System information

22 Deploying Samba on IBM Eserver BladeCenter
Hot-spare blades
MidPlane supports hot-spare blade servers; this operates in the same fashion as the
hot-spare drive. By creating events within IBM Director, you can deploy an operating system
to a hot-spare blade server automatically.
Processor
In a dual processor blade server, if a CPU fails, the following steps are taken by the system.
1. Force failed processor offline
2. Automatically re-boot server
3. Generate alerts
4. Continue operating with the working processor
Memory
There are four memory slots; memory must be installed in matching pairs. The following
redundancy options are available:
򐂰 Chipkill™ ECC - provides correction for up to 4 bits per DIMM
򐂰 Memory hardware scrubbing - corrects soft memory errors automatically
򐂰 PFA - Creates alerts of imminent failure
Disk mirroring
Refer to Chapter 3, “Storage options for the BladeCenter” of the Redpaper The Cutting Edge:
IBM Eserver BladeCenter, REDP3581 for hardware and software mirroring options of IDE
and SCSI disks.
MidPlane
The middle plane provides connectivity between the blades and the modules at the rear of the
BladeCenter unit. There are two connections on each blade server to independent middle
planes for redundancy.
Hot-swap power and cooling modules
By sharing fans, power supplies, cables and other components within a BladeCenter unit and
installing the redundant options for these, your organization can reduce the number of points
of potential failure, thus increasing system availability.
Blowers
Two hot-swap blowers are standard in the IBM Eserver BladeCenter; the blower speed
varies depending on the temperature. A failed blower needs to be replaced within 48 hours.
Note: An Acoustic Attenuation Module can be fitted to reduce noise.