Chapter 8. IBM BladeCenter HX5 385
Figure 8-16 Power consumption in one power domain
A further consideration is when a 2-node or HX5+MAX5 configuration straddles the two
power domains. If you install the primary blade server of a scalable blade complex in blade
server bay 7 of a BladeCenter H chassis, the second node or the MAX5 is installed in bay 8.
Therefore, the scale is split between two separate power domains.
The following situations can occur if there is a power loss to either power domain, depending
on how the scalable blade complex is implemented:
A loss of power to power domain 1 results in both blade servers in the scalable blade
complex going down. If this configuration is a 2-node configuration, you also might get
errors on the reboot and the light path diagnostics reporting that a non-maskable interrupt
(NMI) error has occurred. This result is normal due to losing processors abruptly. The
servers will not come back up until the power is restored or the scalability card is removed.
If the scalable blade complex is implemented in stand-alone mode, a loss of power to any
power domain will still result in both blade servers going down. The blades will not come
back up until either the QPI card is removed or the power is restored.
8.4 Local storage considerations and array setup
In an HX5 2-node configuration, you can choose to use the drives on the primary node, the
secondary node, or both. In this section, we describe creating an array in the LSI 1064. For
drive option and ordering information, see 5.11, “Storage” on page 203.
You can create an array on the primary node, the secondary node, or both. IBM recommends
that the operating system is installed on the primary node for maximum performance.
To Next Page
Loading...
Need help?
General