Compaq ProLiant BL e-Class System Overview and Planning

Compaq ProLiant BL e-Class System Overview and Planning 19
Definitions and Theory
Single Sources of Failure (SSOF)
A server blade enclosure (or rack) can fail if it is dependent on a critical component or resource
that is itself vulnerable to failure. This dependence is deemed a single source of failure.
Redundancy is based on the types of failures a configuration can tolerate and is characterized by
removal of as many single sources of failure as possible. This section describes typical single
sources of failure and the redundancy features needed to survive their failure. These single
sources of failure include power supplies, power cords, PDUs, and AC sources.
Power Supply Failure
The power supply may have an internal failure. This term also loosely applies when the supply’s
power cord is unplugged, the supply has an AC source failure, and when the power supply is not
fully inserted into the server blade enclosure.
Power Supply Redundancy
Power supply redundancy is a feature enabling an enclosure to survive the failure of a power
supply. A 1 + 1 redundant enclosure can tolerate a single power supply failure by re-distributing
100% of the load to the remaining power supply.
Figure 7. Power supply redundancy, single source of failure = PDU and AC source
Figure 7 shows an enclosure that can survive a power supply failure or unplugged power cord.
The ProLiant BL e-Class server blade enclosure has power supply redundancy built in, provided
you have two working power supplies, two attached power cords, and stable AC source(s).
Note: The Power Distribution Unit (PDU), PDU power cord, and AC source are each single
sources of failure. Power supply redundancy implies the AC source or PDU supporting the
remaining power supply has sufficient capacity to handle the re-distributed load.