I’ve been taking some time as of late to contemplate the role of DevOps in the enterprise and more specifically within operations teams. In the broadest of terms, the concept of DevOps is meant to break down the barriers between development teams and operations teams, get them all on the same page and automate as much stuff as possible so that the development team can iterate as fast as possible. After a lot of thought I started to ask myself, what about the great divide between different operations functions?
From the perspective of development teams or application support teams, operations are generally thought of as the group of people that make a big fuss when requirements are not set or someone decided to do something out of the ordinary like attract an extra few hundred users to an application in the course of a few days. The reality is, operations teams tend to be made up of highly specialized admins that often feel the same way about each other as developers and application admins feel about them. There are storage admins, windows and Linux admins, visualization admins, people that take care of user accounts and directory services, dedicated messaging and collaboration teams – the list can go on and on and get more and more complex as organizations get bigger and bigger – that don’t necessarily know what each other are doing and nor do they care. Certain things are out of their sphere of influence, expertise and more than likely their manager.
Any organization that is taking a serious look at bringing in DevOps practices into their IT department needs to have a good look at the operations department as well as development, and perhaps even more so. When all of the specializations within operations are working together, understanding what each other are dong and have a common toolkit in place for process, automation, monitoring and reporting, it’s just as nice to watch as a development team using configuration management to deploy a new test farm.