In the last decade, cloud computing technology has solidified into an easily-recognizable set of services which can be tapped in whole – or in part – by IT admins. Simplicity of use, portability of data and lowered costs for access have justifiably increased the popularity of the cloud, but led to a peculiar issue: Sprawl.
Opponents of distributed computing often point to cloud sprawl as a serious thorn in side of this technology; as more users spin up their own virtual instances, the supposed cost savings enjoyed by companies can start to dwindle. This unique property of the cloud, however, is better thought of as a starting point for evolution – the real goal isn’t to eliminate sprawl, but control it.
What’s Yours is Mine
Just a few years ago, data portability was the silver bullet. Enterprise IT were concerned that data, once placed in the hands of a provider, would never leave again – what good was an agile cloud if it didn’t play nicely with others? But instead of knuckling under, providers got creative and viable open source options began to emerge, common infrastructure threads which allowed data to move as needed, and without consequence to businesses.
Sprawl is the next step for cloud evolution. Here’s why: Cloud computing offers a form of democratized technology – with the right permissions, any employee can spin up a development instance or access resources. But these same (non-IT) employees often forget about the little cloud bubble they’ve created, and instead of going back to it the next time they need compute power, simply create another. The result? A suburban compute landscape, rather than a centralized one.
Changing the Game
Rampant cloud adoption is good news for developers and IT admins alike, but needs to happen under controlled conditions. This is possible in several ways. First is what Read Write describes as a “horizontal hybrid PaaS,” which offers the ability to put some services in the cloud, keep some close to the chest, and switch between providers as necessary. It also gives admins full view of all resource use, allowing them to track users who prefer to spin up new virtual machines (VMs) every time they need a task completed.
Another possibility is management via a public cloud provider. This takes the pressure off IT professionals, allowing them to focus on app development or legacy improvements, rather than worrying about cloud cost overruns. Many startups and established providers now offer a variety of management services, ranging from basic infrastructure support to complete cloud tracking.
Perhaps the most important evolutionary possibility offered by sprawl isn’t machine-based, but human. Employees, familiar with the public cloud thanks to their use of smartphones and tablets, are often confident they understand how to use it in a business setting, yet unaware of the “waste” they leave behind by improperly managing resource use. This knowledge gap gives IT admins the chance to train employees – not simply as ticket-submitters who wait for professional response, but partners in the effective cloud use.
Cloud sprawl is a challenge, but hardly insurmountable, and provides a unique opportunity to collectively advance a company’s IT environment.