Cisco Domain Ten: Domain 2: Virtualization and Abstraction
Last week I introduced our new Cisco Services framework to help guide your data center and cloud transformation – Cisco Domain Ten (SM). I also described the types of challenges you should be thinking about in the Facilities and Infrastructure layer, Domain 1. Now let’s discuss the type of challenges that Domain 2, Virtualization and Abstraction, could present to you. While Cisco Domain Ten can be applied to help you in any data center transformation, I’ll keep focused on showing you how Cisco Domain Ten helps illuminate your path to cloud transformation.
There is a school of though that assumes that if you have a virtualized environment, you have a cloud. Or indeed if you have a hypervisor, you have a cloud. Not true. Well, more precisely, virtualization is a necessary but not sufficient condition to say you have a cloud. I”m not trying to underplay the importance of virtualization here. Not at all. Rather, I’m drawing attention to the Cisco Domain Ten model and saying: there is a lot more to cloud – 10 domains in our framework – and it’s not all about virtualization and hypervisors.
Let’s then focus on virtualization and abstraction. And that second word – abstraction – let’s not forget about that. Virtualization enables a logical abstraction to allow the physical IT resources – network, compute and storage – to be shared among a number of applications. It’s that abstraction that is a major factor in ensuring your applications can seamlessly exploit the elastic nature of cloud, enabled by virtualization and the hypervisor.
“OK” you may say, “Just put in a hypervisor and that’s you on the path to virtualization. Easy. Done. Tell us about Domain 3”. Hold on, hold on, hold on 🙂 Let’s talk about some of the challenges involved in virtualization. And these challenges – at least from a Cisco perspective, others may take a more limited view – take you outside the software layer in the compute platform. Challenges you will face, and/or decisions that you will need to make, include (for example) ….
- Architecting virtualization of not only compute resources, but also network and storage resources.
- Deciding upon the correct percentage of your IT estate that could be – and should be – virtualized. Too little and you’re wasting IT investment, too much and your end users and/or end customers will suffer.
- And with some server platforms designed for virtualization, such as Cisco UCS, how do you exploit your hardware platform to the max.
- Determining which third party software tools can you employ to help you in the challenges of virtualization.
- Designing in support for multiple hypervisors (not uncommon – although I can see how this increases complexity and reduces standardization, which if you remember was a theme in my previous blog).
- Ensuring designed-in holistic virtual machine mobility – movement which will include movement of security policies along with the VM itself. This is where the Cisco Nexus 1000V will have a key role to play in your virtualization architecture.
- Ensuring no single point of failure in your virtualized architecture.
- (You’ll notice I don’t yet discuss application virtualization? – I will come back to this in Domain 8, “Applications”)
Each of these is a not unsubstantial design challenge. We in Cisco Data Center Services can help you in in these areas (for example – remember the above is a limited subset of where we can help you in Domain 2) ….. so get in touch when you need additional expertise to complement your team.
My next blog will touch on Automation and Orchestration, an area I have been heavily involved in this year with Cisco Intelligent Automation for Cloud Deployment Services. Look out for my Tweets with the next update (via @StephenSatCisco)