Incident RoomIncident 2 of 4
The disk that filled itself
TukTuk Rides, a three-wheeler booking app. Ride requests start failing at night. Nothing was deployed.
- Linux
- Docker
- AWS
- GitHub
About 14 minutes
PAGE: ride requests failing, 5xx above 25%
Drivers cannot accept rides. It is 22:40 and demand is at its peak. Two app servers are returning errors.
- Time since page
- 0 min
- Customers hurt
- 0
- Revenue lost
- LKR 0
- Status
- Outage
Your system
- Load balancerHealthy
- App server 1Down
- App server 2Down
- App server 3Healthy
- PostgreSQLHealthy
Evidence and timeline
0 stepsYou are on call and the page just went off. Do not touch anything yet. Open a dashboard or a log from the Investigate list on the right, and find out what is really happening. Every action costs minutes, and every minute hurts customers.
Look before you leap. Good engineers read the evidence first, then act.
Dashboards
Logs
Deploys
Terminal
Ask people
Costs are simulated minutes.

