​
cAdvisor (Container Advisor)

Port: 8080 (internal) Developed by: Google cAdvisor runs alongside the Docker daemon and automatically discovers every running container, collecting real-time resource usage metrics. These metrics are scraped by Prometheus and displayed in Grafana’s Docker Containers dashboard.

​
What It Monitors Per Container

MetricDescription
CPUCPU usage in cores and percentage
MemoryRAM usage, cache, working set
Network I/OBytes sent and received per container
Disk I/OBytes read and written to disk
FilesystemContainer filesystem usage

​
Why cAdvisor + Node Exporter Together?

ToolScope
Node ExporterThe whole server (host-level)
cAdvisorEach individual container
Together they give complete visibility: you can see if the server is stressed and which specific container is causing it.

​
Use Cases

Identify a memory-leaking container: 
sort_desc(container_memory_usage_bytes{name!=""})
Find the most CPU-hungry container: 
sort_desc(rate(container_cpu_usage_seconds_total{name!=""}[5m]))
Check network traffic per container: 
rate(container_network_receive_bytes_total{name!=""}[5m])

​
Grafana Dashboard

The Docker Containers dashboard shows:
  • Top 10 containers by CPU
  • Top 10 containers by memory
  • Network traffic per container
  • All containers on a single timeline view
This makes it easy to spot which service is consuming the most resources at any given time.