What is Observability in Kyma?
Out of the box, Kyma provides tools to collect and ship telemetry data using the Telemetry Module. Of course, you'll want to view and analyze the data you're collecting. This is where observability tools come in.
Kyma collects telemetry data with the following in-cluster components:
Prometheus collects metrics from Pods. Metrics are the time-stamped data that provide information on the running jobs, workload, CPU consumption, memory usage, and more. All metrics relevant for observing the in-cluster Istio Service Mesh are collected separately.
The collected telemetry data are exposed so that you can view and analyze them with observability tools.
NOTE: Kyma's telemetry component supports providing your own output configuration for application logs and traces. With this, you can connect your own observability systems inside or outside the Kyma cluster with the Kyma backend.
You can use the following in-cluster components to observe your applications' telemetry data:
- Prometheus, a lightweight backend for metrics.
- Loki, a lightweight backend for metrics.
CAUTION: the Loki integration got deprecated and is planned to be removed.
- Jaeger, a tracing backend serving as the query mechanism to display information about traces.
- Grafana to provide a dashboard and a query editor to visualize metrics and logs collected from Prometheus and Loki.