Estimation of the Availability of a Complex Repairable System within a Fixed Time Horizon

A complex IT system is an IT system composed of interconnected parts (components) that as a whole exhibit one or more properties not obvious from the properties of the individual parts (components). In the area of IT governance, Service Level Agreements (SLAs) are used to support the planning and management of IT services (i.e., services provided by complex IT systems) in a collaborative way, to maximise the usefulness of these systems while keeping under control the cost of service provision and maintenance. One of the mostly used Quality of Service (QoS) parameters used within SLAs is availability. The availability level for an IT service can influence (a) the business operations enabled by the IT service, or (b) the provisioning of other IT services depending on it. A high availability IT service guarantees that the IT services depending on it, as well as the business operations the service is supporting, can be performed at best without interruption. On the other hand, availability has a cost. An measure for the availability of a repairable system, useful in this context, is the probability that the system availability will be below a certain level (e.g., P (As ≤ 0.995)). This measure can be determined for one single system component in a fixed time horizon (e.g., one month), by the combination of three parameters: