Why Traditional Ops Roles Are Becoming Obsolete: Embrace DevOps and Automation

After reading this, I realized that companies are demanding more from operations staff, even expecting them to be all‑rounders.

Why is ordinary operations compared to a grasshopper after autumn? This refers to ordinary ops; if you are strong enough, you won’t be eliminated. Let yourself become DevOps; let yourself become DevOps; let yourself become DevOps.

Don’t stop learning, don’t rely on outdated skills, and don’t cling to old habits. The future of ops is truly frightening, so let’s look at the author’s experience.

Please spend three minutes reading carefully; you’ll surely gain something.

I’m an IT “geek” who has been following industry trends for almost eight years, working in IT operations. Companies now require ops engineers to know a development language, something unimaginable in 2007.

Back then, ops and development were completely separate; most ops staff only needed to configure common services, perform simple tuning, and write shell scripts for daily tasks such as monitoring and backup. That level of skill was enough to be considered a qualified ops engineer, and many still operate at that level today.

Those requirements were simple, and many fresh graduates could fill the role. Ops was relatively easy to get, though salaries were lower than developers and the department was often seen as peripheral, akin to a “network admin”. I disliked being called a network admin because it felt like fixing computers in an internet café.

Eight years later, I habitually write an article on the last day of each year. Previously I wrote in Beijing during New Year’s, but this time I’m in New York with a wonderful partner. I never imagined that, years ago, my English would be so poor and that I would end up here, just as I never imagined moving from ops to development.

I think my career follows the industry’s big trend: the wind is blowing. Many ops colleagues have already seen that ordinary ops are like grasshoppers after autumn—only a short hop before they disappear. IT automation will soon replace most unmotivated ops staff, and their “retirement” is just a matter of time.

I’ll tell you a real story that might move you.

I come from a small rural village in Shandong. When I was a child, the whole village often experienced power outages. An electrician would climb the pole to fix it, and if he couldn’t, a more professional worker from the county power bureau would be called. Power cuts were a daily occurrence, sometimes several times a month.

Each village had an electrician who received a few hundred yuan a month—a part‑time job because power cuts weren’t constant. The job required little skill, so many people wanted it, but there was only one position per village.

When I was in middle school, the village upgraded the electrical panels and replaced many lines. Power outages dropped from several times a month to once every few months. The village electrician suddenly had little work.

Village leaders thought paying a few hundred yuan each month for an idle worker wasn’t worth it, but they still needed the role. They decided several neighboring villages would share one electrician. When outages occurred, I saw strangers from other villages climbing poles because our electrician no longer worked for us—perhaps due to lack of skill or not giving gifts.

After the electrician was laid off, the village suffered several outages in a month. Investigation revealed someone had deliberately tied two wires together, causing short circuits. The leaders were angry, knew who was responsible, but could not name them publicly; they warned that if it happened again, the police would be involved. Since then, power usage returned to normal.

Eventually, villages stopped hiring full‑time electricians and the power bureau assigned a part‑time electrician to each town, covering dozens of villages. I later left for school and stopped following the issue.

Think about ourselves: are we the electricians? Ops staff are essentially the electricians of a company. Years ago, every company kept a few ops people to maintain a small on‑premise server room.

Now small companies no longer have server rooms or even buy servers; workloads run in the cloud, and a large pool of professional ops staff maintains those services, so many small firms no longer need any ops personnel. Consequently, ops engineers from small firms flow to large enterprises that run hundreds or thousands of servers and employ dozens or hundreds of ops staff. Do these large companies really need that many ops people?

Abroad, the practice is more advanced. I once read that a Facebook ops engineer managed 20,000 servers. I’m not sure the figure is accurate, but I’m confident that with good automation and proper management, such numbers are achievable.

When that day arrives, where will the surplus ops staff go? The answer is simple: follow the industry’s two major trends.

Resource centralization: IT operations will become a foundational service for the entire industry, like water and electricity. You don’t need to buy a generator to have power; one day most IT services will run on a few giant infrastructure providers, dramatically reducing overall IT operating costs.

High automation: I firmly believe that, given the right timing, most tasks currently performed by ops can be fully automated. Machines are more reliable and efficient than humans. Today’s “automation” is still semi‑automatic, requiring human supervision, much like an automatic‑transmission car still needs a driver. Eventually, fully driver‑less cars—and fully automated ops—will become the norm.

In short, I advise all ops colleagues, whether newcomers or veterans, to learn ops development early, replace colleagues’ jobs, and earn their salaries—or wait for them to replace yours, haha.

That’s my view of the industry’s future after many years in the field.