Clouds are made of vapor, right? Does that mean that cloud technology is vaporware? Yes and no. I’ve struggled with forming opinions on enterprise-class cloud computing for a while, and I’ve struggled with basic questions such as what’s really the difference between having your database in a cloud and having your database (and all the apps that go with it, of course) in a co-located facility somewhere out in the ether. This distinction might have been obvious to you for a long time, but it’s only really begun to gel for me over the past few months.
Let’s ignore technology for just a bit and focus on some of the process and business-level differences between having a cloud solution and hosting your applications and databases somewhere else. The differences are fundamental.
Everyone likes to use the example of an electrical grid when you talk about cloud computing. You just plug your stuff into the wall and you’ve got power, right? As in, you don’t need to worry about where the power comes from. Well cloud computing isn’t quite as seamless as plugging your toaster into the wall and just expecting it to work. However, cloud computing is a paradigm shift from what most of us have built our careers around, and I’ve come to believe that cloud computing in its various forms will have as transformational and disruptive impact on business computing as the Internet has had over the last decade.
Today, the corporate data management function spends vast sums of time and money performing backups, worrying about where data is on disk, and worrying about scaling applications up and out after it’s been built. Much of this activity is performed by highly paid staff, and it consumes a large portion of their time. Unfortunately, many of these tasks don’t really provide much in the way of unique or competitive advantage to the business. Backups have to work, of course, but does anyone really want to manage their own backups if they don’t have to? Frankly, I don’t think so. Theoretically, all of those needs simply dissipate like vapor in the cloud. The truth, of course, is that cloud computing technology, and cloud database technology in particular, simply isn’t mature enough to replace all of what happens today. But I’m pretty sure that people used to say that about web development compared to traditional client/server technology a decade ago. However, I don’t think it will take a decade for cloud computing to mature because it simply makes too much sense.
Small businesses typically have at least one person who performs DBA tasks at least half the time. Large businesses, of course, can have many, sometimes dozens, of DBAs. How many servers do you have? And how much do the resources and staff that you need to maintain and support your server environment cost? Well, they cost an awful lot, even for a small business if you’re measuring as a percentage of actual profitability.
Infrastructure tasks don’t vanish in the cloud. Somehow, somewhere, backups need to happen and servers need to exist. But the industrial revolution that started back in the 18th century taught us a thing or two about economies of scale in automation didn’t it? I don’t have statistics or Gartner-endorsed ROI white papers that I can point to that show how much it costs for a single DBA to back up and maintain a single database. But surely economies of scale will apply when this work is magically being done in the cloud, won’t it?
Most of us who come from a technology background tend to focus on technology-oriented aspects of the decision-making process. Sometimes that makes sense. Is the cloud technologically mature right now? Yes and no. But cloud computing has the potential to lower costs of doing business so dramatically that it simply has to work. No one built their own Model T back when Henry Ford revolutionized the car business, and not too many people or businesses generate their own power. Do you really want and need to pay for engineers to do all that’s necessary to make cloud computing work or do you just want to plug your toaster, or application, in and have it work?
This isn’t necessarily a paradigm change that will be driven by IT people. Frankly, many IT people will get dragged along kicking and screaming, perhaps with a pink slip. But the potential cost savings associated with cloud computing are simply too great to ignore. It has to work; therefore, people will find a way to make it work.