kubeletcAdvisor

KubeletcAdvisor refers to the integration of the cAdvisor component into the Kubernetes kubelet. It provides container-level resource usage statistics for the nodes where Kubernetes runs, collecting data such as CPU and memory usage, disk I/O, network traffic, and filesystem utilization for each container and pod on a node. The information gathered by this integration is used by the kubelet itself and by higher-level components that rely on node-level and per-container metrics.

The data surface produced by kubeletcAdvisor is designed to support the Kubernetes monitoring and scheduling stack. Historically, cadvisor within the kubelet exposed a set of endpoints that could be scraped to obtain per-container statistics, which in turn fed into the Metrics Server and other monitoring tools. Metrics such as CPU usage, memory usage and limits, network I/O, and disk I/O are typical outputs, organized at the container and pod level and tied to the node on which the workloads are running. These metrics help inform scheduling decisions, autoscaling, and alerting, as well as general visibility into resource consumption on each node.

Architecture and operation are tied to the kubelet and the container runtime, with data gathered from cgroups and the running containers. Over time, Kubernetes has evolved to rely on the kubelet’s summary API and external monitoring systems more than on the traditional in-tree cadvisor endpoints. Modern monitoring often uses the Metrics Server or Prometheus-based approaches for collecting and aggregating metrics, while kubeletcAdvisor remains a historical and architectural element of how container metrics were traditionally exposed by the kubelet.

See also: cAdvisor, kubelet, Metrics Server, Kubernetes monitoring, node and pod metrics.