Monitoring services

Once observability is enabled, go to the Monitoring tab at the service level in the Qovery console.
This tab gives you real-time and historical visibility into your applications.

Service Health

The Service Health Check provides an overview of running instances and recent events:

  • Qovery events (e.g. deployments, failed deployments) and Kubernetes events (e.g. autoscaler triggered, OOMKilled pod, health check failures).

Events

  • Log errors counter – automatically counts error-level logs. You can jump directly into logs at the exact point where errors occurred.
  • HTTP error rate (499 and 5xx errors) – aggregated over all requests, with the ability to expand and view the breakdown by endpoint and status code.
  • Network request duration (P99) – shows tail latency of your requests, expandable to include P90 and P50.

Resources

Resource usage is tracked per pod:

  • CPU usage – compared against your request and limit.
  • Memory usage – compared against your request and limit.

Resources

Network

Qovery provides network-level observability:

  • Request status by path and error code.
  • Request duration with percentiles (P50, P95, P99).
  • Request size statistics.

Network

If your application exposes a public port, metrics reflect ingress traffic.
If no public port is exposed, metrics reflect internal traffic within the cluster.

Controls

  • Live update: toggle continuous refresh of charts.
  • Custom time frame: select the time range of data to display.
Contents
Resources