As we said previously, Availability is the ratio of uptime to uptime plus downtime. Improving availability means either increasing uptime, decreasing downtime (or both). It is most important to note that any form of fail-over HA clustering can only decrease downtime, it cannot increase uptime (because the failure will mostly be visible to clients). Thus, we describe how to achieve up and down time improvements.