Continuous availability concepts Computer outages are less and less accepted in today's data centers. More and more business applications have to be online 24 hours a day and 365 days a year. Outages can have a serious impact on the performance of an enterprise and should be kept to a minimum. Sometimes outages are caused by hardware, software, and data maintenance work, which needs to be performed at regular intervals. Such planned outages are usually scheduled at a period of low activity. Sometimes outages are caused by faulty hardware, software, and procedures. Such unplanned outages have become quite rare in recent years with advanced multi-user systems. Today, an increasing number of enterprises require that both types of computer outages, planned and unplanned, be eliminated or substantially reduced. An advanced computer system is not a single box, but a complex assembly of such components as processors, peripherals, networks, operating systems, and application software. Consequently, an outage-free system cannot be bought from the shelf. It requires an appropriate systems design, using elements such as redundant and fault tolerant components and appropriate systems management. System availability terms We use three terms to describe the concept of system availability: high availability , continuous operation , and continuous availability . These terms are explained in the following sections. High availability: This is the ability of a system to provide service to its users during defined service periods, at an acceptable or agreed level. These service periods, as well as the definition of an "acceptable" service level, are either stated in a service level commitment by the service provider, or in a service level agreement between end users and the service provider. Typically, a service level above 99.7 percent is accepted as high availability. High availability is maintained by avoiding or reducing any unplanned outages. Continuous operation: This is the ability of a system to provide service to its users at all times, day and night, without scheduled outages to perform system and data maintenance activities. It is obviously difficult to perform change and maintenance work on a system that is supposed to be in continuous operation. However, without preventative maintenance, a system can be in continuous operation, but its availability may not be as high as it should be since it may suffer unscheduled outages more frequently. Even if true continuous operation turns out to be impossible to implement for some applications, a realistic goal might be to increase the defined service period, for example, from 14 hours per day to 18 hours per day. Appendix A. Availability configurations 365