Datasheet

“main” (Installation and Administration) — 2004/6/25 — 13:29 — page 396 — #422
i
i
i
i
i
i
i
i
Figure 19.1: A Simple High Availability Cluster
hit by disaster or power failures. The environmental conditions of the
servers should also be taken into account — (redundant) air condi-
tioning systems are essential.
Hardware Even the most sophisticated software cannot produce a high
availability system without the greatest possible security from failure
on a hardware level. The key hardware components that should be
considered and laid out with the greatest possible redundancy are:
Power Supply If possible, secure your servers using a UPS (uninter-
ruptible power supply) to ensure that a brief power failure can
be bridged and the systems can be shut down correctly in the
event of a longer power failure. The power supply should also
be configured for redundancy.
Network Interfaces Make sure each of your systems has several net-
work interfaces. If one interface fails, another must automatically
396 19.3. Components of a High Availability Solution