Example · Monitoring
Know first when a server slows down
A few services are watched all the time. If one slows down or stops, the engineer gets an alert on their phone with the likely cause from the logs and a suggested step. The engineer confirms, fixes it, and the recovery is logged.
All examplesSteps
- 01All normalFour services are watched, with response times on one screen.
- 02One slows downThe Orders API starts to lag and is flagged.
- 03Alert to phoneThe engineer gets an alert with the likely cause and a suggested step.
- 04Engineer confirmsThe engineer presses handle, so everyone knows it's taken.
- 05Recovered and loggedThe service is back to normal and the whole incident is logged.
Sample data
Service status
Website
Orders API
Database
Payments
Incident log
- 14:32 Orders API response over 2 seconds
- 14:32 Alert sent to Raka
- 14:35 Raka confirmed
- 14:41 Queue worker restarted by Raka
- 14:43 Response back to normal, 190 ms
MonitorTelegram
Alert14:32
Orders API is slow
Response time 2,400 ms since 14:32, usually around 180 ms.
From the logs: database connections are full (100 of 100).
Suggested: restart the queue worker, then check the connection limit.
✓ Raka is handling it · 14:35
The fix stays with the engineer
The system only alerts and suggests. Restarting services or changing settings is done by the engineer, and every step is logged.
Often hear about outages from customers first?
Tell us about the work your team still does by hand. We reply within 1-2 working days.