# Cloud Computing: Wispy or Real? - Customer Relationship Management

This one is going to be fun. Partly because this is such a hot topic and partly because there is an inordinate amount of confusion as to exactly what cloud computing is. In an April 2009 report that was otherwise maligned by a lot of competing parties, McKinsey identified 22 definitions of cloud computing and that was only the formal ones. As of July 2009, there are roughly 22, 200, 000 results for cloud computing when you do a Google search. So you can imagine the interest in it. This is now even higher with the purchase of Sun Microsystems by Oracle in late April 2009. Sun Microsystems is one of the key players in cloud computing.

There is a certain irony in that acquisition because of a statement made by Oracle CEO Larry Ellison at the 2008 Oracle Open World:

The interesting thing about cloud computing is that we’ve redefined cloud computing to include everything that we already do. I can’t think of anything that isn’t cloud computing with all of these announcements. The computer industry is the only industry that is more fashion-driven than women’s fashion. Maybe I’m an idiot, but I have no idea what anyone is talking about. What is it? It’s complete gibberish. It’s insane. When is this idiocy going to stop?

We’ll make cloud computing announcements. I’m not going to fight this thing. But I don’t understand what we would do differently in the light of cloud.

You’ll note that Larry Ellison isn’t planning on rowing into the tsunami. What he says is, I don’t know if this cloud computing thing is anything new or different, but it seems to have some serious mojo. I won’t buck it so I can make a buck on it.

This isn’t all that Oracle has done when it comes to the cloud. They also struck a deal with Google in March 2009 to create Google Gadgets and to use the Secure Data Connector (SDC) so that the gadgets they create can live in the cloud.

More irony in all this talk about the cloud is that consumers who use technology have been living in the cloud for a long time. The Pew Internet and American Life Project released a September 2008 report on “The Use of Cloud Computing Applications and Services” and found that 69 percent of the respondents use web-based word processing, storage, and e-mail services such as Gmail and like the convenience of having access from any device anywhere. The social customer is not hung up on the issues of data and so forth, but at the same time, to be fair, they aren’t an enterprise either. There is a familiarity with cloud services among the population who have been using them, even though they don’t necessarily know them as cloud services.

What does all this “living in the cloud” actually mean? Why is it important?

What Is It?
The contemporary lore as to how cloud computing got its name is that the way that Internet networks are represented in those awful technology diagrams you can’t avoid is by a cloud. Is that true or urban legend? I don’t know, but it certainly is within the realm of possibility. More germane is to figure out what exactly cloud computing is—and what it isn’t.
First, here is the Wikipedia definition, which isn’t bad:

Cloud computing is a style of computing in which dynamically scalable and often virtualized resources are provided as a service over the Internet. Users need not have knowledge of, expertise in, or control over the technology infrastructure “in the cloud” that supports them. The concept incorporates infrastructure as a service (IaaS), platform as a service (PaaS), and software as a service (SaaS) as well as Web 2.0 and other recent (ca.2007–2009) technology trends that have the common theme of reliance on the Internet for satisfying the computing needs of the users.

Next, here is a definition that I’m aggregating given the many other definitions I’ve read, the research I’ve done for this book and elsewhere, and factoring in the consulting I’m involved with that crosses the path of cloud computing:

The use of a pervasively connected Internet-based computing platform to source services, applications, and infrastructure employing a consumption-based (pay-for-what-you-use) pricing model. It is the deployment of your business workload on the Internet.

The Components of Cloud Computing
I trust that definition is satisfactory? If not, I can’t do much about it, but continue to read on regardless. Cloud computing isn’t as vaporous as a real cloud is. There are specific components and services that it encompasses and specific benefits that it has. Before we get into the benefits, let’s peruse the components.

Infrastructure as a Service (IaaS)
This is perhaps the greatest differentiator between SaaS and cloud computing. What cloud computing provides under the aegis of IaaS cloud infrastructure is to deliver a connected grid, either real or virtual, combined with a utility-or subscription-based billing environment. The most intriguing permutation of IaaS is the deployment of a virtual server environment, rather than a physical server. Where SaaShosted models provide a physical server carrying multiple (or single at times) clients on a single instance of the application, the cloud computing model often provides a virtual server such as those provided by EMC’s VMWare, that allow resources to be dynamically reconfigured based on the scale of the workloads at any given time. That said, no matter how much you “virtualize” the server environment, there is going to be a physical server somewhere. It just doesn’t matter geographically where it is.

