In the architecture of a server solution, supply of electrical power is a single point of failure. If electrical power is lost, servers can fail without warning and software corruption can occur even with journaled file systems. UPS devices protect from voltage spikes, surges, overvoltage, undervoltage and other nastiness. They protect the most expensive and critical elements of IT infrastructure. A UPS also gives critical systems enough time to gracefully shut down and switch off before the UPS batteries are exhausted. If possible an email message can be sent to the outside world before everything goes dark.

Systems can also be configured and scripted to return automatically once power has restored and a minimum battery capacity has been reached. Unfortunately, even where a UPS has been purchased, the majority of installations I have come across have little or no configuration in place. It is not just a case of buying a UPS off the shelf and plugging it in. Armed with a little more information and a bit of extra time, a UPS can be used to offer much better protection from power failure events. This document is intended to assist with specifying and deploying UPS for Apple server installations. Implementation Process ” Identify devices requiring UPS protection; ” calculate power draw; ” estimate run time requirements; ” specify required UPS battery capacity; ” select manufacturer and model; ” install on site with appropriate electrical wiring; ” configure UPS communications and software for graceful shutdown and restart sequences; ” scheduled testing; ” document procedures.

Download pdf for Uninterruptible Power Supplies (UPS) Mac OS X Integration Guide Manual