Hitachi Compute Blade 2000 User Manual page 38

Precautions for Safe Use
NOTICE
N + M cold standby function
When the N+M cold standby function is enabled, the power LED on the front panel lights

solid orange, Smart Configure is automatically executed, and then the status LED (CND)
on the server blade lights solid green. Confirm that the power LED of the front panel
lights solid orange to show Smart Configure is completed before pressing the power
button of the server blade for a second or more.
For a system with N+M cold standby configured, you cannot update LPAR manager
firmware versions on all server blades at a time. Then, discrepancies between firmware
versions of server blades are allowed only when the LPAR manager firmware version on
a standby server blade is later than that on the active server blade.
Refer to [Chapter 3: Turn the Power On or Off].
Do not operate the server blade during Smart Configure. If you do, Smart Configure may

fail.
Refer to [Chapter 13: Overview of Smart Configure].
Make sure to apply the same LPAR manager firmware version with the active partition to

the standby partition. If you apply different LPAR manager firmware versions to the active
and standby partition, N+M failover may fail.
Refer to [Chapter 13: Mechanism of the N+M Cold Standby].
Do not set the EFI Shell to the highest booting priority in the EFI Setup menu. If the EFI

Shell is on the top of the boot option, the OS will not successfully boot after N+M
switching and failback.
Refer to [Chapter 13: Procedure for Implementing the N+M cold Standby (1)].
With a standard server blade, executing the Smart Configure changes the SAN boot

priority to the lowest automatically.
Refer to [Chapter 13: Procedure for Implementing the N+M cold Standby (4)].
If you change the LPAR configuration (processors, memory, or device assignment),

make sure to implement [F9]: "Save Configuration" on the LP Menu screen. For details,
see "Save Configuration Changed on LP Screen".
Refer to [Chapter 13: Procedure for Implementing the N+M cold Standby (4)].
When a switching alert is issued by the BSM command, the active partition is forcibly

powered off.
Refer to [Chapter 13: Procedure for Implementing the N+M cold Standby (5)].
xxxviii