How to monitor Docker Swarm cluster and service replicas (tasks) with Prometheus and Grafana?

General
1

Hi everyone,

I have a Docker Swarm cluster running multiple services, and I’d like to monitor both the cluster nodes and the individual service replicas (tasks) using Prometheus and visualize the metrics in Grafana.

Specifically, I want to track:

  • CPU, RAM, and disk usage for each Swarm node

  • CPU and memory usage for each Swarm service (including per-task/replica metrics)

I’ve already set up Prometheus and Grafana, but I’m not sure what’s the best approach to collect metrics from:

  1. The Swarm manager and worker nodes

  2. Each container/task belonging to a service

2

You probably need to run Node exporter and cAdvisor to provide the metrics for Prometheus.

3

is there dashboard to monitor CPU and memory usage for each Swarm service (including per-task/replica metrics)

4

Personally I haven’t tested it. There are also metrics directly from Docker (doc)