Can We Do Without DevOps?

Recently, DevOps has received so much hype and every respectable company sees it as its duty to have a DevOps team on and running. It is considered to be so beneficial to any company that a non-technician scrolling down posts at any tech website would think of it as a universal remedy for all the corporate diseases ever possible. However, given that a definition of DevOps varies from company to company and no one exactly knows what a true DevOps team should do, isn’t there some misconception between the praise that DevOps has been receiving and the real output? Let’s dig deeper and decide for ourselves.

Yet, despite the buzz around the notion, DevOps is said to be a working thing. It guarantees a company-adopter 200 times more frequent deployments, 24 times faster recovery and three times lower change failure rates, according to Puppet’s State of DevOps Report.

That is why, to get a non-biased view on why one should “DevOps-ize” corporate processes, in our article we give the answers to the following:

• What DevOps is?
• What issues is it supposed to help resolve?
• What does a DevOps engineer do?
• What skills does a DevOps engineer need?

All of these questions, when answered, will help us see why the industry has adopted DevOps as a regular practice, why your company needs to try it out and why we can’t do without it the way we used to before—or can we?

What Is DevOps?

According to the AWS, DevOps is the combination of cultural philosophies, practices and tools that increases an organization’s ability to deliver applications and services at high velocity: evolving and improving products at a faster pace than organizations using traditional software development and infrastructure management processes. Well, this definition may not be as clear as the waters of Sardinia, so we’ll clear it up for you a little bit.

The message is that once (before DevOps) the product release was slow and troublesome, while today, with DevOps practices in place, it almost reaches cosmic speeds and high-end quality. All through maturation of work approaches and practices–merge of development and operations–along with the automation, cloud technologies, virtualization and abundance of top-notch tools. DevOps culture to a certain extent enables developers to perform the work of the IT team, and vice versa, which breaks down the operational walls between the teams, eroding a “not my task” mindset at the same time.

What Does a DevOps Engineer Do?

If that’s what is happening today–programmers deal with the configuration issues and the IT teams code–why is a DevOps engineer on the tip of everyone’s tongue? Why come up with a new term if you can write some more new responsibilities into the dev’s/IT guy’s contract and leave the names as they are? The thing is, the transformation of the processes and roles was so substantial, that in time they merged into a completely new position, which is neither a coding IT guy, nor a configuring developer.

To sum it all up, a DevOps engineer is a so-called “centaur” performing duties of a system administrator between coding sessions. His/her primary objective of the project is to find a solution that will make product deployment fast and efficient and people’s (devs and ops) lives easier.

What Skills Does a DevOps Engineer Need?

According to the Qulix’ practice, an ideal skillset for a DevOps is:

• A technical graduate.
• Intermediate or advanced English.
• General programming skills.
• Proficient PC-user knowledgeable about networking, Git, Linux, Docker, Nginx, Bash.
• Strive for constant perfection and thirst for the new.
• Curios, with inner drive to get to the bottom line, enthusiastic.
• Strong work ethic and time management skills.
• Open and a great team player.

These will do for a start. At the same time, within the same company various departments may require from their DevOps engineers to have a command of different technologies. For example, another department of ours expects from their candidates to know TCP\IP, DNS, Nginx, Tomcat, Git, Bash, Groovy, Ansible, Docker, Ubuntu, Debian, Macos, Gradle, Maven, CI/CD.

As you can see, an ideal fit for the position of a DevOps engineer within a specific department is something rare. All because DevOps engineers have to have so much on their plate, both from technical and soft skills perspective. Why? First of all, as a system administrator to a certain degree, a DevOps engineer has to operate numerous technologies, that sometimes far outweigh the number of those a regular developer uses on a daily basis. In this connection, I once asked a friend of mine working in a DevOps team, “Of all the difficult things that you have to deal with as a DevOps guy, which one is the most difficult one?” And his response was “Everything.” 

The “everything” of my friend needed further explanations, for which I headed to another DevOps engineer of Qulix Systems, Denis Golovnev. He told me he opted for a DevOps position because he wanted to get to know the whole SDLC closer and in depth, as “this position requires a 360-degree knowledge of the SDLC processes.” DevOps guys are among those few specialists who stay on the project from start to finish, to maintenance. That is why “a DevOps engineer won’t only have to dig deeper into his/her own area of knowledge, but also target adjacent spheres to see how the product is created and how his/her actions can create more value for the client,” explained Golovnev. Thus, this made the tech part clear to me.

What about soft skills?

“The most difficult thing about DevOps is to arrange the processes in a manner that will suit everyone; the one that will encourage the initiative and individual decision-making process, to a reasonable extent,” said Golovnev. “To achieve that, highly developed soft skills are a must to get to the root of the problem and enable an engineer to outline clearly his/her point of view to everyone.”

On the other hand, DevOps team will inevitably face both technical and organizational issues, which should not be confused.

“One should not try to settle human-related issues through technical means,” said Golovnev. This is another reason why a candidate without highly developed soft skills is very unlikely to get the position. Soft skills are the absolute killer thing for a DevOps candidate.

To Wrap It Up

We’ve been gradually migrating from slow and troublesome product releases and delivery of separate work chunks to a speedy, orderly and consistent performance. Most definitely, this transition isn’t the consequence of DevOps practices implementation alone. But without it, without the culture of speedy collaborative delivery, dynamic issues resolution and ultimate automation, which are typical of the DevOps philosophy, the shift would be impossible.

Furthermore, the revolution starts from minds, whereas the means comes later. If a typical DevOps engineer is, as according to Golovnev, “the one in love with his/her job, insane as to automate everything within his/her reach, get to the core of any constrain that impedes fast product delivery or reduces effectiveness,” no wonder we, as consumers, and the whole industry need them. Together with the outstanding tools and technologies, a DevOps engineer is a soldier that, in 100% of cases, shoots to kill (any production constraint).

One issue, however, remains unclear. With all the variances as to what true DevOps means even within the same company, how will we reach a unified DevOps definition? Most definitely, some unanimity would be useful for the candidates that try to land the position but end up rejected due to the above-mentioned differences in DevOps visions among the companies.

On the other hand, lack of unambiguity is justified due to a wide range of technologies as well as different approaches used by different companies. So far, skillset variance will stay a key problem in regards to DevOps hiring. Although our main question–can we do without DevOps?–is out of the question.