cAdvisor — Real-Time Container Resource Monitoring by Google

What it is

cAdvisor (Container Advisor) is an open-source container monitoring tool maintained by Google. It collects, aggregates, and exports resource usage and performance data from running containers. cAdvisor runs as a single container and auto-discovers all other containers on the host, providing per-container CPU, memory, network, and filesystem metrics.

The tool targets DevOps engineers, SREs, and platform teams running Docker or Kubernetes workloads who need lightweight, always-on container metrics without deploying a full monitoring stack.

How it saves time or tokens

Without cAdvisor, gathering per-container metrics requires parsing docker stats output, writing custom collection scripts, or deploying heavyweight agents. cAdvisor provides a single Docker container that auto-discovers all running containers and exposes metrics in Prometheus format. No configuration files, no agent installation on the host, no custom scripts.

How to use

1. Run cAdvisor as a Docker container:

docker run -d --name=cadvisor \
  -p 8080:8080 \
  --volume=/:/rootfs:ro \
  --volume=/var/run:/var/run:ro \
  --volume=/sys:/sys:ro \
  --volume=/var/lib/docker/:/var/lib/docker:ro \
  gcr.io/cadvisor/cadvisor:latest

1. Open http://localhost:8080 to view the built-in web UI with container metrics.

1. Point Prometheus to http://localhost:8080/metrics to scrape container metrics for dashboards.

Example

# prometheus.yml - scrape cAdvisor metrics
scrape_configs:
  - job_name: 'cadvisor'
    scrape_interval: 15s
    static_configs:
      - targets: ['localhost:8080']

With this Prometheus config, container metrics flow into Grafana dashboards for visualization and alerting.

Related on TokRepo

• Monitoring Tools -- Observability and monitoring tools for infrastructure
• DevOps Tools -- Container orchestration and deployment tools

Common pitfalls

• cAdvisor's built-in web UI is useful for debugging but not suitable for long-term monitoring. Pair it with Prometheus and Grafana for production dashboards.
• Volume mounts must be read-only and match the host filesystem layout. Missing mounts cause incomplete metrics.
• On Kubernetes, cAdvisor is embedded in the kubelet. Running a standalone cAdvisor alongside kubelet can cause duplicate metrics collection.