Private Cloud: From "What" To "Why" ... To "How"

I've been living with the private cloud journey since well before my first public post at the beginning of 2009.

The appeal of the private cloud model has steadily grown in the last 18 months, and shows all signs of becoming the primary cloud adoption model for most typical enterprises.

And, over that time, I've seen my dialogue with customers and partners evolve from...

"what is a private cloud and how is it different from other proposed cloud models"

to ...

"why you should consider a private cloud model for your enterprise"

to the more recent

"... and here is how you start planning your journey to the private cloud".

It's a natural evolution of the discussion, if you think about it.  I thought it'd be a good opportunity to bring everybody up to date as to where the conversation appears to be at present, and where it's likely to head.

A Cloud Is A Cloud Is A Cloud?

So much energy gets expended on arguing over fine shadings of cloud definitions  -- as if they really matter in the big scheme of things.  As I mentioned in a previous post, these debates tend to get settled in the marketplace when people vote with their IT budgets.

Not to bore you, but my working definition of "cloud" goes something like this:

-- built differently, using dynamic pools of virtualized resources
-- operated differently, with a focus on service delivery rather than underlying components
-- consumed differently, in a way that's convenient for the consumer vs. the provider

I tend to call a cloud "private" if it is mostly under the control of an enterprise IT organization, much in the same way that traditional IT is under control.  I tend to call a cloud "public" if the service provider is largely in control vs. the tenants.

The concept speaks more to IT strategy, process and supporting architecture vs. debates about where it's located (internal, external), how it's paid for, etc..  The reason is simple: other forms of familiar economic infrastructure (power, transportation, financial systems, etc.) already follow this model; it's safe to say IT will be very similar over time.

If that's the "what" of private clouds, let's turn to the "why".

The Rationale For Private Cloud

If you've ever taken a look at a high-level enterprise IT, you'll notice that the vast majority of money is spent making the infrastructure do pretty much what it did last year.  Too much of every IT dollar is spent keeping the lights on; far too less is spent creating new business value.

So, argument #1 for a private cloud model is "a better investment strategy for IT infrastructure". That means either spending less on IT infrastructure delivery, or -- more likely -- the opportunity to invest IT dollars in things that really create value.

Many organizations are focused on growth and innovation.  Very frequently, IT capabilities are needed for these new initiatives, and -- simply put -- speed matters.  A successful private cloud model will remove most of the typical IT infrastructure friction associated with these newer growth and innovation efforts.

So, argument #2 for a private cloud model is "deliver speed, flexibility and agility for new business initiatives".  Some people phrase this as repositioning IT from being seen as a blocker to being an seen as an enabler for the business.

Finally, there's a ton of cruft that's been built up in the legacy physical world: applications designed for another era, security and compliance systems that aren't keeping up, skill sets and roles that aren't aligned, inability to leverage external services, and so on.  Many customers see a private cloud model as an opportunity to create a legacy-free environment based on modern technologies and operational principles.

That makes argument #3 for a private cloud model "an opportunity to re-envision how IT delivers its services back to the organization".  In many ways, it's the proverbial clean sheet of paper.  And that's powerful stuff indeed.

Generally speaking, when speaking with enterprises, I can usually make the case that the private cloud model is the right cloud model for them -- that part isn't too hard.

And I usually can make a decent case around the three arguments for doing so -- a better IT investment strategy, deliver speed and agility back to the business, and re-create IT as a service delivery organization.

Which leaves us with the "how do I get there?" question.

The Journey To The Private Cloud

No coincidence, this is the theme that EMC (and our ecosystem) is focusing on currently.  We've got a good definition of the end state, we have powerful reasons for getting there, the technology is available -- now, how do we get our customers there in a logical and progressive way that recognizes their pragmatic realities?

Hint: at a macro level, it really isn't about technology any more. 

You're free to debate me on this, but here's my pitch: it's here.  It works.  It does what it's supposed to do.  Yes, it always could be faster, better, cheaper, more complete, more integrated, etc. -- but there are more than enough examples of organizations moving ahead with what's on the truck today in the marketplace.

No, the discussion is rapidly moving squarely to people and process -- generally speaking, it's now an IT leadership challenge more than anything.  How will you move your organization forward to this model? 

Now, if you're looking for reasons not to move forward, I'm sure you can find plenty of them -- insert long and familiar list here -- but there are enough diverse organizations moving aggressively forward (large and small, multiple industries, public and private sectors, multiple geographies, etc.) that it's started to hit a critical mass.

