Five nines availability, sometimes called five 9s, is a reference to the percentage of time that a computer system is promised to be accessible and operational. The term is often used to describe products that have been designed for high availability, with features that help it handle unexpected problems on the fly, without crashing or compromising the integrity of the data. A system with five nines availability will be operational 99.999% of the time – which means that in any given year you can expect about 5.26 minutes of total downtime, planned or unplanned.
There are three engineering design principles that help achieve this kind of high availability. First, adding redundancy to eliminate the single points of failure – this will keep any single component from crashing the entire system. Next, ensuring there is a reliable crossover between the redundant systems, to prevent the crossover point itself from becoming the single point of failure. And last, making sure failures are detectable as they occur. A recent survey revealed that the top reason for downtime of enterprise IT systems was the failure to monitor the relevant components. Another important factor in achieving five 9s is reducing the reliance on human intervention. The most common cause for unplanned system downtime is human error.
Tools such as SUSE Linux Enterprise High Availability Extension are designed for five 9s availability and can virtually eliminate unplanned downtime using high-availability clustering technologies that can work in physical and virtual environments.