Like most advanced data architectures these days, DevOps poses a dilemma for the enterprise: Do you build it from within, leveraging your already considerable investment in IT technology, or do you outsource it to the cloud, retaining only the barest infrastructure for your critical needs?
This dichotomy is growing even wider now that third-party cloud and platform providers are building end-to-end DevOps as a Service environments that promise to put DevOps in action quickly and easily and relieve the enterprise of many rote tasks required for continuous application support.
Organizations have long had the option to spin up their own Platform as a Service environments and simply tailor them toward DevOps, but the newest breed of DaaS aims for a highly targeted offering backed by enterprise-class tools. A case in point is the new Agile Stacks service that allows a full DevOps ecosystem to be built within minutes using more than 30 pre-tested and integrated cloud and platform components. The service is available in beta on AWS and is built around the DevOps Control Plane, which offers template-based configuration linked to a centralized hub that handles script automation and other functions.
The cloud, of course, is probably the quickest and easiest way to implement DevOps at scale. According to service provider CloudHesive, DaaS offers automated processes right away, as well as the latest virtual development tools. In addition, cloud services allow organizations to forget about infrastructure to focus more directly on app performance and core product support, and also provide a universal platform with centralized communication so all DevOps teams and projects can be managed under the same basic framework. In short, it is the fastest way for organizations with little or no DevOps experience to build out a digital services model to rival the largest players in their respective industries.
But this does not mean the DaaS model is without flaws. As Engine Yard’s Christopher Rigor noted, you’ll still need a fairly sophisticated level of experience with systems integration, workflow and other facets of software development, not to mention expertise with the tools that are available from your DaaS provider. As well, you cannot use the cloud to clone a production environment that you don’t already have, and even creating one for test purposes can lead to a wide range of unforeseeable problems. And then there are the security risks associated with third-party infrastructure services, particularly on the transport layer. This isn’t to say DaaS is unworkable; just that the enterprise should weigh the value of rapid deployment, easy collaboration and simplified management against these potential problems, and perhaps even consider a hybrid DevOps architecture that leverages the advantages of internal and external resources.
Part of this mixed bag of DevOps capabilities may include other service-based offerings, such as container management. Microsoft is the latest provider to launch Kubernetes as a Service, and word is that Amazon will follow shortly, says ReactiveOps’ Corey O’Brien. With KaaS, all the difficulty involved in launching container clusters is removed, giving the enterprise an easy way to push services and microservices to users within a broader DevOps architecture. But again, nothing is trouble-free in modern data environments. For one thing, while KaaS can help monitor an app, it doesn’t set up monitoring and alerting for you and cannot peer into your broader, non-containerized infrastructure.
As with any service-based offering, finding the right provider is probably the key element to achieving successful outcomes. Some providers like to run a very hands-off service, which may suit more experienced clients just fine. Others like to get very involved in clients workflows, which is helpful for novice users.
No matter which approach suits your goals, however, it must be remembered that while your provider answers to you when it comes to meeting performance expectations, you answer to your users. If service is inadequate on your application layer, few people will care that your DaaS provider is at fault. Chances are, it will be easy enough to launch someone else’s app.
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.