Containers are said to be the key to effective devops in the enterprise. But if this is the case, why are so many enterprises still launching the bulk of their applications on virtual machines more than four years after the Docker container platform made its debut?
Part of the reason can be attributed to simple inertia. Most large enterprises evolve around a culture of risk-aversion, so they want to make sure that every step forward doesn’t result in two steps back. But some of the delay can be attributed to the need to integrate containers into existing data architectures, which has proven to be much more difficult than previously believed.
For the development side of devops, of course, containers are a no-brainer. As Red Hat’s Gordon Haff noted to The Enterprisers Project recently, the advent of containerized microservices means that coders no longer have to build their projects from scratch but can simply assemble what they need from previously written bits of code. This modularity provides a great deal of flexibility to add and remove functions as needed, and will be particularly useful as automated services emerge in Big Data environments and the Internet of Things.
The problem, says Adobe’s Matt Asay, is that the technology is still trying to make peace with the operational side of the devops model. According to Cloud Foundry, container adoption jumped from a measly 22 percent of the enterprise workload last year to a slightly less measly 25 percent while virtual machines, which already hold more than 100 million x86 workloads, gained 8 percent. In many cases, then, organizations are allowing developers to work with containers, but when it comes to deploying these products into production environments, they turn to virtual machines. This is the main reason why AWS teamed up with VMware last month: The cloud is primarily about operations, and so far at least, the enterprise chooses to operate on virtual architectures.
Still, there are those who would prefer that the enterprise pursue an all-container, or at least a container-heavy, approach to operations. In the past month, we’ve seen Microsoft announce a new initiative with Red Hat to more easily integrate Windows Server and the Azure Cloud with the OpenShift Container Platform, which both companies say will enable streamlined development and deployment across mixed Server/Linux infrastructure. And more recently, the once cloud-averse Oracle has joined the Cloud Native Computing Foundation to support the Google-backed, open-source Kubernetes container management stack, which should make it easier to integrate Oracle’s enterprise platforms with public clouds.
Still, there are many hurdles to leap before containers can make real progress on the operational side. According to CIO.com’s Marc Wilczek, container architectures need to confront issues related to capacity and lifecycle management, as well as networking and monitoring or face the prospect of a swirling mass of services and microservices that consume resources at an exponential rate. Advanced analytics should help with all of these issues, but until that technology is perfected, organizations would be wise to tread cautiously when implementing native container architectures in production environments.
It seems, then, that containers are pointing out the stark reality of devops: Just because you can combine these two functions under a single marketing term does not mean they are one and the same. Both will continue to pursue separate goals and utilize tools and infrastructure in different ways.
Rather than deploying containers to achieve a devops model, perhaps it would be wiser to improve the devops model to accommodate containers.
Arthur Cole writes about infrastructure for IT Business Edge. Cole has been covering the high-tech media and computing industries for more than 20 years, having served as editor of TV Technology, Video Technology News, Internet News and Multimedia Weekly. His contributions have appeared in Communications Today and Enterprise Networking Planet and as web content for numerous high-tech clients like TwinStrata and Carpathia. Follow Art on Twitter @acole602.