Kubernetes observability stack¶
The Kubernetes observability stack refers to the integrated ecosystem of tools used to gain visibility into the internal state of Kubernetes clusters, typically encompassing Metrics, logging, and Tracing.
Core Components¶
Metrics and Monitoring¶
Prometheus is the standard tool for monitoring and time-series data collection within Kubernetes^[400-devops-06-kubernetes-k8s.md]. It typically operates with specific labels, such as release: prometheus, to organize metrics^[400-devops-06-kubernetes-k8s.md].
Logging¶
The EFK stack (commonly referring to Elasticsearch, Fluentd, and Kibana, though Fluentd is often substituted or complemented by Fluent-Bit in this context) is used for centralized log management^[400-devops-06-kubernetes-k8s.md].
Visualization¶
Kiali is utilized to visualize the Kubernetes topology and integrate Service mesh observability^[400-devops-06-kubernetes-k8s.md]. It allows users to view the structure of the mesh and the relationships between services, often relying on labels like app: xxxxx to identify and group applications^[400-devops-06-kubernetes-k8s.md].
Service mesh¶
While not strictly an observability tool itself, the Istio Service mesh provides the underlying infrastructure (often via sidecar proxies) that generates the telemetry data required for advanced observability^[400-devops-06-kubernetes-k8s.md].
Related Concepts¶
Sources¶
- 400-devops-06-kubernetes-k8s.md