ProLiant BL p-Class Interconnect Overview - White Paper

3
Each ProLiant NC series NIC has been designed, developed, and manufactured to meet the needs of
ProLiant customers. HP's testing, certification, and validation of NICs within ProLiant servers further
enhance this value. Additionally, the NC series NICs provide the following features:
• Preboot eXecution Environment (PXE) and Wake on LAN (WOL) capability
• High-availability teaming including Network Fault Tolerance (NFT), Transmit Load Balancing (TLB),
and Switch-assisted Load Balancing (SLB)
• Teaming Configuration GUI for Microsoft
®
operating systems
• Auto-negotiation for speed and duplex
• Support for management agents and sophisticated management tools such as Insight Manager 7
• TCP Checksum and segmentation offload to reduce the load on the CPU for overall improved
system response (NC7780 and NC7781)
ProLiant BL p-Class server blade enclosure
The ProLiant BL p-Class server blade enclosure is a 6U (10.5-inch) chassis with two outside
interconnect bays and eight interior server bays that support various combinations of ProLiant BL p-
Class server blades (Figure 2). The bays are designed so that the server blades and interconnect
blades slide in and blind mate to the server blade enclosure backplane for power and data
connections. The signal backplane routes both Ethernet and Fibre Channel signals from the server
bays to the interconnect bays while completely isolating these signals from each other.
Figure 2. ProLiant BL p-Class server blade enclosure
Blade enclosure network signal routing
Each server bay supports up to four Ethernet signals. This is sufficient for the BL20p and BL20p G2
server blades with four NICs each, and it is ample for the BL40p server blade, which has six NICs
spread over four server bays.
Figure 3 illustrates the network signal routing from a single server blade bay to the interconnect bays.
Individual category 5e specified signal traces on the passive backplane route the network signals as
Ethernet from the server blade NICs to the interconnect bays. To provide redundant network
connections and maximize network availability, half of the network signals from each server blade
bay go to each interconnect bay, regardless of the slot width of the server blade or the number of
NICs it supports.