Incident

An incident is any unplanned interruption to, or reduction in the quality of, a service. Examples include a full outage, a slow checkout, a failed deploy that breaks login, or an expired certificate. Incidents have a start time, a detection time, a resolution time and a severity.

Why it matters

Handling incidents consistently limits the damage. A defined process covers detection, triage, communication, mitigation and review, and keeps customers informed through a status page. Treating each event ad hoc leads to confusion about who is in charge and what was changed.

Example

A typical timeline:

09:12 monitor detects HTTP 503
09:15 on-call acknowledges, declares SEV2
09:30 cause found: bad config deployed at 09:05
09:38 config rolled back, service recovers
09:50 incident closed, postmortem scheduled

Common mistakes

• Debugging silently for an hour with no customer communication.
• Fixing the symptom and skipping the review, so the same incident recurs.
• Mixing up incident, problem and bug. An incident is the disruption, while a problem or bug is the underlying cause.
• Declaring too late. Opening an incident that turns out minor costs little.

Related

• Incident commander
• Severity levels
• Postmortem
• Status page

यह पेज सामान्य जानकारी है, जिसे हमारी संपादकीय टीम ने लिखा है और दस्तावेज़ीकरण बदलने पर अपडेट किया जाता है। सर्वर और प्रदाता अलग-अलग होते हैं, इसलिए बदलाव पहले सुरक्षित तरीके से जाँचें। देखें: संपादकीय नीति और अस्वीकरण.