The Amazon Elastic Computer Cloud (EC2) is an excellent example of how this works. To use it, you create what Amazon calls an AMI (Amazon Machine Image) that includes your existing applications, libraries, data, and configuration settings. This is loaded into their Amazon Simple Storage System (S3) repository, which consists of redundant servers in multiple data centers across the world. This is a virtual version of your business workload stored in a virtual storage area that is created by redundant connected physical servers in many locations.

Platform as a Service
Platform as a service (PaaS) in the cloud is designed for the facilitation of the development and subsequent deployment of applications without the physical hardware and software. The obvious benefit here is that it eliminates the overhead costs of that development and deployment, such as operating system maintenance or network connectivity. Most of all, you don’t have to buy software or hardware.

Salesforce.com’s Force.com is an ideal example of this kind of platform, though it uses Apex, which is a proprietary development language. Another example would be the .NET platform from Microsoft as it has been reconfigured for Microsoft’s cloud computing effort, Azure.

Cloud Services
This component is perhaps the most difficult to truly define because of conflicting ideas on what cloud services are. For example, in that controversial McKinsey report, they distinguish between “the cloud” and “cloud services.” To them, “the cloud” has to comply with three requirements:

1. Hardware management is highly abstracted from the buyer.
2. Buyers incur infrastructure costs as variable OPEX (operating expenses).
3. Infrastructure capacity is highly elastic (up or down).

The “true cloud, ”as McKinsey calls it, has to meet all three requirements. Cloud services only need to meet numbers 1 and 3. They identify Amazon EC2, Microsoft Azure, and Google as “true clouds, ”with Zoho, salesforce.com, and Gmail as “cloud services.” I think that McKinsey might be overcomplicating it though I agree that Zoho and Gmail are cloud services. salesforce.com is more than that.

A cloud service is a product, service, or solution that is consumed via the Internet. Interoperable web services that communicate from machine to machine in a distributed environment are cloud services when they are available through the use of web-based software. Rearden Commerce using web services to provide business services like airport car rental or Zoho providing sales force automation via the Web and not as a hosted service would be ideal examples of how a cloud service works. It is characterized by an interface that is directly accessed via a browser from wherever you are.

Storage
This is an important distinguishing characteristic when it comes to understanding the difference between SaaS and cloud computing. In the cloud, data is stored online as a service—meaning that you pay for the amount of data you store at a rate typically based on gigabytes per month. This is not based on the purchase of a hard drive. In effect, you’re not paying for the real estate or the storage locker, you’re paying for the weight or volume of what you’re storing. Of course, data is stored in the data centers of SaaS-based hosts, but you aren’t typically charged for use of the physical storage as long as you keep within reasonable limits. The basic storage cost is built into the subscription price.

One of the simplest examples, though hardly a paradigm of how to do it well, is Apple’s MobileMe. You can upload files to the Web in the MobileMe storage space—up to 10GB—and then synchronize it with your data on your PC or, more likely, Mac. Amazon’s S3 is also a web services–based cloud storage capability. That’s where the AMI is stored, as you saw above.

Applications
The easiest way to think about cloud-based applications is to just think about Zoho. Zoho is a cloud-based collaboration suite that provides dozens of products, all run through your browser. Zoho has 29 applications, add-ins, and plug-ins accessed from a web browser. Take a look at Figure for the home page, which is not only a listing for the applications but a portal for accessing them.

Zoho applications—a list and a portal

Theoretically, all the Zoho applications could be accessed through a single computer in any location from anywhere—or a single application could be delivered via a grid of distributed computers. The user has no idea how it is being delivered or accessed and most likely doesn’t really care.

Web 2.0 Characteristics
While the majority of cloud computing players have social features built in, it isn’t a mandatory component for cloud computing environments. However, this hasn’t stopped some of the vendors from developing social cloud services. For example, IBM launched a collaboration and social networking cloud service called LotusLive Engage, which links LotusLive applications like Live Meeting or Live Activity—their cloud apps—to more traditional on-premise Lotus applications like Notes, Domino, or Lotus Connections. They call it, somewhat cutely, “Click to Cloud.” The idea is that it allows employees to bridge the firewall by rolling all applications into the cloud whether they are onpremise or on-demand. It even provides an application called Content Collector that manages e-mail and instant messaging in the cloud, something like managing Gmail and MSN Messenger.

