How to Make DevOps Work for You

By: Tomer Levy

The core idea behind DevOps is to create a bridge between development and operations. The benefits are clear: organizations can deliver features more quickly and enhance the quality of product deployment.

But this process comes with challenges because DevOps is a fundamentally different way of thinking for any organization that undertakes it. Dealing with budgeting and resource constraints, overcoming the “developers versus operations” mentality, and adopting new skills are all obstacles that organizations will need to conquer.

These challenges may be too difficult for traditional enterprises to overcome. But there are ways to adopt the DevOps methodology and culture in a way that will reduce friction between development and operations while changing your culture and mindset to make DevOps work for you.

Building Trust and Teamwork

It’s no secret that development and operations view things differently. Although they both want the same result, their approaches are different. This sometimes leads to failures to adopt DevOps cultures and instead results in frustration and even developer burnout when not implemented correctly. The breakdown is caused by developers having the responsibility of streamlining code from development all the way through testing to production and the stress of having to work in an operations-style fast-paced and frequently-changing environment.

Even when developers welcome the innovations that the DevOps methodology offers, operations may feel that they are the only party really interested in the adoption of these new methodologies and that developers still view them as only an “IT support” group. Such thoughts obviously harm the trust and collaboration between the parties.

Bringing Ops into the Mix

One key factor in making DevOps work is having ops personnel integrating themselves into R&D teams. In addition, creating a culture of teamwork between development and operations requires having multi-skilled teams and further embedding dev and ops into each other’s teams. A designated ops person injects themselves into everything the engineering team is doing as early as possible and is involved from the beginning in the product development of any new features that launch.

Ops integration means ops coming into R&D discussions; ops using the same version control system as development so that anybody can reproduce the production environment based on what is in version control; and ops giving continuous feedback to let developers know there is an issue as early and quickly as possible. This feedback in particular will help to minimize any surprises in later testing or deployment.

Hiring SREs

Another approach is to look at a relatively new role known as Site Reliability Engineers (SREs). Site Reliability Engineering was created to help to bridge the gap between development and operations. As a type of support team for developers and the middleman between development and operations, SREs aim to ease any built-up pressure, enable effective development, and use the methods of software engineers to address an operation’s set of challenges.

Senior Management Must Buy-In

Considerable time and effort are needed to make DevOps work. It’s an opportunity for senior management to invest in restructuring processes for more accountability, and reconsider where time is spent and what priorities are. Without this investment, DevOps’ chances for success are minimal. Special ‘guerilla’ teams may need to be set up with dedicated personas to ensure adoption, and teams need the time and support to acquire skills and integrate the tools. In the end, evidence of improvement can help bring added ROI and improvement to the organization, such as increased frequency of deployments, improved service quality and performance, and reduced time-to-market.

Getting the Right Tools in Place

Aligning dev and ops teams while using common tools and methodologies together helps to enhance collaboration, speed up deployment, and reduce errors and downtime. For that purpose, you should ensure that you are using the tools that fit your organization and the frameworks that you use. These include using continuous integration tools such as Jenkins, moving from traditional subversion systems to git, building and maintaining code review tools, and using issue tracking along with release management systems and logging and monitoring tools.

One important tool for DevOps management is log analytics through the use of platforms such as the popular open source ELK Stack. The ELK Stack is the most common log analytics open source solution in modern IT operations, but there is more to it than just service monitoring.

Reaching DevOps Nirvana

The adoption of tools and methodologies such as the ones mentioned above are critical for reaching solid modern IT operations. With these tools and methodologies, an organization can create DevOps culture and mechanics that enable and inspire the pace of innovation – something every organization needs today. Ultimately, one team is created where everyone works together and communicates in an efficient environment, with the goal of producing the best solutions possible, as quickly as possible.