The Importance Of Process and Organizational Change

If you follow this blog, you'll probably remember the 3-phase model we're using to conceptually describe how an organization might approach this journey.

Phase 1 is IT Production: the apps that IT owns on their own: file servers, test and dev, maybe email.  IT doesn't have to go ask anyone to go virtualize these environments.

In some sense, it's almost too easy here: promote the widespread use of virtualization and a few related technologies, do relatively little in the way of signficant process change, and big paybacks result.

Phase 2 is Business Production: these are the named apps that business people depend to run important parts of their business.  Stakes are much, much higher here.

This is where process change is required to be successful.  It's not only new tools for things like management and backup and security; it's entirely new processes to complement them.

Phase 3 is IT-As-A-Service: various offers for infrastructure-as-a-service (IaaS), one or more platforms-as-a-service (PaaS), perhaps even software-as-a-service (SaaS) using virtualized desktop experiences as an example.  And, once abstracted as a service, the option emerges whether to provide that service internally, or use a compatible service provider to do so.

Not only are we talking substantial process change, we're often talking significant organizational changes as well: new roles, new skills, new responsibilities.  The job titles change -- the relationships change as well.  Even in a moderate-sized IT organization, that's heavy stuff indeed.

It's one thing to evaluate technology and vendors, select your favorite, and deploy it.  It's another thing entirely to re-cast an entire organization around a new model.

Perhaps a better way to represent the journey, then, would be to characterize each step in terms of required process and organizational changes to be successful.

Phase 1 would be "minimal" -- IT is still IT as we've always known it, they're just using virtualized technology vs. the physical stuff for a portion of the environment.

Phase 2 would be "significant" -- multiple IT disciplines have to change the way they get their job done, and learn to work together in new and potentially uncomfortable ways.

Phase 3 would be "transformational" -- entirely new IT disciplines are established (usually prefaced by the word "cloud") -- cloud architect, cloud capacity planner, cloud business analyst, cloud GRC, etc. with far less reliance on the individual traditional disciplines.  New metrics and KPIs are put in place; compensation systems are aligned with new outcomes.

And, as this new realization takes hold, many IT leaders are realizing that they might need some outside help to get where they're going sooner than later.

And, yes, we've recently seen a strong demand for consulting services along these lines.

The Value Of A Good Consultant

I know, there can often be a inherent distaste for bringing in outside consultants.  We at EMC tend to make the same sort of unflattering comments I'm sure that many of you make when an outside consultancy is brought in to assist.

That being said, if it's a really big / important / complex topic -- that's where a good consulting organization can really add value.

Part of the problem is that every organization is different.  No cookie-cutter recipe will usually do.  Sure, there are some common themes and repeatable patterns, but no two IT organizations -- or their situations -- are alike.

Transformational skills are often in short supply in many organizations.  Usually, there's plenty of operational and program management talent around, but the required higher-order skills can be hard to come by:

• making the case to the business -- and the IT organization

• finding a place to demonstrate the success of newer approaches, accelerating change

• starting the organizational change management efforts required to move the rest of the org forward

• creating new governance models that lessen resistance and further accelerate change

• baselining and documenting improvements to KPIs on behalf of the IT organization.

This can be very daunting stuff to many.  An engagement with a consulting team that's done it before -- and brings that "outside credibility" that might be lacking -- can be invaluable in the transformation.

Yes, EMC Does Consulting

Many folks aren't aware of this, but EMC has been quietly assembling a 1000+ army of consultants.  To be clear, these aren't the folks who do the traditional infrastructure-oriented design/implement projects, but are more at the business level.

A decent portion of this consulting team is now involved in "JttPC" (journey to the private cloud)

transformational engagements.  One of their more public customers is -- wait for it -- EMC's IT organization, whose journey can be followed here.

Compared to demand, that's just a very small drop in a very large bucket, so we're also working with a growing array of technology industry consultants to meet substantial customer demand: folks like Accenture, CSC, Cap Gemini, Fujitsu and many others.

Not all of this consulting is being done with traditional enterprise IT: a growing portion is now with an entirely new class of service providers that have sprung up in the marketplace. 

They too need to get from where they are today to where they'd like to be in the near future, something I'll be talking about more on my other blog.

Every time I meet with the folks from EMC Consulting, I hear more wonderful stories about customers and partners we're working with to help them along in their journey.  It's cool stuff.

What makes this so interesting is that everyone is largely heading to the same place; they're just taking slightly different paths to get there.

Exciting times indeed ...