Utility Consumption Pricing
Just to be clear, this is not just a subscription model. Subscription or even traditional pricing is an option that can be provided to the users. But the idea behind the pricing model is pay-for-use. What you use is what you pay for—no more, no less. For example, Amazon EC2 pricing works like this:

• On-Demand Instances This is pricing by the hour to run your AMI. It varies depending on geography (U.S. and Europe); size of the “instance” from small to very large; whether the instance is running on Linux or Windows, with special pricing for an instance running on a virtual SQL server, Microsoft server, or IBM.

• Reserved Instances This is an option that lets you make a one-time payment with a one- or three-year commitment that significantly reduces your usage (hourly) cost. Same variances apply.

• Internet Data Transfer There is additional cost for data transferred into the system or out of the system. There is a single price for data transferring into the system. Outgoing data pricing varies based on estimated amounts of data, from 10 terabytes to 150 terabytes per month. There can be variations based on region and zone.

• Elastic Block Storage (EBS) This is the persistent storage that exists for the life of the instance. When the instance ends, so does the provided storage. The charges vary by location (U.S. and Europe). They are for the amount of gigabytes used per month and for each million I/O requests. They also charge for the number of EBS snapshots made per month that are stored on Amazon S3—their storage “backup” for EC2 in this case- though it does more than that.

• Elastic IP Addresses This has nothing to do with what holds up your pants. It is a static IP address that’s assigned to your account and that you can use to remap your instance in case of failure. The charge is a bit unusual. You’re charged for non-use. When you are using it there is no charge. However, if you’re remapping it to your instance—you’re charged a “remapping fee” in effect.

Actually, if you distill the pricing, you pay for use of storage, bandwidth, volume, location, and AMI activity. All in all, even though I didn’t specify the numbers, it is quite small per unit, though since we are dealing with an enterprise, you could end up paying a fair chunk of change. But it still is less than the costs of your own environment.

Benefits of Cloud Computing
To look at the benefits of cloud computing, we’re not going to do something so mundane as another chart. This time, we’re going to take a case study, examine it, and break out the benefits based on the results of the case study. If you have any questions, please just blurt them out.

Case Study:DISA RACE. What?
The Defense Information Systems Agency (DISA) is responsible for the management of IT for the Department of Defense (DoD). Historically, the DoD had a problem with buying expensive equipment for product development and testing. When the effort was completed, the equipment remained on the books depreciating with only roughly 15 percent of its capacity having been used. A huge, wasteful, but still necessary expenditure.

In 2009, to deal with this problem, in the name of efficiencies, DISA launched RACE—the Rapid Access Computing Environment. It was aimed at creating a development environment that could be accessed from the cloud through a portal for only $500 a month. Because of the low subscription price, the users could charge the cost on a government agency’s credit card—which allows purchases under$2, 500 without too much paperwork.

What made RACE so attractive was that it was a complete development environment, providing all the tools, services, and infrastructure that were needed. Capacity planning wasn’t necessary. Spending ungodly amounts of taxpayer dollars wasn’t necessary. Rather than spending months to set up the physical servers and install the software and then watch much of it go to waste when the testing was complete, it took no more than 24 hours and usually considerably less to provision the entirety of the infrastructure and applications needed.

When the effort was underway, the environment could be accessed at the office or at home, on whatever device it needed to be accessed from. When it was done, the environment just retreated to the cloud-no muss, no fuss, no wasted capacity.

Breaking Down the Benefits
The benefits of cloud computing become apparent when you start to analyze RACE.

