Skip to main content
In this walkthrough, you’ll learn how to enable Docker Swarm’s Auto-Lock feature to encrypt Raft logs and TLS keys on disk. With Auto-Lock enabled, any manager restarting or rejoining the cluster must provide the unlock key—adding a robust layer of security.

​
1. Enable Auto-Lock on Your Swarm

For an existing Swarm cluster, run: 
docker swarm update --autolock=true
This outputs a one-time unlock key, for example: 
SWMKEY-1-izfTZG1yXBjIOY3VBkIHFDI+WcnpqeJKYV6daZW3o
To enable Auto-Lock during cluster creation, use:
docker swarm init --autolock=true

​
2. Store the Unlock Key Securely

Save the key in a safe location. For demo purposes we’ll use /tmp/swarm-unlock.key. In production, consider a secrets manager or vault. 
echo "SWMKEY-1-izfTZG1yXBjIOY3VBkIHFDI+WcnpqeJKYV6daZW3o" > /tmp/swarm-unlock.key
chmod 600 /tmp/swarm-unlock.key
Losing this key means you cannot unlock your Swarm managers. Always back it up securely.

​
3. Quick Reference: Swarm Auto-Lock Commands

CommandDescription
docker swarm init --autolock=trueInitialize a new Swarm with Auto-Lock enabled
docker swarm update --autolock=trueTurn on Auto-Lock for an existing Swarm
docker swarm unlockUnlock a manager node after restart or rejoin

​
4. Verify Cluster Health on Manager 1

Even with Auto-Lock active, manager1 can query node status without unlocking: 
[root@manager1 ~]# docker node ls
ID                        HOSTNAME       STATUS  AVAILABILITY  MANAGER STATUS  ENGINE VERSION
kvbht486wmj881wp5vqxp53 * manager1       Ready   Active        Leader          19.03.8
u8imabedhzsu4cawtoz6jh32   manager3       Ready   Active        Reachable       19.03.8
s2ymqdbtfal661imydx31rlno  manager2       Ready   Active        Reachable       19.03.8
38oehhk79ss5rk2coejcavha   worker1        Ready   Active                        19.03.8
k4gc50oc0n8k6jm3f6bm2bph   worker3        Ready   Active                        19.03.8
1pqddmh2fcoy79vq9najr841d  worker2        Ready   Active                        19.03.8

​
5. Test Auto-Lock on Manager 2

  1. Restart Docker on manager2: 
    [root@manager2 ~]# systemctl stop docker
[root@manager2 ~]# systemctl start docker
  2. Attempt a Swarm command (should fail): 
    [root@manager2 ~]# docker node ls
Error response from daemon: Swarm is encrypted and needs to be unlocked...
  3. Unlock the Swarm with your saved key: 
    [root@manager2 ~]# docker swarm unlock
Please enter unlock key: [paste contents of /tmp/swarm-unlock.key]
  4. Confirm the node list again: 
    [root@manager2 ~]# docker node ls
ID                        HOSTNAME       STATUS  AVAILABILITY  MANAGER STATUS  ENGINE VERSION
xbvhtg486wmj881wp5vkqx53 * manager1       Ready   Active        Leader          19.03.8
u8imabedhzsu4cawtoz6jh32   manager3       Ready   Active        Reachable       19.03.8
s2ymqdbtfal661imydx31rlno * manager2       Ready   Active        Reachable       19.03.8
38oehhth79bsfs7kco2jcvah   worker1        Ready   Active                        19.03.8
k4gcc5ooc0nm8xgl36fmb2pd   worker3        Ready   Active                        19.03.8
1pqddmhd2f0y7vq9najr841d   worker2        Ready   Active                        19.03.8
Congratulations! Manager 2 has rejoined securely with Auto-Lock enabled.

Watch Video