Cisco UCS Blade Chassis
The Cisco UCS 5108 chassis incorporates unified fabric; integrated, embedded management; and Fabric Extender technology. It uses fewer physical components, has no need for independent management, and enables greater energy efficiency than traditional blade server chassis. This simplicity eliminates the need for dedicated chassis management and blade switches, reduces cabling, and enables Cisco UCS to scale to 20 chassis without adding complexity.
The Cisco UCS 5108 blade server chassis is the first blade server chassis offering by Cisco and is 6-RU. The chassis can mount in an industry-standard 19-inch (48-cm) rack and uses standard front-to-back cooling.
At the front of the chassis, shown in Figure 16-7, are the server slots. It can accommodate up to eight half-width or four full-width Cisco UCS B-Series blade server form factors within the same chassis.
Figure 16-7 Cisco UCS 5108 Blade Chassis (Front)
Also, at the front are installed the power supplies. The Cisco UCS 5108 blade server chassis supports up to four fully modular PSUs that are hot-swappable under certain power redundancy configurations.
The power management and redundancy modes are as follows:
- Non-redundant: Cisco UCS Manager turns on the minimum number of power supplies needed and balances the load between them. If any additional PSUs are installed, Cisco UCS Manager sets them to a “turned-off” state. If the power to any PSU is disrupted, the system may experience an interruption in service until Cisco UCS Manager can activate a new PSU and rebalance the load.
- N+1: The total number of PSUs to satisfy non-redundancy, plus one additional PSU for redundancy, are turned on and equally share the power load for the chassis. If any additional PSUs are installed, Cisco UCS Manager sets them to a “turned-off” state. If the power to any PSU is disrupted, Cisco UCS Manager can recover without an interruption in service.
- Grid: Two power sources are turned on, or the chassis requires greater than N+1 redundancy. If one source fails (which causes a loss of power to one or two PSUs), the surviving PSUs on the other power circuit continue to provide power to the chassis.
The power policy is a global policy that specifies the redundancy for power supplies in all chassis in the Cisco UCS domain. This policy is also known as the PSU policy.
At the back, as shown in Figure 16-8, are installed the FEXs for the uplink connectivity to the Fabric Interconnects and the fan modules.
Figure 16-8 Cisco UCS 5108 Blade Chassis (Back)
As the blade servers use passive colling, which means that they do not have their own cooling, the fan modules of the chassis take care of the cooling of the chassis, the servers, the power supplies, and the FEX modules. The chassis has a total of eight slots for fan modules to provide the needed redundant cooling capacity.
In the chassis are also installed the FEX modules. They are installed at the back of the chassis because there are two slots for the two FEX modules. To build a Cisco UCS cluster, you need to have two Fabric Interconnects. The cluster link between them does not carry any data communication, because each of the Fabric Interconnects is a separate data path. This is very important because, upstream, the Cisco UCS connects to two different infrastructures—the LAN and the SAN. Each of them has different design requirements for supporting redundancy and increased reliability. A very important principle is always to have two physically separate paths between the initiator and the target. In the LAN infrastructure the approach is almost the opposite—the goal is to interconnect as much as possible and then leave it to the switching and routing protocols as well as to technologies such as port channels and virtual port channels (VPCs) to take care of the redundancy. However, inside the Cisco UCS, it is one and the same physical connectivity, over which communication to both the LAN and the SAN occurs. And this also means that there must be a way for the internal Cisco UCS infrastructure to support the redundancy design for both the LAN and SAN. As the LAN is more flexible, the challenge is to be compliant with the more restrictive design of the SAN infrastructure. That’s why, inside the Cisco UCS, it is important to preserve two physically separate paths of communication. Because of that, there are two Cisco UCS FEXs per chassis: one for the communication path through each of the Fabric Interconnects. Put in a different way, each UCS FEX connects to only one UCS Fabric Interconnect.