1. The RACE environment is provisioned in no more than 24 hours and typically well under that. This means that there is no vendor selection process, no technology implementation period, no capacity planning necessary.
2. The $500 per month subscription fee for unlimited use is a controllable cost that can be easily charged to a government credit card, bypassing what can often be an onerous procurement process. 3. Since the RACE environment is based on economies of scale, when the customer base grows, the price will decrease. 4. The development environment is complete and custom fit to the requirements of the agency that needs it. Since the environment actually resides in the cloud, there are no maintenance costs that concern the agency—they are covered in the monthly subscription fee. There is no overhead, keeping costs down. 5. It is accessible anywhere through a web portal. That means that the users will be able to work in optimal environments of their (and the agency’s) choosing. 6. Finally, when the development and testing are complete, the environment is just rolled back into the cloud so there is no underutilized capacity—utilization is a theoretical 100 percent. Additionally there is no cleanup necessary on the part of the agency when done. Potential Problems with Cloud Computing This doesn’t mean that cloud computing is perfect. There are definitely problems, though they tend to fall into the category of “immature, ”not “horrible breakage.” Even in that regard, there aren’t that many egregious ones. First, economies of scale, which is a very professional sounding phrase, are pretty much theoretical at this point since most companies are not working in the cloud. Additionally, the cloud hosts each have unique cloud infrastructures so there isn’t exactly easy interoperability in the cloud. Some of the concerns are the same concerns that people have had about SaaS: data security and control, downtime, and so on. In fact, McKinsey took a look at the uptime of cloud hosts and found it ranged from 99.5 percent to 99.9 percent, which led to concerns about the lack of availability of cloud-vendor service level agreements. Most of the issues are due to an immature model trying to find its way. But cloud computing seems to be growing up fast. Gartner Group issued its annual technology forecast at the end of March 2009 and, despite what they saw as a major slowdown in IT growth (3.9 percent), one of the few areas where growth was on fire was cloud computing. Expenditures on cloud computing were projected to grow 21.3 percent from 2008’s$46.4 billion to 2009’s \$56.3 billion.

Case Study:Cloud Computing Does It All
BrandSCAN is an interesting company that focuses on using contemporary media for market intelligence and brand insight. What makes them interesting is not their product, which is interesting unto itself, but their development model. What they’ve done is take the entire production process from concept to product and combine outsourcing services with the cloud. Bruce Culbert, industry thought-leader and managing partner of BPT Partners, LLC, will take it from here:

BrandSCAN’s founders hired a U.S.-based development company who had their programming operations offshore. The founders of the company created concept, design, and user interface documents, which were then given to the development company to turn into a product. The product idea was straight forward. Collect actionable insight, information, and conversations from the Web and social media. Then create a functional easy UI that integrates with key CRM solutions like salesforce.com. The back end is run by a .NET architecture and a SQL database.

The idea was to take this configuration and, using Force.com, create a framework that not only used salesforce.com but also could integrate various products via AppExchange. The back end application would use SOA and be hosted on Amazon’s EC2 cloud. What BrandSCAN was able to do was to start with a concept and go to product release without any investment in fixed IT costs or in infrastructure. Because it runs in the cloud, the scalability is very flexible, meaning it can be deployed starting small and scaling higher—and still be priced based on usage.

Similar to RACE but in a commercial cloud environment, Brand- SCAN is a good example of a small company that simply figured out how to be cost-effective and remarkably efficient. Not a bad combo.

The Players
It’s very hard to detail the players in this space. You’ve seen a lot of Amazon in this chapter because they were one of the first vendors who figured out that the cloud was a good place to be. They had a lot of unused capacity—and voilà, EC2 and S3 were born.

But there are several other players. EMC, which owns the virtualization powerhouse VMWare, recently introduced VSphere4. That gives companies the opportunity to create a private cloud environment behind their firewall—which might beg the point of the cloud but is still a viable idea. salesforce.com uses its PaaS framework Force.com to provide its cloud infrastructure. Google has been providing consumer cloud services like Gmail for a long time and is moving into the enterprise, though exactly what direction they are taking to do that remains to be seen.

Perhaps most intriguing is the April 2009 acquisition of Sun Microsystems by Oracle. Even though Oracle probably didn’t buy them just for their cloud services, it is a part of what Sun provides. Given the comments of Larry Ellison earlier in this chapter, it will be fascinating to see what Oracle does with Sun’s cloud capabilities.

Microsoft released its Azure cloud infrastructure, which they describe as a “cloud services platform” in late 2009 after announcing it in late 2008. This one has a lot of promise because it has been released in “flavors” such as Azure for web developers, corporate developers, system integrators, independent software developers, and business. They’ve already thought through who was going to use their platform and created the environments and toolsets to meet the requirements.

There are others-companies like IBM, 3Tera, and Hewlett- Packard. In fact, a list compiled by John M. Willis, in his IT Management and Cloud blog, found 48 companies in 2008 that had some relationship to the cloud as a server or a service provider or who had an application that lived in the cloud. We can expect to see lots more vendors in the near future. Take note here. The cloud is not only here to stay, but in a few years (I’m hedging a little) might be the way to go.