On-Call Rotation
An on-call rotation is a schedule that assigns responsibility, usually on a rotating basis among team members, for responding to alerts and incidents outside normal working hours to keep a service running.
Definition
An on-call rotation is a schedule that assigns responsibility, usually on a rotating basis among team members, for responding to alerts and incidents outside normal working hours to keep a service running.
Overview
Most production services need someone reachable at any hour to respond when something breaks, since outages don't wait for business hours. An on-call rotation formalizes that responsibility: team members take turns — often weekly — carrying a pager or phone alert for a defined shift, with a clear escalation path to a secondary responder or manager if the primary doesn't acknowledge an alert in time. Well-run on-call programs try hard to keep the burden sustainable: reasonable shift lengths, compensation or time off for disruptive pages, and a strong emphasis on reducing the volume of low-value alerts so engineers aren't paged for things that don't actually need immediate human attention. Alert fatigue — being paged so often that engineers start ignoring or delaying responses — is one of the most common failure modes of a poorly tuned on-call program, and it's addressed by better observability and alerting design, not just by asking people to tolerate more noise. On-call engineers lean heavily on runbooks to resolve familiar problems quickly and on established incident management processes for anything more serious. Tools like PagerDuty and Opsgenie handle the mechanics of scheduling, escalation, and notification delivery. Reducing on-call toil — through better automation and more resilient system design — is considered a core, ongoing responsibility in site reliability engineering (SRE) rather than something teams simply have to live with.
Key Concepts
- Rotating shift schedule spreading after-hours response responsibility across a team
- Defined escalation paths to a secondary responder if the primary doesn't respond
- Integration with alerting and paging tools for automated notification delivery
- Explicit attention to alert quality to avoid fatigue from excessive or low-value pages
- Often paired with compensation, time off, or shift limits to keep the load sustainable
- Continuous feedback loop with runbooks and postmortems to reduce future page volume
Use Cases
Frequently Asked Questions
From the Blog
How to Design a Tool Schema an LLM Will Call Correctly
A tool schema is a prompt, not just an interface contract. This shows how to name tools for unambiguous selection, type parameters so wrong values are impossible, write descriptions that say when not to call, and design error messages the model can actually recover from.
Read More AI & TechnologyHow to design tool schemas an LLM agent can call reliably
Agent tool-calling failures are usually schema failures. Learn how naming, typing, enums and error messages make the model pick the right tool and arguments.
Read More AI & TechnologyBuilding Your First AI-Powered App with the Anthropic API
The fastest way to understand AI engineering is to build something real. This project- based guide walks you through building a writing assistant powered by Claude — from your first API call through streaming responses, a FastAPI backend, a simple frontend, and deployment.
Read More ProgrammingThe JavaScript Event Loop Explained Simply
The JavaScript event loop is the mechanism that lets single-threaded JavaScript handle async work by running queued callbacks whenever the call stack is empty.
Read More