rycus86/container-metrics repository overview

⁠Simple container metrics in Go

Collect Docker⁠ container and daemon statistics with Prometheus⁠.

⁠Motivation

I'm running a Home Lab⁠ on resource constrained devices, and wanted a simple metrics collector with low CPU, memory and disk requirements as an alternative to (the otherwise fantastic) Telegraf⁠ collector.

The end result is a small static binary, packaged as a Docker image⁠, weighing only 3 MB (compressed).

⁠Usage

The easiest way to get started is to use the Docker image available.

$ docker run -d --name container-metrics \
	-p 8080:8080 \
	-v /var/run/docker.sock:/var/run/docker.sock:ro \
	rycus86/container-metrics #[flags]

The application accepts the following flags:

• -p or -port: HTTP port to listen on (default: 8080)
• -i or -interval: Interval for reading metrics from the engine (default: 5s)
• -t or -timeout: Timeout for calling endpoints on the engine (default: 30s)
• -l or -labels: Labels to keep (comma separated, accepts regex)
• -d or -debug: Enable debug messages
• -v or -verbose: Enable verbose messages - assumes debug

You can also build the application with Go, currently tested with version 1.10, then simply run it on the host:

$ ./container-metrics -p 8080 -i 15s -l com.docker.compose.service,com.mycompany.custom

⁠Metrics collected

Currently, the following Gauge metrics are exported.

⁠Engine metrics

• cntm_engine_num_images: Number of images
• cntm_engine_num_containers: Number of containers
• cntm_engine_num_containers_running: Number of running containers
• cntm_engine_num_containers_stopped: Number of stopped containers
• cntm_engine_num_containers_paused: Number of paused containers

⁠Container CPU metrics

• cntm_cpu_usage_total_seconds: Total CPU usage
• cntm_cpu_usage_system_seconds: CPU usage in system mode
• cntm_cpu_usage_user_seconds: CPU usage in user mode
• cntm_cpu_usage_percent: Total CPU usage in percent

⁠Container memory metrics

• cntm_memory_total_bytes: Total memory available
• cntm_memory_usage_bytes: Memory usage
• cntm_memory_usage_percent: Memory usage in percent

⁠Container I/O metrics

• cntm_io_read_bytes: I/O bytes read
• cntm_io_write_bytes: I/O bytes written

⁠Container network metrics

• cntm_net_rx_bytes: Network receive bytes
• cntm_net_rx_packets: Network receive packets
• cntm_net_rx_dropped: Network receive packets dropped
• cntm_net_rx_errors: Network receive errors
• cntm_net_tx_bytes: Network transmit bytes
• cntm_net_tx_packets: Network transmit packets
• cntm_net_tx_dropped: Network transmit packets dropped
• cntm_net_tx_errors: Network transmit errors

⁠License

MIT