08-08-17 | Blog Post
Cloud sprawl is a major issue for many IT organizations, and leads to many problems such as rising costs, security vulnerabilities and threats to data integrity. It’s become such a concern that optimizing cloud costs is the top initiative for 53 percent of organizations, according to RightScale’s 2017 state of the cloud report. But what is cloud sprawl, why did it come about and how can you keep it in check?
In a nutshell, cloud sprawl is excess machines or workloads running, oftentimes without the company’s knowledge. Sprawl happens when an organization fails to properly monitor manage or control its cloud environment. This leads to headaches across the entire IT department because its effects are felt by everyone–from the IT admin who has to round up all the “wild” cloud instances in the environment, to the CIO who must own up to the high cost of running those extra unnecessary workloads.
As Nick Lumsden put it, sprawl exists because of a storage problem more than a compute problem. With the rise of Big Data, suddenly sprawl became an expensive storage and storage density problem, especially as workloads continue to move into the public cloud. People need somewhere to put all that data, and they need resources assigned to it. Why not store it “cheaply” in the cloud?
What does that mean for Joe Company? Well, sprawl not only presents a financial problem in the form of rising costs to support workloads that aren’t even being used, but it also presents security and data integrity problems. Unmanaged or unmonitored workloads means more potential attack vectors for hackers. Forgotten workloads that are running in QA or dev/test environments are likely to be using real production data, which, if it doesn’t already threaten customer data when it’s being used in live dev/test environments, certainly poses a major risk in idle, unwatched environments. In fact, 80 percent of organizations admitted to using and storing production data in dev/test environments, according to Delphix.
Sprawl also exists in the form of over-allocated resources for a particular workload. This includes idle or forgotten workloads that no longer need resources attached to them, non-persistent workloads that are running 24/7, and workloads that have been given more memory, computing and storage power than they really need. All of these examples lead to increased costs and decreased efficiency.
Unfortunately, there aren’t many service providers interested in eliminating cloud sprawl as much as a client organization. This is because service providers make more money when there’s more machines running. Why shoot themselves in the foot?
However, this problem doesn’t apply to everybody. It’s possible to find a service provider who will work with you to optimize your workloads AND help you manage your costs. Online Tech’s SprawlGuard™ technology is designed to do just that. Using advanced usage analysis to determine your monthly spend, so you can adjust as needed. We’ll also alert you to abandoned or idle workloads, giving you more cost optimization and total control over your environment.
Cloud sprawl is when you’re running more cloud instances than you need, which leads to wasteful spending and can hamper productivity and also poses a security risk. Sprawl can pose itself in a variety of ways, including idle or forgotten workloads, non-persistent workloads and over allocated workloads. While most service providers encourage sprawl (or at the very least, don’t discourage it), there are those out there who will work with you to not only optimize your current workloads to manage costs but send you alerts when you’re close to exceeding your monthly budget and help seek out those workloads that may be contributing to cloud sprawl.
