In Part 1 of this blog series, I introduced Gartner's definition of Fabric based infrastructure (FBI).  I explained the benefit of data center fabrics with the promise of continuing in a follow on blog.  Some readers rightly pointed out to me that I had not yet touched on the automation benefits. This video featuring Gartner analyst, Donna Scott and Cisco CTO Paul Perez provides a great overview of the benefits namely:

• Capital cost savings with infrastructure consolidation
• Lower operating costs with automation
• Speed of implementation and infrastructure deployment
• Better SLAs with faster recovery or migrations

Let's dig a bit deeper and start by looking at the difference between a FBI server and a run of the mill server.  FBI essentially lets us define the profile of a server in software. The profile here refers to as many as 120 attributes of a physical server stored as meta data in a profile. These attributes include BIOS version, LAN connection parameters, SAN connection parameters, UUID, MAC Address etc.

In the case of run of the mill servers some of these attributes remain the same throughout the life of the physical server. You may be able to alter other attributes with manual operations through proprietary user interfaces.  As shown in the figure above, the server identity (service profile) of a FBI server is abstracted from the physical server.

The benefit  (shown above) is that modifying the identity and purpose of a server is now a breeze. It is as easy as dropping a software profile on to a physical server.  You can now pool resources for multiple workloads and adopt just in time provisioning and de-provisioning.  Unlike other solutions, which continue to build layers of management software to stitch together legacy infrastructure, FBI enables on-demand provisioning from shared pools of infrastructure resources. With the introduction of the Cisco VIC 1225 in the second half of last year you can even make Cisco UCS rack optimized C-Series servers part of the FBI and manage them with service profiles.

Bottom line: Cisco UCS is to physical servers as a hypervisor is to virtual servers.

The Cisco UCS Manager running in the Cisco UCS Fabric Interconnect manages, stores and restores service profiles.  It goes further to expose an API, lends itself to modifying updating and moving service profiles, programmatically.   As server life cycles (http://blogs.cisco.com/datacenter/server-life-cycle/) become more dynamic with cloud deployments these capabilities are indispensible.  The power of APIs simply cannot be underestimated these days.

The APIs are the underpinnings of automation and orchestration of server infrastructure and essential for management of data centers with error free operation and high availability. With API's and service profiles, data center managers can enforce policies and standardization which in turn reduce errors, tedious trouble shooting and in the end save time. From a maintenance stand point server failures can now be fixed much faster. Since there is no manual configuration you can significantly lower mean time to recovery (MTTR).  An older blog lists customers who have quantified these benefits. The Cisco UCS API is open and you can even dabble with it using the UCS Platform Emulator even before investing in the physical infrastructure.

In the third and final installment of this blog series, I will discuss fabric extension into virtual machines. I will also discuss the benefits it brings to application workloads running on Cisco UCS servers.