase high availability (HA) begs the question of what is normal or low availability. For noncritical applications, an availability of 95% may be perfectly acceptable, for example, for an organization’s internal staff holiday/vacation scheduling system, whereas 99.5% may be totally unacceptable for others, for example, airline reservations. According to internal data corporation (IDC), a system is considered to be highly available if, when failure occurs, data are not lost and the system can recover in a reasonable amount of time. What is reasonable will, of course, depend on the criticality of the application. What this definition does not cover is how often failures can be tolerated and in what, which raises the question high availability of what? The availability of the central processing unit (CPU), the storage subsystem, the whole system, the building containing the information technology (IT) equipment, the staff, and so on?
