NOTE: This post discusses just a fraction of the incredible content covered in my upcoming DCICT course for CBT Nuggets.
I can recall my shock in 2009 when Cisco Systems entered the server hardware market! I suppose it was similar to when I saw Amazon try (and sadly fail) at making smartphones. A Vice President at HP certainly was surprised and famously stated: “A year from now the difference will be (Cisco) UCS (Unified Compute System) is dead and we have had phenomenal market share growth in the networking space.“ Fortunately for Cisco Systems, he could not have been more wrong. In the 4th quarter of 2016 alone, Cisco did nearly 1 billion of server sales!
Cisco not only entered this market but has produced several variants already including rack mount servers, blade servers, and the hyper-converged Unified Computing System (UCS) of which there have already been three generations of technology.
In this post, let’s take an overview of the main products and technologies that make up the Cisco UCS umbrella.
Management Software
Cisco UCS Manager – this is the software for managing a single UCS domain. Don’t think this necessarily means a small environment, however, since this could mean up to 160 blade or rack mount servers in that management domain. And of course, each of those many servers could be running a huge number of virtual servers and/or containers. You have options when working with this software thanks to a GUI (Graphical User Interface), an XML Application Programming Interface (XML API), and a Command Language Interface.
The Cisco UCS Manager GUI
Cisco UCS Central Software – this software permits you to manage multiple domains located in the same campus, or even distributed worldwide. This provides the scalability required for very large enterprises.
Cisco UCS Director Software – since there are many different integrated systems included in the Cisco UCS world featuring equipment from the likes of EMC, Hitachi and more, the UCS Director Software helps you automate integrated infrastructure orchestration and management. Elements managed by this software include networking, hardware compute, operating systems, virtual machines, and storage.
Connection Technologies
Cisco SingleConnect Technology – connect your LAN, SAN, and management networks using one physical connection. Remember, this includes the connectivity for both your physical and virtual servers.
Cisco Direct Connect Technology – this advancement permits you to connect various servers in your overall system directly to the Fabric Interconnects. This allows you to manage these servers using a single cable for both management and data traffic. If you are not familiar with the UCS Fabric Interconnects, these are described below.
UCS Series Hardware
Cisco UCS Blade Server Chassis – these chassis can mount in industry-standard racks and use standard front to back cooling. They are so flexible in that they accommodate full-width blade servers or half-width blades. You can even mix and match these in the chassis. Cisco’s goal with these UCS chassis was to feature fewer physical components, eliminate the need for independent management of systems, and to increase energy efficiency.
The Cisco UCS 5108 Blade Server Chassis
Cisco UCS Fabric Extenders – the idea here is to scale the system without unnecessary complexity. Fabric Extenders bring the unified fabric into the blade server enclosure, providing multiple 10 Gigabit Ethernet connections between blade servers and the fabric interconnect, simplifying diagnostics, cabling, and management. As its name implies, this device extends the I/O fabric between the Fabric Interconnects (covered below) and the Cisco UCS Series Blade Server Chassis. This enables a lossless and deterministic Fibre Channel over Ethernet (FCoE) fabric to connect all blades and chassis together. Since the fabric extender is similar to a distributed line card, it does not perform any switching and is managed as an extension of the Fabric Interconnects. This approach removes switching from the chassis, reducing overall infrastructure complexity and enabling Cisco UCS to scale to many chassis without multiplying the number of switches needed, reducing TCO and allowing all chassis to be managed as a single, highly available management domain. The Cisco UCS Fabric Extenders also help manages the chassis environment (the power supply and fans as well as the blades) in conjunction with the Fabric Interconnect. Therefore, separate chassis management modules are not required. The Cisco UCS Fabric Extenders fit into the back of the Cisco UCS Blade Server Chassis. Each Cisco UCS chassis can support up to two fabric extenders, allowing increased capacity and redundancy.
