High Availability
The quest for High Availability, or HA, has become a little likened to the search for the elixir of life.
As reliance on computing systems has grown, so has the need to supply the provision of a level of service required.
In computing, availability is usually reckoned to be the period of time required for the system to respond to users, say, a one second response time, or the duration of when the services are available, say, 18 hours per day, 6 days per week.
In the search for system perfection, no matter how reliable the software and systems are, problems can occur that can bring down ones server or applications.
High Availability systems are built to recover automatically from server or component failure.
The HA system in its simplest form, is easiest described as back-up after back-up. For instance if the system consists of two identical redundant web servers behind a load balancer, the incoming traffic will be equally divided between the web servers, but if one of these servers fails, the load balancer will redirect all traffic to the functioning online server.
Since any component may fail, the goal is to design systems which can predict and isolate problems before a failure happens, and in which, should that failure happen, it is quickly detected and corrected.
In the present business environments, High Availability computing is becoming the requested norm, rather than an optional bonus. It provides insurance against downtime leading to loss of business and allows the provider to offer competitive and superior customer services.
With the growing demand for ’24-7’ computerised services, HA can provide new business opportunities particularly for those which depend in depth on computers.
The expenses that HA incurs are gradually falling as the technology to minimise the likelihood of component or system failures trickles its way towards mainstream.
Leading high availability providers include independents & the largest IBM maintenance providers outside of IBM themselves.
With their own, carbon offset data centres, high availability is the expected norm rather than a bespoke add on.