Swarm Setup 2 node Cluster

In this guide, you'll learn how to set up a Docker Swarm cluster with one manager and two workers, perform basic node operations, and verify cluster status. By the end, you'll be ready to add, promote, or drain nodes in your Swarm.

## Prerequisites

1. Three machines (physical or virtual), on-premise or cloud.  
2. Static IPs assigned:
   - **manager1**: 172.31.46.126  
   - **worker1**: 172.31.46.127  
   - **worker2**: 172.31.46.128  
3. Install [Docker Engine](https://docs.docker.com/engine/install/) on each host.  
4. Open firewall ports for Swarm communication:

| Port     | Protocol | Description                          |
|----------|----------|--------------------------------------|
| 2377     | TCP      | Cluster management                   |
| 7946     | TCP/UDP  | Node-to-node communication           |
| 4789     | UDP      | Overlay network (VXLAN) traffic      |

<Callout icon="triangle-alert" color="#FF6B6B">
Ensure your firewall rules allow traffic on these ports; otherwise, worker nodes cannot join the Swarm.
</Callout>

## Verify Docker and Swarm Status

After installation, run:

```bash
docker system info
```text

Look for:

```plain
Swarm: inactive
```text

If `Swarm: inactive`, the node is not yet part of a Swarm cluster.

## 1. Initialize the Swarm Manager

On **manager1**, execute:

```bash
docker swarm init --advertise-addr 172.31.46.126
```text

This:

- Configures the host as the Swarm manager  
- Prints the `docker swarm join` command with a token for workers  

Re-run `docker system info` to confirm:

```plain
Swarm: active
```text

## 2. Join Worker Nodes

Copy the join command from the manager’s output and run on each worker:

```bash
# On worker1
docker swarm join \
  --token SWMTKN-1-xxxxxxxxxxxxxxxxxxxxxxxx \
  172.31.46.126:2377

# On worker2
docker swarm join \
  --token SWMTKN-1-xxxxxxxxxxxxxxxxxxxxxxxx \
  172.31.46.126:2377
```text

<Callout icon="lightbulb" color="#1CB2FE">
If you lose the join token, regenerate it on the manager:

```bash
docker swarm join-token worker
```text
</Callout>

## 3. Verify Cluster Nodes

On the manager, list all nodes:

```bash
docker node ls
```text

Example output:

```plain
ID                            HOSTNAME   STATUS  AVAILABILITY  MANAGER STATUS  ENGINE VERSION
91uxgq6i78j1hu5v7moq7vgz *    manager1   Ready   Active        Leader          19.03.8
2lux7z6p96g6vtx0h6a2wo2r       worker1    Ready   Active        \<none>          19.03.8
w0qr6k2cee3ojawmflc26pvp3      worker2    Ready   Active        \<none>          19.03.8
```text

| Column            | Description                                                                                   |
|-------------------|-----------------------------------------------------------------------------------------------|
| **STATUS**        | Ready: node is healthy and participating                                                     |
| **AVAILABILITY**  | Active: scheduler assigns tasks<br>Pause: no new tasks, existing continue<br>Drain: tasks moved off |
| **MANAGER STATUS**| Leader: primary manager<br>Reachable: manager node able to take over<br>&lt;none&gt;: worker |

The `*` marks the node where the command was run.

## 4. Inspect a Node

Get detailed info on any node:

```bash
docker node inspect manager1 --pretty
```text

Sample output:

```plain
ID:             91uxgq6i78j1hu5v7moq7vgz
Hostname:       manager1
  State:        Ready
  Availability: Active
  Address:      172.31.46.126
Manager Status:
  Address:      172.31.46.126:2377
  Raft Status:  Reachable