Wall Street Journal Investigates What's Behind the Cloud
Confirming what capacity management experts have believed for years, the Wall Street Journal writes that the cloud is more complex than it might first seem, and that such complexity easily translates to costs.
Is the cloud a one-way ticket to slimmer budgets and increased performance? Considering the hype that cloud computing technology still enjoys, one could easily think so. However, a look that the Wall Street Journal took into the cloud over a year ago should have left no doubts that this kind of simple solution to complex problems of IT cost management simply doesn’t exist. Unfortunately, it’s a lesson that many companies have yet to learn.
In February 2015, the newspaper interviewed industry veterans who had fully or partially migrated their IT operations to the cloud in order to gauge how the services operate in practice. Contrary to common notions about the technology, the Journal concluded that, “Absent proper expense controls... the cloud can be exceptionally wasteful of expensive resources.” 17 months later, the industry still wrestles with the issue.
How the Cloud Becomes Expensive
The cloud can provide for essentially “on-tap” IT resources to its customers. But as WSJ’s Clint Boulton writes, this creates an environment where, in practice, many companies tend to leave the tap on: “Ordering too much computing power can be as easy as over-ordering at a restaurant or leaving the water running at home.” Yet it can be difficult for organizations to spot when they’ve spent too much because, unlike with water bills, they often lack a clear frame of reference for what constitutes a “normal” cost for a given amount of usage.
Many companies have trouble determining how much cloud capacity they will need, as well as managing that capacity as their needs evolve. As a result, organizations routinely overprovision cloud resources, not as an accidental slip up, but as a planned decision. In some cases, this simply includes purchasing larger virtual servers than necessary; in others, enterprises don’t realize that cloud servers should be right-sized by the hour -- unlike physical servers, you can downsize the cloud during off-peak operating hours. Indeed, perhaps the cloud’s biggest advantage is the ability to turn it off.
Adrian Cockroft, the cloud architect who oversaw Netflix’s transition to Amazon’s AWS service, told WSJ that “If you build applications that assume the machines are ephemeral and can be replaced in a few minutes or even seconds, then you end up building an application that is cost-aware.” This happens less than half the time, however; the Journal estimated then that as many as 60% of enterprise-level cloud servers could be reduced or eliminated.
Making Costs Visible
In other instances, organizations seemingly understand their cloud spend, but focus on application performance to point of excluding capacity management altogether. Poorly-performing applications give the impression of efficiency in such cases because they’re sucking up cloud resources, giving the impression that their cloud service is sized proportionally to their needs. If the cloud is like a running tap, these IT infrastructures are plagued with leaky pipes.
Did WSJ find a solution to these issues? At the end of the day, IT operators were forced to dig through their systems to find underutilized services and flag clear instances of overspending (such as cases in which engineers left systems running 24/7). Ultimately, the message was cautionary: “We’re going to be more careful about what we put in the cloud,” said Jamie Cutler, CIO of QEP resources. His firm had contracted with a cloud vendor only to realize later that the vendor couldn’t provide a customized function he required.
But luckily for companies today, hindsight isn’t the only view into capacity that’s 20/20. With TeamQuest’s Vityl suite of capacity management tools, organizations can ensure that their cloud spend is appropriately sized for their current and future business needs. When the chief issue with the cloud is a lack of knowledge surrounding capacity needs, the best solution is one that leverages comprehensive IT systems information to lend decision-making power.
Although the cloud doesn’t guarantee savings, there’s no reason why you shouldn’t be able to do the work necessary to achieve them.