Overview

• Opsgenie is an incident notification service that alerts responsible personnel via phone call, SMS, and email when an alert (incident) matching pre-defined conditions occurs. 

Purpose

• To ensure that the person in charge can clearly recognize the incident and easily assess the situation and take follow-up actions. 

Features

• Team-based operation
  • Teams are organized per system, and each team's Admin can configure routing rules (types and timing of alerts), escalation, and on-call schedules. 
  • Only alerts relevant to one's own team are displayed and handled. 
  • In addition to team-based settings, individual users can set personal alert rules. 
• Operation according to the on-call schedule 
• Alerts are sent to the person in charge based on the characteristics of the incident. 
• Supports over 200 integrations – Slack, Grafana, Webhook, API, etc. 
• Only users registered in Opsgenie can receive alerts. (Alerts can be configured for specified users.) 
  • For each system unit, the Owner account adds users with Admin or User roles. (https://docs.opsgenie.com/docs/user-roles)

How it works

1. Service incident occurs 
   1. Detected by the monitoring tool, which creates an alert in Opsgenie 
2. Alerts are sent to the on-call person in charge for each system (e.g., system Owner) – via phone call, SMS, email 
   1. Alerts are delivered through international calls, and phone calls use an English ARS. (If the description contains Korean, that part is skipped during the reading.) 
   2. There must be a designated alert recipient (on-call person or Admin) for each system (team). 
   3. Escalation rules must also be configured. 
   4. Depending on the severity, alerts can also be sent to the business team or customer contacts. 
3. An alert message is posted to the system-specific alert channel in Slack. 
4. When the person in charge acknowledges the incident (clicks the Acknowledge button, views the alert, assigns a responsible person, etc.), the alert is closed and escalation stops. 
   1. If not acknowledged, the alert is escalated to the next recipient according to the escalation rule.