On-Call
Adjective · Noun · Slang & Abbreviations
Definitions
The practice of assigning an engineer to be reachable and ready to respond to production incidents for a set period, including nights and weekends. Teams usually run a rotation with a primary and a secondary responder, an escalation policy that pages the next person if an alert goes unacknowledged, and paging tools such as PagerDuty or Opsgenie. Good on-call depends on runbooks for known failure modes, alerts that fire on user-visible symptoms or service-level objective burn rather than on every internal metric, and blameless postmortems that turn incidents into fixes. The chief failure mode is alert fatigue: noisy, unactionable pages train responders to ignore them. Google's SRE practice caps operational work at half an engineer's time so that reliability improvements happen instead of endless firefighting. Whether on-call is compensated, and how, is a recurring point of contention across the industry.
In plain English: Being the person who has to answer their phone and fix problems if something breaks at night or on weekends.