Skip to main content
On-call management ensures someone is always available to respond to critical alerts. Kodo helps you create schedules, rotate responsibilities, and escalate when needed.

​
Core Concepts

​
Schedules

A schedule defines who is on-call and when:
  • Rotation: Team members take turns being on-call
  • Shifts: Time periods (e.g., weekly, daily, custom)
  • Overrides: Temporary changes for vacations or swaps

​
Escalation Policies

When alerts aren’t acknowledged, escalate to the next person: 
Alert Fired
    ↓ (5 min)
Primary On-Call
    ↓ (10 min, not acknowledged)
Secondary On-Call
    ↓ (15 min, not acknowledged)
Engineering Manager

​
Creating an On-Call Schedule

  1. Go to Dashboard → On-Call → Create Schedule
  2. Name your schedule (e.g., “API Team On-Call”)
  3. Add team members
  4. Choose rotation type:
    • Weekly: Each person on-call for a week
    • Daily: Rotate daily
    • Custom: Define specific shifts
  5. Set handoff time (e.g., 9:00 AM Monday)
  6. Save schedule

​
Schedule Types

​
Weekly Rotation

Most common for small teams: 
Week 1: Alice
Week 2: Bob
Week 3: Charlie
Week 4: Alice (cycle repeats)

​
Follow-the-Sun

For global teams, hand off as business hours end: 
US Hours (9 AM - 5 PM EST): US Team
EU Hours (9 AM - 5 PM CET): EU Team
APAC Hours (9 AM - 5 PM JST): APAC Team

​
Layered Schedules

Combine schedules for 24/7 coverage: 
{
  "layers": [
    {
      "name": "Business Hours",
      "schedule_id": "sched_business",
      "hours": { "start": "09:00", "end": "18:00" }
    },
    {
      "name": "After Hours",
      "schedule_id": "sched_afterhours",
      "hours": { "start": "18:00", "end": "09:00" }
    }
  ]
}

​
Escalation Policies

Define what happens when alerts go unacknowledged: 
curl -X POST "https://kodostatus.com/api/oncall/policies" \
  -H "X-API-Key: your_api_key" \
  -H "Content-Type: application/json" \
  -d '{
    "name": "Critical Alert Escalation",
    "steps": [
      {
        "targets": [{ "type": "schedule", "id": "sched_api_team" }],
        "timeout_minutes": 5
      },
      {
        "targets": [{ "type": "user", "id": "user_eng_manager" }],
        "timeout_minutes": 10
      },
      {
        "targets": [{ "type": "channel", "id": "ch_pagerduty" }],
        "timeout_minutes": 15
      }
    ],
    "repeat_count": 2
  }'
This policy:
  1. Notifies whoever is on-call from the API team schedule
  2. Waits 5 minutes for acknowledgment
  3. If not acknowledged, notifies the engineering manager
  4. Waits 10 more minutes
  5. If still not acknowledged, triggers PagerDuty
  6. Repeats the entire escalation 2 more times
Escalation steps support multiple notify methods: web_push, voice_call, email, and sms. Add voice_call to automatically phone responders when alerts go unacknowledged. See Voice Alerts for setup and billing details.

​
Schedule Overrides

When someone needs to swap shifts or take time off: Create overrides in the Dashboard or via API:
  1. Go to Dashboard > On-Call > Schedules
  2. Click on a schedule and select Add Override
  3. Choose the user, time range, and reason

​
Who’s On-Call Right Now?

View current on-call in Dashboard → On-Call → Now

​
Notification Preferences

Team members can set their notification preferences:
  • Push: Browser push notifications (all plans)
  • Voice Call: AI-powered phone calls via Voice Alerts (Team+)
  • SMS: Text messages for all alerts
  • Email: Fallback for non-urgent
Configure notification preferences in Dashboard > On-Call > Preferences or via API: 
curl -X PATCH "https://kodostatus.com/api/oncall/preferences" \
  -H "X-API-Key: your_api_key" \
  -H "Content-Type: application/json" \
  -d '{
    "critical": ["phone", "sms"],
    "major": ["sms", "push"],
    "minor": ["email"]
  }'

​
Best Practices

Distribute on-call burden evenly. Consider time zones and personal situations.
Give people time to respond—5 minutes for critical, 15+ for major. Too short causes stress, too long delays response.
Ensure on-call engineers know how to respond to common alerts.
Analyze escalation patterns monthly. If alerts frequently escalate, investigate why.
Minimize non-critical alerts during nights and weekends.