Cloud Computing is the latest technology rage. Is this hype or here to stay? Is Cloud Computing much cheaper than a corporate data center? What are security concerns? Is McKinsey's report "Clearing The Air On Cloud Computing" bringing much needed reality check to all the cloud computing hype? What are the white papers and reports that are must-reads for Cloud Computing?
We have invited Cloud Computing Expert Bernard Golden to shed light on these topics. Bernard Golden has over twenty years experience in the technology field, having worked in global consultancies, enterprise software companies, and large IT organizations. Today he is CEO of HyperStratus, Silicon Valley-based consulting firm that helps its clients define and implement their cloud strategy and systems. Bernard is the author of Virtualization for Dummies, the most popular book on the subject ever published. He also serves as the Virtualization and Cloud Computing Advisor for CIO Magazine. Bernard is a popular speaker, appearing at many conferences like CloudWorld, OSCON, and EDUCAUSE.
Ambal Balakrishnan: The term "cloud computing" means a variety of things. How would you define Cloud Computing?
Bernard Golden: Cloud computing has many specific definitions offered by many people. Common to all of them are three core characteristics:
1. Huge resources: a cloud provider offers enormous scalability by having large amounts of computing resources available, making it seem that there are no practical limits for system scalability;
2. No commitment: resources are available without any need for an up-front commitment by the user; getting going requires nothing more than an easy signup and perhaps a credit card; and
3. Pay by-the-drink: Resources are easily obtained and released and payment is based on actual use rather than a fixed access fee
Ambal Balakrishnan: What does your company HyperStratus do in the Cloud Computing space?
Bernard Golden: HyperStratus does cloud computing consulting, helping organizations in these areas:
- Developing their cloud computing strategies
- Planning their cloud computing initiatives
- Reviewing their application portfolio for cloud opportunities
- Selecting appropriate cloud providers
- Designing their cloud-based application architectures
Ambal Balakrishnan: What in your opinion is the difference between Cloud Computing & Grid Computing?
Bernard Golden: Many people feel they are similar or even the same. In my view, grid-based applications spread a single application across a number of machines, impose a particular application design, and execute parallel processing. While certain cloud providers provide an application framework and thereby an overall application design approach, parallel processing is not required, nor is it a requirement that applications run across a pool of machines.
Ambal Balakrishnan: What are the top three challenges & considerations for organizations looking to adopt cloud computing model?
Bernard Golden: People often raise a number of concerns about cloud computing: the role of SLAs; the true TCO of cloud apps, and even the ease of application migration. In my view, there are three primary challenges for organizations to truly leverage cloud computing; two are present today, and one will be a longer-term challenge. Those challenges are:
1. (Present today) The skinny straw of network bandwidth: bandwidth between corporate premises and cloud providers is much lower than intra-data center bandwidth. This makes it difficult to move large amounts of data to and from cloud environments.
2. (Present today) Privacy requirements are unclear: The laws and regulations that affect company data stored in the cloud are not fully understood and make it difficult to know what applications may be moved into cloud environments. This raises risk concerns and restricts cloud adoption.
3. (Longer-term) Software licensing: As companies move applications into the cloud, they will confront software licensing conditions designed for perpetual installation on a fixed machine base. In a cloud environment where the number of machines dedicated to an application can scale up and down rapidly, those software licensing assumptions no longer hold. There will be a significant time during which the conditions of software use will conflict with typical licenses.
Ambal Balakrishnan: Is there a growing market acceptance of applications served from the cloud? If so, why?
Bernard Golden: There is a growing acceptance of SaaS applications. As companies have experienced success with one or more SaaS offerings like Salesforce or Google Docs, they are more willing to consider other SaaS applications. And, of course, for companies with thin IT resources, SaaS products offer a way to gain the benefits of software without requiring IT headcount.
.jpg)
Ambal Balakrishnan: What is the potential market for Cloud Computing?
Bernard Golden: I don't think we've even begun to recognize the potential market. Something to keep in mind is that every platform shift in computing (e.g., mainframe to minicomputer, minicomputer to client/server, etc.) has vastly reduced the cost of computing and exploded the uses computing is put to. Cloud computing, with its centralized, inexpensive data centers, as well as ultra-efficient automated system management, will drive down the cost of computing just like those previous platform shifts. Consequently, areas in which computing was too expensive will now be economically viable. In economist terms, the lowered cost of cloud computing will bring forth elasticity effects and dramatically increase the use of computing.
This is a rather long-winded way of saying that there will be incredible demand for cloud computing.
Ambal Balakrishnan: There seems to be 2 polarized camps for Cloud Computing. One camp is saying that Cloud Computing is mostly marketing hype and users should keep their information in their own hands. What is your take on this?
