Continuous availability is a standard, or goal, for many information technology (IT) professionals, especially server administrators, with regard to data access. This standard means that data on a server or other service can be accessed at all times, even if some type of error or malfunction occurs with software or hardware. Such availability can be implemented in a number of different ways, often involving redundancy in hardware and software, backup systems in case of failure, and software that handles transitions between systems invisibly. Continuous availability has become increasingly important as vital and non-vital businesses have become more dependent on digital information and technology.
The primary goal for continuous availability is that data or a service is available to customers and employees at all times. An online retail business, for example, wants such availability to ensure that customers can access the company’s website and complete purchases at all times. Airline traffic controllers need continuous availability for systems that track flights to ensure airplanes are able to safely navigate the skies and land and take off without issue. There are a number of ways in which network administrators and other IT professionals strive for this type of availability.
Redundancy is, perhaps, the key concept in creating a system with continuous availability. This means that for each software and hardware component of a system, there is at least one backup that can be utilized in case the primary component fails. If the fans for a bank of servers fail, for example, it can lead to the entire server overheating and crashing, which then causes a loss of availability for clients connected to that server. By having backup fans for a server, however, continuous availability can be better maintained. Other measures can then be used to notify technical support that the fans have stopped working, so they can be repaired while the backup fans are used.
Continuous availability is often measured as a percentage of time a server or system is available. This is usually in terms of one year, in which 99% availability would mean the system is down for 87.6 hours each year, while a system with 99.99% availability is only down for about one hour each year. While only one hour of downtime per year may seem like very high availability, one hour of downtime for an air traffic control system could be catastrophic. Downtime is caused by either planned or unplanned outages, which refer to times when a system may be down for scheduled maintenance or is down due to a failure within the system.