Metrics and Monitoring¶
Cloud-native applications are complex and require lots of network interactions. Metrics provide visibility into these interactions, which enables applications teams to take swift action to address any issues that may arise.
All instances registered in Traefik Pilot can benefit from metrics, making it possible to diagnose and record changes in the usage of the instances or services behind Traefik through graphs spanning a day, a week, or a month.
Visualizations of the current data can be accessed by selecting the Metrics tab in the Traefik Pilot dashboard.
Available metrics include:
- Requests per second
- Total requests
- Request duration
- Open connections
For requests per second, both a maximum and an average value are graphed. These values are computed from the total requests.
When you first access the metrics view, you will see metrics at the user level by default. You can also choose to see metrics at the instance level or the proxy level. If you remove the filter, you will see the metrics at the user level again.
Metrics are sent every 5 minutes by the instances, but the charts display:
- One point per hour for the Last day
- One point every 6 hours for the Last week
- One point every day for the Last month
If a Traefik proxy fails to send a heartbeat signal to Traefik Pilot for a full 10-minute interval, Pilot can issue an alert that the proxy has lost access to the internet. See the Alerts section for more information.