1.1 What is Cisco UCS?
To have a global idea, let’s have a look to a definition found in a UCS Tech Sheet provided by Cisco, which states the following:
The Cisco Unified Computing System™ is a next-generation data center platform that unites compute, network, storage access, and virtualization into a cohesive system designed to reduce total cost of ownership (TCO) and increase business agility.
Here is the key word “data center platform“. Actually, UCS is a platform composed by both hardware and software. Over the Cisco UCS hardware product portfolio has been evolving and growing, and as a result, the Cisco UCS hardware product portfolio has grown. The hardware part consists of rack and blades servers, and the associated components to house the blade servers (Servers Chassis) and their connectivity (FEX extenders and Fabric-Interconnect).
To have a simpler view, we have some rack servers, some blade servers hosted in chassis, all of them are racked in a cabinet. The chassis has got two redundant IO Modules, aka FEX (Fabric Extenders) in UCS Vocabulary, which connects it to a specific switch called the Faric Interconnect. Actually, there are two Fabric-Interconnects (FI), not only one. Bascially, you connect each FEX to each FI. Here is a global overview of how the cabling should be done, but we’ll go deeper in the next post.
As mentioned earlier, the IOM modules in Cisco UCS are not called IOM, but have the fancy name of FEX. It is worth to mention and insist that each FEX need to be connected to one FI, so FEX A will go to FI-A, and FEX-B will go to FEX-B.
Now that we have a brief description in mind of what is UCS, let us describe in details its components and the different terms used in a UCS environment
1.2 UCS Components and Terminolgy
UCS Chassis: This is simply the chassis that hold the UCS blade physical servers. We may have have up to 8 half-size servers pushed horizontally into the chassis.
Server: The actual physical piece of hardware that contains all necessary hardware components (RAM, NIC, HDD, etc…). It is worth noting that a server is a useless bunch of hardware, if not assigned a Service Profile. The server can not even be powered on without a Service Profile.
Fabric Extender (FEX): It is the IOM module that multiplexes and forwards all traffic using a cut-through architecture over one to four 10-Gbps unified fabric connections to the fabric interconnect. In a correct setup a fabric extender is connected to only one fabric interconnect. The other fabric extender in the blade chassis should be connected to the other fabric interconnect.
Fabric Interconnect: This is the core of the UCS architeture, as all management is done on it, through the embedded UCS manager which is used to manage the entire UCS system.The FIs are deployed in cluster in an active-active pair that supports failover.
Here is a global picture showing all UCS system in one shot
That’s all it for the UCS physical components. In the next post, the initial deployment and cluster creation will covered.
In this post we had a review of what is Cisco Unified Computing System (UCS) and what are the physical components that come with it. In the next post, we will see how to initialize and create our UCS cluster, and how to proceed with the postpone configuration.