Home > Cloud Computing, Cloud Security, Virtualization > Internal v. External/Private v. Public/On-Premise v. Off- Premise: It’s all Cloud But How You Get There Is Important.

Internal v. External/Private v. Public/On-Premise v. Off- Premise: It’s all Cloud But How You Get There Is Important.

February 24th, 2009 Leave a comment Go to comments

I've written about the really confusing notional definitions that seem to be hung up on where the computing actually happens when you say "Cloud:" in your datacenter or someone else's.  It's frustrating to see how people mush together "public, private, internal, external, on-premise, off-premise" to all mean the same thing.

They don't, or at least they shouldn't, at least not within the true context of Cloud Computing.

In the long run, despite all the attempts to clarify what we mean by defining "Cloud Computing" more specifically as it relates to compute location, we're going to continue to call it "Cloud."  It's a sad admission I'm trying to come to grips with.  So I'll jump on this bandwagon and take another approach.

Cloud Computing will simply become ubiquitous in it's many forms and we are all going to end up with a hybrid model of Cloud adoption — a veritable mash-up of Cloud services spanning the entire gamut of offerings.  We already have today.

Here are a few, none-exhaustive examples of what a reasonably-sized enterprise can expect from the move to a hybrid Cloud environment:
  1. If you're using one or more SaaS vendors who own the entire stack, you'll be using their publicly-exposed Cloud offerings.  They manage the whole kit-and-kaboodle, information and all. 
  2. SaaS and PaaS vendors will provide ways of integrating their offerings (some do today) with your "private" enterprise data stores and directory services for better integration and business intelligence.
  3. We'll see the simple evolution of hosting/colocation providers add dynamic scalability and utility billing and really push the Cloud mantra.  
  4. IaaS vendors will provide (ala GoGrid) ways of consolidating and reducing infrastructure footprints in your enterprise datacenters by way of securely interconnecting your private enterprise infrastructure with managed infrastructure in their datacenters. This model simply calls for the offloading of the heavy tin. Management options abound: you manage it, they manage it, you both do…
  5. Other IaaS players will continue to offer a compelling suite of soup-to-nuts services (ala Amazon) that depending upon your needs and requirements, means you have very little (or no) infrastructure to speak of.  You may or may not be constrained by what you can or need to do as you trade of flexibility for conformity here.
  6. Virtualization platform providers will no longer make a distinction in terms of roadmap and product positioning between internal/external or public/private. What is enterprise virtualization today simply becomes "Cloud."  The same services, split along virtualization platform party lines, will become available regardless of location. 
  7. This means that vendors who today offer proprietary images and infrastructure will start to drive or be driven to integrate more open standards across their offerings in order to allow for portability, interoperability and inter-Cloud scalability…and to make sure you remain a customer.
  8. Even though the Cloud is supposed to abstract infrastructure from your concern as a customer, brand-associated moving parts will count; customers will look for pure-play vetted integration between the big players (networking, virtualization, storage) in order to fluidly move information and applications into and out of Cloud offerings seamlessly 
  9. The notion of storage is going to be turned on its head; the commodity of bit buckets isn't what storage means in the Cloud.  All the chewy goodness will start to bubble to the surface as value-adds come to light: DeDup, backup, metadata, search, convergence with networking, security…
  10. More client side computing will move to the cloud (remember, it doesn't matter whether it's internal or external) with thin client connectivity while powerful smaller-footprint mobile platforms (smartphones/netbooks) with native virtualization layers will also accelerate in uptake

Ultimately, what powers your Cloud providers WILL matter.  What companies adopt internally as their virtualization, networking, application delivery, security and storage platforms internally as they move to consolidate and then automate will be a likely choice when evaluating top-rung weighting when they identify what powers many of their Cloud providers' infrastructure.

If a customer can take all the technology expertise, the organizational and operational practices they have honed as they virtualize their internal infrastructure (virtualization platform, compute, storage, networking, security) and basically be able to seamlessly apply that as a next step as the move to the Cloud(s), it's a win.

The two biggest elements of a successful cloud: integration and management. Just like always.

I can't wait.


*Yes, we're concerned that if "stuff" is outside of our direct control, we'll not be able to "secure" it, but that isn't exactly a new concept, nor is it specific to Cloud — it's just the latest horse we're beating because we haven't made much gains in being able to secure the things that matter most in the ways most effective for doing that.
  1. No comments yet.
  1. No trackbacks yet.