As stated in part one, sometimes DevOps implementation can go wrong. In nearly every case, however, problems are avoidable. Here’s the second part of our list of some bad scenarios and how to avoid them. To see the first part of the feature, click here.

3. Poor DevOps coverage. Many companies making the switch to DevOps understandably wish to save money. But sometimes along with bargain rates comes systems that lack complete coverage.
   As an example, a company that relies both on metrics and external active monitoring cannot merely implement a metrics based solution. Even a low-visibility hole in coverage, such as that coming from outside interference, can create an outage.
4. Underutilizing people and process. DevOps can be seen as a solution to human error and poor judgement. But it is, in fact, a very human process. While tools and automation assist developers and operators greatly, people and the culture that DevOps inspires within them is the key to successful deployment.
   Even a system that seems to be technically perfect needs to have a team behind it. That team must be part of the collaborative and communicative culture of DevOps. Technical issues are always solvable, but if the core of the system—the people—are not wholeheartedly involved in the process, the system will indubitably break down.

DevOps is the most promising solution to software delivery acceleration. When realized properly it can become indispensable to a business, but disaster is often just around the corner for those looking to cut corners. The best DevOps often comes from the companies most invested in a smooth and permanent transition.

For more critical DevOps tips, download our popular white paper: DevOps Misconceptions and Best Practices.