As more and more enterprises become comfortable with the concept of using the public cloud to replace portions of their on-premise infrastructures, I've seen growing surprise among IT organizations that the cloud does not necessarily save them any money. I'm not entirely sure what has given rise to the widely held impression that the cloud is always cheaper than on-premise infrastructure, but focusing solely on cost misses the point of what the public cloud is really good for.
The most obvious benefit of the public cloud is that it lets you access a very small slice of an enterprise-class infrastructure at prices that directly reflect the size of the slice. For small businesses that want a couple of servers or a small amount of storage for backup, the cloud can usually offer what you need much more cheaply than any similarly priced on-premise option, even over a long period of time.
The reason is those small workloads are not large enough to fully use the on-premise hardware and software that you would have purchased for internal deployment. If your needs grew to the point where you could fully use an enterprise-class backup, storage, and virtualization infrastructure, the apparent cost benefit of the cloud would melt away.
However, that does not mean the cloud isn't a worthy option -- it simply wouldn't be cheaper anymore.
Today's cloud strategy may give way to tomorrow's on-premise approach
Recently, I saw a great example of this play out. Several years ago, a software development company needed to spin up a few Linux servers to act as a test bed for the application it was developing. Rather than try to secure the capital to purchase new hardware for what seemed like a short-term need, it opted to fire up a few servers on Amazon's EC2 platform. Since then, the application has become the company's bread and butter, and those first few servers grew into a relatively static army of hundreds of servers.
Today, the price of deploying those servers on-premise is less than what it would cost to construct an on-premise virtualization infrastructure capable of running the cloud instances and the expensive connectivity to the cloud for users to access the infrastructure. Plus, an on-premise deployment at the company's current usage scale would provide substantially better overall performance. Though the initial use of the cloud was an exceptionally good one -- leveraging both the pay-as-you-go and elasticity advantages that the cloud can deliver -- this particular company has effectively grown out of those benefits and is very likely to move the bulk of its workloads back on premise.
That company's experience may be somewhat unusual, but that's ultimately the point -- the cloud is a tool and, like any tool, has a specific set of benefits and disadvantages that prospective users need to weigh carefully. In the end, cost is certainly a factor, but it isn't the only one.
This article, "The cloud isn't always cheaper -- and that's OK," originally appeared at InfoWorld.com. Read more of Matt Prigge's Information Overload blog and follow the latest developments in storage at InfoWorld.com. For the latest business technology news, follow InfoWorld.com on Twitter.