CIO

How to Plan Your Cloud Computing Mix

Experimentation will be a vital phase in determining the right hybrid cloud path

Experienced CIOs have learned the hard way that achieving tangible benefits early in the technology lifecycle is no easy matter--whether its OO, CMMI, ITIL or SOA. Cloud computing shows promise and demands attention, but the related hoopla needs to be tempered with a good dose of business sense. The cloud, regardless of its variety, should never be considered an all-or-nothing proposition. Rather, it's a melting pot in which the mixture's properties will gradually change over the coming years.

Firms of all shapes and sizes are at least exploring the idea of sending infrastructure, applications, and information to the cloud -- and they would be taking a big risk if they weren't doing their due diligence.

The cloud trend should continue to gain traction as both technology and business leaders embrace some of the practical realities involved in introducing not just a new set of technologies, but accompanying processes and financial models.

At the same time, though, CIOs will face a significant hurdle in convincing business partners that there's more than meets the eye in terms of cost and effort. Remember a few years ago when people were asking, "Why aren't we using an ERP system?" Yogi Berra might say it's deja vu all over again.

Moving from a traditional hosting model to full-scale cloud computing model in one step is neither realistic nor prudent. So for the next several years, most enterprises will operate variations of a hybrid cloud model, one in which public and private infrastructure clouds and traditional data-center hosting coexist as part of the enterprise infrastructure ecosystem.

This mixed environment will give CIOs a lot of room to experiment with low-risk public cloud applications before seriously exploring vendor solutions.

The "Purpose-Built" Cloud

Experimentation will be a vital phase in determining the right hybrid cloud path for each organization.

Public clouds, owned and operated by third parties, offer greater scale and, as a result, greater cost savings. But they are immature, providing far less control over SLAs, increased security and compliance concerns, and complex back-end integration.

Private clouds are attractive because they offer similar benefits to those of a public cloud, but add increased control and advanced customization.

With many pros and cons to consider, the three operating models--traditional hosting, public clouds, and private clouds--will likely coexist for at least the next three to five years, especially for large and mature enterprises. In addition, each model will contribute infrastructure, middleware, and business applications to the enterprise technology portfolio.

I expect we will see companies travel in two directions. High-volume, high-performance firms that also must meet robust regulatory and security mandates (e.g., Wall Street firms and other large highly-regulated enterprises) are likely to shift from the current hosting model to a hybrid cloud model that combines hosting with elements of a private cloud.

These companies may selectively adopt public cloud services from vendors such as Salesforce.com, but for the most part, public clouds will not be able to meet their needs for performance, security, and compliance.

On the other hand, small to mid-size companies in less regulated industries, such as retail, may choose to augment their current hosting model with the public cloud--able to extract its benefits without the need to meet more stringent requirements.

Page Break

Firms of all shapes and sizes are at least exploring the idea of sending infrastructure, applications, and information to the cloud -- and they would be taking a big risk if they weren't doing their due diligence. As I recently wrote with two of my colleagues, our clients fall into one of six increasingly complex categories in their cloud exploration and adoption.

At one end, clients are surveying the market or might be using an off-the-shelf cloud application or service (the stock Salesforce.com or Google's Docs and Mail services come to mind). Clients at the other end of the spectrum are already using multiple cloud applications that are neatly integrated with corresponding in-house applications.

As a first step in sorting through the hype, an organization must gain a clear view of its systems portfolio in the context of business needs and pain points. Without it, a company will feel the weight of mismatched cloud services and applications--in terms of both cost and flexibility.

The results of a six-step readiness assessment will guide a company as it engages a few targeted cloud providers to discuss high-priority candidate systems for the cloud. In order to "clear the lens," an organization must:

  • Understand the business demands for new functionality.
  • Evaluate the current application and architecture portfolio, in terms of business gaps and technology health.
  • Determine candidate applications by understanding relevant cloud offerings.
  • Evaluate opportunities for infrastructure-only cloud services.
  • Assess management and operations readiness.
  • Identify early cloud adopters.

The pace and direction of each organization's cloud computing migration strategy will differ. But the need for an organization to understand its readiness for a new approach to delivering applications is a common thread.

With a comprehensive assessment in hand, a CIO can begin evaluating the company's systems architecture in earnest, and building a roadmap that demonstrates the business value in shifting pieces to the cloud, the company's readiness, and potential vendor fit.

In other words, you'll want to know the probable outcome before dropping elements into the melting pot.

Chris Curran is Diamond Management & Technology Consultants' chief technology officer and managing partner of the firm's technology practice. He writes the CIO Dashboard blog, and can be reached at Chris.Curran@diamondconsultants.com or @cbcurran on Twitter.