Manager and worker nodes in Docker - Commands & Configuration

Introduction

When running multiple Docker hosts together, you need a way to control and organize them. Manager nodes control the cluster and assign tasks, while worker nodes run the tasks. This setup helps run apps reliably across many machines.

When you want to run a web app on several servers to handle more users without downtime

When you need to update your app without stopping all servers at once

When you want to balance work evenly across multiple computers

When you want to add or remove servers easily without breaking your app

When you want automatic recovery if one server fails

Commands

This command starts a new swarm cluster and makes this machine the manager node. The advertise address is the IP other nodes use to connect.

Terminal

docker swarm init --advertise-addr 192.168.1.100

Expected Output

Swarm initialized: current node (abcdef123456) is now a manager.



To add a worker to this swarm, run the following command:

    docker swarm join --token SWMTKN-1-0abcdef1234567890 192.168.1.100:2377

→

--advertise-addr - Sets the IP address that other nodes use to connect to this manager

Run this command on a worker node to join the swarm cluster controlled by the manager at 192.168.1.100.

Terminal

docker swarm join --token SWMTKN-1-0abcdef1234567890 192.168.1.100:2377

Expected Output

This node joined a swarm as a worker.

→

--token - The secret token that allows this node to join the swarm as a worker

Run this on the manager node to see all nodes in the swarm and their status.

Terminal

docker node ls

Expected Output

ID                            HOSTNAME            STATUS              AVAILABILITY        MANAGER STATUS      ENGINE VERSION

abcdef1234567890abcdef1234    manager-node        Ready               Active              Leader              20.10.12

1234567890abcdef1234567890    worker-node1        Ready               Active                                  20.10.12

Key Concept

Manager nodes control the swarm and assign tasks, while worker nodes only run the tasks assigned to them.

Common Mistakes

Trying to join a worker node without using the correct join token

The swarm will reject the node because the token is like a password to join the cluster.

Always use the exact join token provided by the manager node when running 'docker swarm join'.

Running 'docker node ls' on a worker node

Only manager nodes can list all nodes; workers do not have permission to see the cluster state.

Run 'docker node ls' only on a manager node.

Summary

Initialize a swarm on one machine to make it the manager node.

Join other machines as worker nodes using the join token from the manager.

Use 'docker node ls' on the manager to see all nodes and their status.