Bernard Golden: Those who characterize cloud computing as mostly hype have short memories. It was barely a decade ago that many people characterized the Internet as mostly hype. Where data should be located is a separate question, and as I indicated in the earlier question, is one of the significant challenges to cloud computing. However, I don't believe that data privacy/security will block the growth of cloud computing.
Ambal Balakrishnan: How important do you think Cloud Computing is going to be in the next 5 years?
Bernard Golden: Very important. We are only seeing the glimmers of its potential. As I said earlier, the biggest companies in technology are pouring tens of billions of dollars into this area. Clearly, they recognize its importance.
Ambal Balakrishnan: What are your Cloud Computing predictions for 2009-2010? What percentage of technology companies would start using cloud technology in the next five years?
Bernard Golden: Technology companies are already including cloud computing as part of their roadmaps. I think only a small percentage of technology companies will not incorporate cloud computing into their products and services.
Ambal Balakrishnan: What are the benefits (to customers) of Cloud Computing? Can smaller and mid-sized businesses take advantage of these benefits as well?
Bernard Golden: The SME market is a natural for cloud computing, since these companies typically spend less overall on IT and are also usually short-staffed in IT. Anything that reduces their computing burden while still offering application functionality will be embraced by this market segment.
Ambal Balakrishnan: What do you see as the biggest potential dangers of Cloud Computing?
Bernard Golden: I don't see any particular dangers that will not be addressed or compensated for. People tend to forget that every platform shift was accompanied by people raising "dangers" that the new platform presented. For the Internet, it was "you want your data to travel across public networks?" For client/server it was "you want your data to be spread out on a bunch of machines spread throughout your offices?" While each of these issues was real, the benefits of the platform overrode the concern about the issues and instigated solutions.
Ambal Balakrishnan: What are your thoughts on Amazon Elastic Compute Cloud (Amazon EC2)?
Bernard Golden: Amazon has done a great job jumpstarting this market and they are the clear market leader at this point. Now they're going to have company (competitors) and they'll need to respond. Amazon's offering is impressive and easy to use, and they've rapidly improved it. They'll be a formidable competitor to the new market entrants.
Ambal Balakrishnan: Who are the key players to watch in the Cloud Computing space?
Bernard Golden: Obviously, the platform players like Amazon, IBM, Microsoft, et al are key players. The virtualization players like Vmware and Citrix are important as well. The key area, and hardest to forecast, are the market areas that will be filled by startups, whose products will address new characteristics of cloud technology that the big players can't or don't. These are areas like system management, system lifecycle management, and so on. Small companies that I keep my eye on include RightScale, 3Tera, and so on.
Ambal Balakrishnan: What is your expert opinion on the 'Above the Clouds: A Berkeley View of Cloud Computing' white paper by UC Berkley.
Bernard Golden: I think this is an excellent white paper and well worth reading by anyone interested in cloud computing. I blogged about it in my CIO blog Cloud Computing: What UC Berkeley Can Teach You
Ambal Balakrishnan: McKinsey & Co. report titled 'Clearing the Air on Cloud Computing' claims Clouds are not cost-effective in all cases? What are your thoughts on this claim?
Bernard Golden: Their overall point is that a really well-run data center could match (and beat) the costs of Amazon Web Services; from this, they conclude that companies should avoid using external cloud providers and focus on high-intensity server virtualization. No one disputes that cloud computing may not be appropriate for every application; however, they focused on only one of the three characteristics of cloud computing, which is cost. They specifically abjure attempts to create internal clouds, which are directed at the other two characteristics, scalable resources and easy access without long-term commitment. Merely meeting a cost benchmark without addressing those other two areas is suboptimal.
Ambal Balakrishnan: Please list any Cloud Computing analyst reports (or other reports, documents) you think are "must reads" for anybody looking into Cloud Computing
Bernard Golden: Two important ones that address energy use, which is a significant driver for data center efficiency improvement: one by Microsoft personnel Incenting the Right Behaviors in the Data Center the other by Steve Denegri on Microsoft & Google: Cloud Computing Dominance Through Renewable Energy. I address this topic in my blog post Power: One Cloud Cost Advantage That May Be Irresistible
Ambal Balakrishnan: Bernard, thanks for taking the time to share your expert insights on Cloud Computing.
Bernard Golden: Thank You Ambal.
You can read Bernard's comments on Marketspace's Envisioning the Cloud: The Next Computing Paradigm whitepaper at Cloud Computing Meets Washington: Lots of Data Security and Privacy Questions. To learn more follow Bernard Golden's Cloud Computing Blog at CIO Magazine.
Posted by April 22nd 09
July 2010
