From Heroes to Rotas: How DevOps Teams Share the Hard Jobs

From Heroes to Rotas: How DevOps Teams Share the Hard Jobs

Spend enough time around DevOps learners and you notice a shift in what people ask about. A few years ago the questions were mostly tools: which pipeline, which container platform, which monitoring stack. Now a growing number are about people: how do we share the pager, how do we stop the same two engineers answering everything, how do we let a newcomer take a real incident without panic? What follows is a look at that change: why shared on-call and rotating hard jobs have moved from nice idea to ordinary expectation, and what it means if you are learning DevOps today. It is an observation from the shape of the conversation rather than a survey, so treat it as a field note, not a dataset.

The trend: from heroes to rotas

The old picture of operations had a lone expert at its centre, the person who knew where everything was buried and answered the phone at any hour. Teams celebrated that person, and often over-relied on them. The newer picture, visible in how teams describe themselves and what trainers are asked to cover, treats the hero as a symptom. If a service only survives because of one person’s memory, it is fragile. The healthier pattern is a rota in which responsibilities move around, supported by notes and mentoring.

Several forces push in that direction, and none of them requires a number to explain.

  • Teams own what they build. When developers are expected to support their own services, on-call stops being a separate department’s burden and becomes a shared duty.
  • Systems are more distributed. No single person can hold the whole picture, so knowledge has to be spread deliberately.
  • Burnout is better understood. Managers recognise that a permanent pager-holder is a retention risk.
  • Remote and flexible work. Teams spread across locations and schedules need explicit, written handovers rather than hallway knowledge.

What learners are missing

Most training is good at tools and thin on shifts. Students can build a pipeline and still never have been responsible for it at night. That gap is what the rotation fills in a real job. It is also why articles such as this one on what students still have to learn beyond the basics of DevOps matter: tools are only half the education. The other half is judgement, communication and the habit of leaving good notes.

A rota gives you all three in compressed form. You learn to decide with partial information, to explain a fault to someone who is worried, and to write down what you did so the next person can act on it. These are the qualities that separate a person who has completed a course from a person who can be trusted with a live service.

Problem-solving is a practised skill

Critical thinking in operations does not arrive from reading alone. It grows from repeated contact with real faults, and as this guide to sharpening your reasoning when DevOps problems get messy explains, the habit of forming and testing hypotheses is something you train. Rotation guarantees the repetition. If you only meet a fault once a year because someone else always answers, your problem-solving muscle stays weak.

What good practice looks like now

  1. Everyone in the pool. Senior engineers take turns too. A rota that exempts them teaches the wrong lesson.
  2. Shadow before solo. New people observe, then lead with a guide, then fly alone.
  3. Handover notes by habit. Short, dated, specific, and read before anything else.
  4. Rest after hard shifts. A rota that ignores recovery simply spreads exhaustion.
  5. Follow-up work shared as well. The write-up and the fix do not always go to the same two people.
  6. A regular question. What did we learn this cycle that we would not have learned otherwise?

What a day looks like on a healthy rota

Picture an ordinary week on a team that has absorbed the idea. On Monday the outgoing engineer reads her notes aloud in a short call and flags one flaky job. The new person has a guide on standby, who is also next in line. Tuesday is quiet, so the new person uses the lull to read old handover notes, discovering a pattern nobody had spotted. On Thursday an alert fires in the evening, the on-call engineer handles it calmly with a message to the guide, and the follow-up ticket is assigned to the person who held the shift rather than whoever is nearest. On Friday a short retrospective captures one new thing learned. Nothing in the week is dramatic, which is exactly the point.

Where the trend is still uneven

Not every organisation has caught up. Some small teams still run on one person’s goodwill. Some larger ones have a rota on paper, but a quiet culture in which the same people always volunteer, or in which swaps are made privately and unfairly. Others adopt a rota without training, which frightens newcomers and feeds a belief that on-call is a punishment. The trend is real, but practice lags behind intention, and the gaps are where learners can make a difference by asking good questions in interviews.

What this means if you are learning DevOps

  • Ask prospective employers how their on-call works, and who takes part.
  • Practise incident response in labs, and write short handover notes after each exercise.
  • Volunteer for supported shifts early in a role.
  • Keep a learning list from every shift, as raw material for interviews and reviews.

These habits will serve you whatever tools are fashionable next year, because they concern how you work with people and uncertainty, not which product you use.

The role of outside partners

The trend has also shaped what support providers offer. DevOpsSupport pairs consulting and round-the-clock support with training courses for startups and larger teams, a combination that reflects the same shift: teams want help running things and want their own people to grow in the process. The most useful partnerships leave a team better able to share its hard jobs, rather than more dependent on someone else.

What teams say they gain

Teams that have moved to shared rotas tend to describe the same benefits in conversation. The pager feels less personal, since it belongs to the team rather than to a hero. New people reach confidence sooner, because they practise with support. Documentation improves as a side effect, since everyone needs it. And the people who used to carry everything report being able to take a holiday without carrying the phone. None of these is a measured result; they are the recurring themes of the stories people tell.

A note on caution

Rotation is not a cure for everything. A rota will not fix a system that pages people constantly, a tool stack nobody understands or a culture that punishes mistakes. It will, however, make those problems visible to more people, which is often what is needed before they can be tackled. Treat the rota as a way of spreading both the load and the evidence.

One more trend deserves a mention: the rise of the blameless review, in which a team studies a fault without hunting for a culprit. It pairs naturally with rotation, because when everyone takes the pager, everyone understands how easily a good engineer can meet a bad night. Sympathy, it turns out, is a by-product of shared duty.

Where it is heading

Expect the conversation to keep moving toward people and practice. Expect job descriptions to mention shared ownership and good documentation. Expect learners who can say, honestly, that they have taken a shift, written a handover and helped a colleague through a first incident, to stand out. And expect the old hero model to look increasingly like a risk that careful teams quietly retire.

If you are building your own learning plan this season, add one item that has nothing to do with software: find a way to take a real turn at the hard jobs. It will teach you more about operations than another tool tutorial, and it will give your future colleagues a good reason to be glad you joined.