Skip to content


Recover from Cluster crash

Start new cluster

docker swarm init --force-new-cluster --advertise-addr

Check other Nodes automatically join new cluster

docker node ls

If they don't, use the join command provided, ie:

docker swarm join --token SWMTKN-1-1korweqob1x2530nc34uex84j03pa1e3x7qg5z2pt4bvt2fo5i-0aoijiczvso7socyzlrucbu8y

List 'Down' nodes

docker node ls

Remove 'Down' nodes

docker node demote <node_id>
docker node rm <node_id>

Recover from Service crash

  • Run deploy stacks again
  • Update each stack/service through Portainer

Rebalance containers on Nodes

docker service update --force <stack>_<service>


docker service update --force portainer_portainer proxy_proxy portainer_agent