What are the drawbacks of the cloud?
Here are some honest, straight forward limitations to cloud-based apps:
Space: While cloud computing is not totally new; it’s new enough that there are limitations. For example, Windows SQL Azure (the cloud-based storage option for Azure-based apps) has a limited database size. There are recommended techniques to get around the limitation, but it involves more work by programmers (i.e., time and money) to get a larger application up and running. That’s why we advise customer sticking with your existing server or data center if your goal is simply to add new features to an existing application.
Legacy Spaghetti: I’m not going to mislead you: moving an enterprise’s legacy IT to the cloud can be difficult. As Andrew McAfee at MIT has written, “it forces tough decisions about consolidation and standardization. Most organizations that have been around for awhile have a hodge podge of hardware, operating systems, and applications, often described as ‘legacy spaghetti.’ It can’t simply be transferred to the cloud but first must be untangled and simplified.”
Conversely, it’s the perfect time to move to the cloud if you need to completely rewrite an application or develop a totally new one. Ditto when hardware is getting old. While it’s true that connecting cloud-based applications back to legacy databases or other information inside a company’s walls is doable, but it requires extra steps/work. In many cases, it’s often more cost-effective to move to the cloud instead of investing in expensive new hardware for your data center, which leads me to:
Cost: Another concern about moving to the cloud can be the cost. There have been conflicting calculations from such respected organizations as McKinsey and Microsoft on the costs of putting an organization’s entire data center in the cloud. However, MIT’s McAfee believes the focus on cost is misguided for two reasons:
(1) Most companies don’t spend massive amounts on technology, so even substantial changes in the IT budget won’t make a large difference on the income statement (Gartner estimates that for S&P 500 companies, all IT-related costs amounted to just 3.2 percent of revenue, on average, in 2009.) and,
(2) Over time, the economics of building and running a technology infrastructure will favor the cloud.
On point number two, McAfee reasons that economies of scale can be realized by cloud providers in the areas of hardware, bandwidth and power. He also writes that, “because they buy gear all the time, they can take continual advantage of the computing cost declines predicted by Moore’s Law,” citing Amazon Web Services and its dozens of price reductions in the past three years, “even though it does not face intense competitive pressure.”
Security: I’ll share McAfee’s response to security questions, which I find to be completely compelling:
“The security of the cloud is frequently questioned. It’s true that transmissions can be intercepted; firewalls can be breached; viruses, worms and other forms of malware can invade. Perhaps most unsettling, the people responsible for digital infrastructure can steal secrets or get sloppy and let thieves in. However, this is true for every computer network, including the ones that companies run themselves. . . .The only way to have 100 percent computer security is have zero computers. The next best approach is to constantly monitor the threat landscape: buy or build the best technologies to protect devices, networks and transmissions; and hire and retain top digital security specialists. Cloud computing vendors are better able to do this than all but the very largest and most security-conscious organizations.”
Bottom line: there are definite pros and cons related to cloud computing. My next (and final post on this topic) will look at who can help with cloud computing questions and transitions.