Debunking Five Cloud Computing Myths

For the third year in a row, cloud computing is one of the top three technology investments for CIOs. However, there are many misconceptions of “the cloud”. Indeed, in my travels through public speaking sessions and corporate training seminars on cloud computing, I have encountered five common myths or misconceptions. It’s high time to debunk them.

Myth #1: “The Cloud” is Just One Big Cloud

With the exception of Amazon Web Services which is constantly expanding its data center presence, there is no single cloud of record.  Companies and vendors are standing up cloud computing infrastructures which they make available to the public or internal stakeholder audiences such as employees or suppliers.  

In fact there are hundreds if not thousands of “clouds” in the United States alone (especially when one considers private cloud infrastructures).  For example on the software vendor side, Oracle, HP, IBM, Teradata (full disclosure: the author works for Teradata Corporation) and others are building and maintaining their own clouds. And of course there are B2C “clouds” such as iCloud and DropBox. So the next time someone says, “I’m performing analytics in the cloud”, you may wish to ask “which one”?

Myth #2: One Day Soon, All Our Data Will Be in the Public Cloud

Many cloud experts and prognosticators believe the march to public cloud computing infrastructures—for everyone (corporations and consumers) is inevitable.  Reasons for this line of thinking range from the growing size and complexity of data volumes (i.e. who can afford all this storage?) to public cloud providers should be able to monitor, manage and secure IT infrastructures better and cheaper than individual companies.

While I don’t doubt that public cloud computing will take more market share in the future, I certainly am under no illusion that one day soon all data will be stored in the public cloud—mainly because of bandwidth costs for data transport and costs of doing all your processing on a pay-per-use basis. And of course, recent government snooping revelations help me easily predict that plenty of data will stay right where they’re currently located.

Myth #3: Cloud is Cheaper than On-Premises Computing

This particular myth is a big misconception to overcome.  Corporate buyers hear the word “cloud” and assume it equates to cheaper IT costs. This statement may be true on a low utilization basis—meaning you only plan on using compute power infrequently—but on a full utilization basis you’ll most likely pay more for computing on a pay-per-use basis than maintaining your own IT infrastructure and applications.  For a deeper discussion on this topic, visit “The Cloud Conundrum: Rent or Buy?”

Myth #4: Cloud Computing Means Someone Else Now Has My IT Headaches

Of course, while moving your workloads to “the cloud” means that another vendor—that “someone else”—is responsible for monitoring, maintaining and supporting the information technology infrastructure, it certainly doesn’t mean that your IT headaches go away. In fact, while you may no longer have day to day responsibility for availability, software and hardware upgrades and more, you never really lose complete “responsibility” for IT.

Instead, your day is now consumed with vendor, contract (SLAs) and incident management, workload balancing, application development (for the cloud), and security items such as roles, profiles, authentication processes and more.  Long story, short; you don’t abdicate responsibility for IT when you move workloads to the cloud.

Myth #5: If it’s not Multi-Tenant, It’s Not Cloud

I hear this particular comment quite a bit. Really, the person suggesting this “truth” is stating that the real beauty of cloud computing is taking a bunch of commodity hardware, virtualizing it, and pooling resources to keep costs down for everyone. To be sure, resource pooling is a key criteria for cloud computing, but virtualization software isn’t the only route to success—(i.e. workload management might fit the bill just fine).  

In addition, while multi-tenant most commonly means “shared”, it’s important to define how many components of a cloud infrastructure you’re actually willing to “share”. To be sure, economies of scale (and lower end user prices) can result from a cloud vendor sharing the costs of physical buildings, power, floor space, cooling, physical security systems and personnel, racks, maintaining a cloud operations team and more. But I’ll also mention that there are customers I’ve talked to that have zero intention of sharing hardware resources—mostly for security and privacy reasons.

These are just five cloud computing myths that I’ve come across. There are certainly more that I failed to mention. And perhaps you don’t agree with my efforts to debunk some of these themes?  Please feel free to comment, I’d love to hear from you!