Is 24/7 On‑Call Ops Really Terrifying? Real Insights from Chinese Ops Professionals

Several Zhihu contributors discuss whether operations (Ops) truly require a terrifying 24‑hour on‑call schedule.

Microsoft's approach: Deploy three separate teams in different time zones (West 8 (Seattle), UTC (UK/Ireland), East 8 (Shanghai)). Each team works an eight‑hour shift, handing off to the next zone, achieving continuous coverage without any individual working round‑the‑clock.

Typical company experiences: In some firms, managers may message at 3 am asking about service status, and employees face harsh feedback if they don’t respond promptly. However, many companies implement structured on‑call rotations, allowing staff to rest after handling incidents.

Key practices highlighted across answers include:

Having primary‑backup pairs so one can cover for the other.

Rotating on‑call duties, often giving each person a week off per month.

Ensuring service level agreements (SLAs) and disaster‑recovery capabilities so that a few failed nodes don’t cause major outages.

Fostering a team culture that emphasizes planned work over constant firefighting.

Automation and mature open‑source tools are emphasized as ways to improve efficiency: high‑availability setups, load‑balancing, front‑end/back‑end separation, and API governance can resolve most technical issues without manual intervention.

When incidents do arise—such as urgent business requests or catastrophic alerts—teams respond quickly, but the total on‑call time per month often stays under three hours for many respondents.

Overall, while 24/7 availability is a requirement for many Ops roles, especially in larger organizations, it can be managed through multi‑timezone teams, clear rotation policies, automation, and robust monitoring, making the workload far less frightening than the myth suggests.