Cloud computing: A new era of IT opportunity and challenges

It doesn't take long to get a good feel for the potential of cloud computing and how it can offer ready access to entirely new business capabilities, less expensive IT resources, and unrivaled flexibility for businesses of every size. Since becoming a hot topic early last year as major vendors, including top firms such as Amazon, Google, and Microsoft, jumped on the bandwagon with a wide-range of offerings, cloud computing has consistently stayed on the industry's radar. One of the bigger challenges IT departments will face this year is whether they can take the plunge with cloud computing quickly enough to benefit their organizations as a whole.

With leading companies still joining the movement -- including IBM, HP, and Salesforce -- cloud computing has moved from a cottage industry to one of the bigger growth areas in the computing business, just as the industry as a whole begins to take serious lumps from the recession.

The onus is now on businesses to take advantage of cloud computing to cut costs and become more agile. In the process, they will have some hard choices to make -- and some intriguing ones as well -- if they want to access the many advantages that cloud computing platforms can provide.

There are also some non-trivial challenges involved in adopting cloud computing that must be watched closely as well. These includes a long list of issues such as the security and privacy of business data in remote 3rd party data centers, the dreaded concerns about platform lock-in, worries about reliability/performance, and even fears about making the wrong decision before the industry begins to mature.

However, in a business environment where change is almost mandatory in order to survive, cloud computing appears to offer significant economic benefits if the risks can be offset. Hence, one of the bigger challenges IT departments will face this year is whether they can take the plunge with cloud computing quickly enough to benefit their organizations as a whole.

ZDNet's own Phil Wainewright has covered some of the more interesting issues swirling around cloud computing of late including the default lock-out that occurs in the event of the demise of a cloud computing provider as well as the brewing SLA battles between the major providers. This underscores how the cloud computing space is where the new platform wars are forming and it's sizing up to be as big or bigger than earlier ones. The good news for now: In a wide-open new industry, there is no clear leader today and choice prevails.

This brings up the side discussion of what actually constitutes cloud computing, since everyone seems to be applying the label to anything that runs on the network. Is it Web hosting of your application code? Is it a software platform as an on-demand service? Do SaaS applications count as cloud computing? The answers to all these questions are a qualified yes; the answer hovers roughly around the outsourcing of computing of any kind (CPU, storage, apps, etc.) using a shared cost, commodity utility model. In general, you know if you're involved with cloud computing of some kind if you're receiving a bill for computing services being done for you somewhere else but which you can access directly.

I used the term commodity utility model since cloud computing providers aren't monopolies today (unlike a lot of the other utility services we use in business) and currently compete actively with each other on features and pricing. This means that CPU cycles, bandwidth, and application logins in the cloud will be extremely cheap and extensively commoditized for the foreseeable future. This constant competition creates continuous pressure to drive down the costs and increase the capabilities of cloud computing platforms in a way that just doesn't happen naturally within IT organizations today. In other words, just like most businesses don't generate their own power or create their own financial institutions to keep their money in, increasingly they won't keep their computing in largely parochial, private capabilities that can't leverage the economies of scale, innovation, and efficiencies of dedicated providers.

Enterprise cloud computing: Some assembly required

As a new industry, there is a lot of choice in cloud computing today and will continue to be more until the inevitable shake-out occurs and the winners begin to emerge. That also means there is no dominant model for how cloud computing should be delivered and this is resulting in some interesting fragmentation in the market already. Some cloud computing offerings are so generic (Amazon's Elastic Compute Cloud, aka EC2) that they are nothing more than sophisticated on-demand hosting services while others offer nearly everything you need to create software, as long as you use their programming model, frameworks, tools, and management systems (Google App Engine). Which model a business chooses will have deep ramifications for how they can take advantage of cloud computing and because of this, most organizations will likely have multiple providers.

Is it time to declare the death of the enterprise data center? No, not quite yet, but it's coming.

Let's also make no mistake, most IT executives currently think very few cloud computing solutions are properly enterprise ready today. Though a number of them have come a good bit of the way towards addressing enterprise needs, it's currently up to IT departments to close the remaining gap in terms of combining everything needed to "complete" their cloud computing effort, from management tools to data synchronization strategies and security audits. Consequently, the smart bet at the moment is to be building initial competency with cloud computing and becoming a sophisticated consumer of 3rd party cloud services while the industry evolves and begins to settle.

For now, here are the competencies IT departments need to be developing to both embrace the opportunities and deal with the challenges of cloud computing today:

Enterprise cloud computing competency areas

• Ability to standardize and compare economic models for IT/cloud services. Businesses should be able to compare their traditional IT cost structures to the CPU/hour, gigabit/month utility pricing model common in the cloud computing business. Hidden costs such as management, governance, and transition costs including staff training should be factored into the equation. The raw pricing of cloud computing services is extremely compelling but will be dampened by the fully loaded costs and organizations should be using the most accurate numbers when determining how long it will take to recoup the investment in a move to cloud computing.
• Effective prototyping and vendor selection processes. Making a long-term bet with a given cloud computing platform isn't required to start taking advantage of what they have to offer. Smart organizations will temper their adoption with a wait-and-see attitude combined with proactive prototyping and a selection process that will help them select the best set of cloud computing services and tools. Early trials with non-mission critical applications will be a useful way to help everyone in the organization get up to speed on cloud computing technologies and how they integrate with existing IT processes. An extra bonus if the trial actually helps validate the economic model and saves budget dollars this year in the process. Also critical will be finding the dark side of cloud issues including performance implications, latency, and other issues involved in moving services farther away from where they are being consumed, as will getting a clear sense of the strengths and weaknesses of the cloud computing platforms used in the prototypes/pilots.
• Technical ability to adopt and execute with cloud computing services. Once it makes economic sense, the next hurdle will be achieving wider technical fluency in the selected cloud computing platforms, which tend to emphasize open source technologies or newer Web-style programming languages and application models that are often less well known in the enterprise world. Architects, developers, testers, operations, enterprise security, and networking staff will all need to be capable of working effectively with the cloud computing services. They must be given the time and resources to acquire these skills without distracting from their core activities.
• Incorporation of cloud computing into strategic IT planning. Important questions arise when cloud computing is brought into the picture: How will it affect service-oriented architecture (SOA) strategy? How are disaster recovery plans impacted? What about backups and legally mandated data archiving policies? What is the risk profile for using cloud computing services and what is the mitigation strategy? What is the potential for platform lock-in and how can it be avoided? All these questions and more must be asked and addressed, particularly as cloud computing is embraced for more important business applications and IT functions.

Already the cloud computing landscape is becoming fairly complex with the emergence of topics such as public vs. private clouds, hosting vs. platform-as-a-service, cloud interoperability, non-relational cloud databases, and cloud SOA. I'll explore these subjects soon including a round-up of the latest enterprise cloud computing products and services in an upcoming post. For now, cloud computing is one of the few bright spots this year where IT departments can deliver valuable bottom-line results to the business with a fairly highly degree of confidence.

Are you finding cloud computing a compelling option to run your business this year? Why or why not?