The words Vibe OnCall uses, in the sense the product uses them. Korean console labels are shown where they differ from a literal translation.
| Term | Meaning |
|---|---|
| Alert (얼럿) | A raw signal from a monitoring source — a Datadog monitor, a Grafana rule, an email, a webhook call. Alerts have a severity, a source, metadata, and a dedupKey. See Alerts. |
| Incident (인시던트) | The unit of response. Groups the alerts, people, war room, timeline, and updates for one problem. States: Open → Acked → Blocked → Observation → Resolved → Archived. See Incident Lifecycle. |
| Notification (알림) | Any message sent to a person — Slack or Teams DM, email, SMS, push, phone call. Not the same as an alert. |
| Page / paging | Notifying a responder urgently through their configured channels until they acknowledge. |
| Acknowledge (Ack) | A responder confirms they have the alert or incident. Stops further escalation for it. |
| Severity | SEV0 (highest) to SEV5, plus SIGNAL (lowest, informational). One scale for alerts and incidents. |
| Priority | The three notification bands used in personal preferences and policies: High = SEV0–SEV1, Medium = SEV2–SEV3, Low = SEV4–SEV5 and SIGNAL. |
| Source | The webhook or inbound email address through which a monitoring tool sends alerts in. First stage of Routing. |
| Attribute | An optional Routing rule that enriches matching alerts before escalation: set a severity, add metadata, or copy fields from a Connection. |
| Notification rule | The Routing stage that decides which alerts are escalated and to whom (users, schedules, escalation policies), plus an ELSE fallback. |
| Routing (라우팅) | The single Settings screen — Sources → Attributes → Notification rule — that replaced the older Alert Policies and Routing Rules pages. |
| dedupKey | A stable string that identifies one logical alert stream; repeated alerts with the same key merge into the active alert. See Alert Deduplication. |
| Alert group (얼럿 그룹화) | Related alerts collected during a fixed time window and evaluated once, creating at most one incident. See Alert Grouping. |
| Alert burst | Several alerts admitted by the same automation within a short window. With burst grouping on, the assistant investigates them in one pass under the first alert that crossed the threshold. Distinct from an alert group, which merges alerts into one incident; a burst only limits how often the assistant runs. See Run Assistant Agent. |
| Mute | Pausing repeat paging for one alert for a preset window (15 min to 24 h). Does not resolve or acknowledge it. |
| Maintenance window (유지보수 기간) | A scheduled period during which matching alerts do not page. |
| Heartbeat (하트비트) | An endpoint a job pings on a schedule; an alert fires when the pings stop. |
| Schedule / rotation / override | A schedule defines who is on call; a rotation hands shifts over automatically; an override covers a shift temporarily. |
| Escalation policy (에스컬레이션 정책) | Ordered tiers of who to try next, with wait times. Pages only when targeted by a notification rule or paged manually. |
| Notification policy vs Notification preference | Policy: the minimum channels and delays an admin sets for a team or organization. Preference: a person's own channel choices per priority. |
| Scope (범위) | The selector at the top of the Settings sidebar: Global (organization-wide, Org admins) or a specific team. Routing, grouping, and other settings apply at the selected scope. See ‣. |
| War room | The Slack or Teams channel created for an incident, where responders work and where Vibe OnCall keeps the live summary. |
| Postmortem | The written review after an incident: timeline, root cause, actions. Vibe OnCall can draft it from the incident record. |
| Service (서비스) | An entry in the built-in service catalog: owners, responders, stakeholders, and a free-text description the AI reads during triage. |
| Responder / Stakeholder | Responder: pulled into the war room to fix. Stakeholder: kept informed with updates. Both can be set per service. |
| Automation action (자동화 액션) | A trigger + conditions + actions rule that runs when something happens (alert created, incident created, war room created…). |
| Assistant agent | An AI agent in Agent Studio — the built-in RCA agent or a custom one — that an automation can run against alerts. |
| Connection vs Integration | Connection: an HTTP endpoint used to enrich alerts or to be called from an automation. Integration: a connected product — Slack, Teams, Datadog, Grafana, SSO. |
| Tabletop exercise | A rehearsal incident driven by Vibe OnCall from a runbook, used for training and testing response. |
| Org admin / Member / Read-only · Team admin / Member | The organization-level and team-level roles. See Roles and permissions. |
Last reviewed: 2026-09-30 · Still stuck? Email [email protected]