The Rear of the Chassis with Fabric Extenders Installed
Cisco UCS Fabric Interconnects – these critical devices support a single point of connectivity and management for the overall UCS system. Because it is such a critical component of the system, it is often deployed in redundant pairs. As an example, consider the 6332 Fabric Interconnect which provides:
LAN and SAN connectivity for all servers within their domains
Bandwidth up to 2.56 Tbps
32 40-Gbps ports in one 1 rack unit (RU)
Support for 4×10-Gbps breakout cables
Ports capable of line-rate, low-latency, lossless 40 Gigabit Ethernet and Fibre Channel over Ethernet (FCoE)
Centralized unified management with Cisco UCS Manager
Efficient cooling and serviceability
The Cisco UCS 6332 Fabric Interconnect
Cisco Nexus Fabric Extenders – these optional components are third-generation devices that support LAN and SAN connectivity to the UCS system. They offer ultra-high and flexible bandwidth options. Thanks to the Nexus Fabric Extenders, you can take advantage of the latest data center technologies including:
Virtual Port Channels
Enhanced Virtual Port Channels
FabricPath
Unified Fabric
Application-Centric Infrastructure
Virtual Extensible VLAN-based topologies
Versatile TCAM
Cisco R-Series Racks – these racks are optimized for Cisco UCS featuring a custom-design for the Cisco infrastructure, including computing, network, and power
while they comply with EIA-310-D rack standards.
Cisco UCS B-Series Blade Servers – the approach here is a blade server for any purpose and any scale. Models are tailored for scale out, enterprise-class, or mission-critical deployments. As an example, the enterprise-class Cisco UCS B480 M5 Blade Server delivers support for the Intel Xeon Scalable processors; up to 6 terabytes (TB) of memory; four SAS, SATA, and NVMe drives; M.2 storage; up to four GPUs, and 160 Gigabit Ethernet connectivity for I/O throughput.
Cisco UCS C-Series Rack Servers – again, an approach for various workloads and scale. Consider the Cisco UCS® C480 M5 Rack Server that delivers:
A 4RU form-factor
The latest Intel Xeon Scalable processors with up to 28 cores per socket and support for two-or four-processor configurations
2666-MHz DDR4 memory and 48 DIMM slots for up to 6 TeraBytes (TB) of total memory
12 PCI Express (PCIe) 3.0 slots
Six x8 full-height, full-length slots
Six x16 full-height, full-length slots
Flexible storage options with support up to 32 Small-Form-Factor (SFF) 2.5-inch, SAS, SATA, and PCIe NVMe disk drives
Cisco 12-Gbps SAS Modular RAID Controller in a dedicated slot
Cisco UCS Virtual Interface Cards (VICs) – as described above, these interface cards permit simplified computing connectivity thanks to Cisco SingleConnect support. This unifies LAN, SAN, and systems management into one simplified link for rack servers, blade servers, and virtual machines. Second and third generation cards even feature lower latency thanks to usNIC technology. usNIC (user-space NIC) is Cisco’s low-latency computer networking product for Message Passing Interface (MPI) over 10 Gigabit Ethernet in high-performance computing. It operates at the OSI Model’s data link layer (Ethernet frames) or the network layer (UDP packets) to eliminate the overhead of TCP within a data center.
Cisco UCS Invicta Series – while officially End of Life from Cisco Systems, you still might find mention of these servers in certification and of course you might find them installed in the field. The idea behind these servers is ultra fast performance through the use of NAND flash memory for sustained high throughput, a high rate of I/O operations per second (IOPS), ultra-low latency, and fast write performance.
Cisco Integrated Infrastructure – through partnerships with other networking giants, Cisco has offered integrated systems including:
FlexPod – a pre-validated data center platform built on Cisco UCS, the Cisco Nexus family of switches, and NetApp data management solutions
Vxblock Systems – provide a wide range of solutions to meet requirements for size, performance, and scalability; built with compute and networking from Cisco, storage from Dell EMC, and virtualization technologies from VMware
