Is Cisco UCS “Stateless” or “State Run”?

Cisco design hierarchy places an over dependence on a proprietary network.

Cisco’s own words:

“The core concept of a stateless computing model is to separate the access to the application from the execution of the application.”

->    Separate intelligence from the application

“The compute node is just an execution engine with CPU, memory, disk, flash, or hard drive.”

->    A server is just a drone to execute orders

Source:  Cisco Cloud Computing –   Data Center Strategy, Architecture,  and Solutions
http://www.cisco.com/web/strategy/docs/gov/CiscoCloudComputing_WP.pdf

Cisco’s self-proclaimed “stateless” computing architecture is a hierarchical one based Cisco’s network-centric view of the world that drives their design philosophy. Cisco has a lot of expertise in designing networking equipment, but does this expertise necessarily translate to servers? If you look at a typical datacenter design of Cisco equipment, you will see a hierarchical design strategy where all the traffic and management is controlled by the switch or set of switches. Now Cisco has embedded UCS Manager in a top-of-rack switch (based on Nexus design) called a Fabric Interconnect. Cisco’s embeds service profiles in this top of rack switch to abstract the server hardware. Cisco states that the core concept of a stateless computing model is to separate the access to the application from the execution of the application and that servers are just “execution engines”. Cisco servers are relatively “dumb” and tied to a proprietary hub. The design hierarchy separates intelligence from the application.

In my opinion there are several problems with Cisco’s design: Data and control are on the same path making this a point of failure as well as a bottleneck. Additionally with Cisco the intelligence is separated from the application where it is needed most.

HP believes servers need to be more intelligent. Now with HP ProLiant servers Intelligence is now even closer to the application making them more autonomous, easier to manage and better optimized.

This is the required architecture to scale servers to cloud scale and makes HP servers a better choice than Cisco for the next generation datacenter.

 

About Prasenjit Sarkar

Prasenjit Sarkar is a Product Manager at Oracle for their Public Cloud with primary focus on Cloud Strategy, Oracle Openstack, PaaS, Cloud Native Applications and API Platform. His primary focus is driving Oracle’s Cloud Computing business with commercial and public sector customers; helping to shape and deliver on a strategy to build broad use of Oracle’s Infrastructure as a Service (IaaS) and Platform as a Service (PaaS) offerings such as Compute, Storage, Java as a Service, and Database as a Service. He is also responsible for developing public/private cloud integration strategies, customer’s Cloud Computing architecture vision, future state architectures, and implementable architecture roadmaps in the context of the public, private, and hybrid cloud computing solutions Oracle can offer.