0

We are trying to run swarm cluster on Docker 1.12. We are using VMs (Kernel Version: 3.13.0-100-generic Operating System: Ubuntu 14.04.5 LTS). Swarm was created successfully and we have joined a worker to this cluster. enter image description here

Now have created a network docker network create --driver overlay --subnet 10.130.0.0/16 customnw. The network get created sucessfully. enter image description here

Than we created a service on the leader docker service create --name jboss11 --network customnw --replicas 3 -p 8080:8080 tutum/jboss

Service created

Containers are not getting started enter image description here

on observing the events logs we found that the containers are created and destroyed countinously

root@dockernetra:~# 2016-12-05T15:28:43.771650569+05:30 container destroy 3e72622f1bbf6a01d6dc34be22b6417e185c009819d34532d973d1a97441ed0d (com.docker.swarm.node.id=0tlikj98tb2bnfwl9tgw3ldbc, com.docker.swarm.service.id=b5bahzvf698ujor4dmiyltodt, com.docker.swarm.service.name=jboss12, com.docker.swarm.task=, com.docker.swarm.task.id=ef749838q9ya24uirpai1km6z, com.docker.swarm.task.name=jboss12.1, image=tutum/jboss:latest, name=jboss12.1.ef749838q9ya24uirpai1km6z) 2016-12-05T15:28:48.423961590+05:30 container destroy e6359659b674aa0df8fb7324aeedb434fd4b80e81471a80ae049e5b7d88f4b7c (com.docker.swarm.node.id=0tlikj98tb2bnfwl9tgw3ldbc, com.docker.swarm.service.id=b5bahzvf698ujor4dmiyltodt, com.docker.swarm.service.name=jboss12, com.docker.swarm.task=, com.docker.swarm.task.id=2isfwz3osgtn67e9p9wjhomk2, com.docker.swarm.task.name=jboss12.1, image=tutum/jboss:latest, name=jboss12.1.2isfwz3osgtn67e9p9wjhomk2) 2016-12-05T15:28:53.172836589+05:30 container destroy 7675fdfec3aac974ab63c3498a4fba59f4e9c1a200a4e4910296ed861b4b3e9f (com.docker.swarm.node.id=0tlikj98tb2bnfwl9tgw3ldbc, com.docker.swarm.service.id=b5bahzvf698ujor4dmiyltodt, com.docker.swarm.service.name=jboss12, com.docker.swarm.task=, com.docker.swarm.task.id=cm0i7mxrnj2awwgchxkxzwn43, com.docker.swarm.task.name=jboss12.1, image=tutum/jboss:latest, name=jboss12.1.cm0i7mxrnj2awwgchxkxzwn43) 2016-12-05T15:28:58.203511731+05:30 container destroy bcf832404745516923e033376c0f737ea955efa0c6ec6606f055a00ac8bb4b3f (com.docker.swarm.node.id=0tlikj98tb2bnfwl9tgw3ldbc, com.docker.swarm.service.id=b5bahzvf698ujor4dmiyltodt, com.docker.swarm.service.name=jboss12, com.docker.swarm.task=, com.docker.swarm.task.id=ddiwqcuh6qcyabgm8jr73nma5, com.docker.swarm.task.name=jboss12.1, image=tutum/jboss:latest, name=jboss12.1.ddiwqcuh6qcyabgm8jr73nma5)

Vaibhav Jain
  • 119
  • 3
  • 16
  • I am complete novice in docker and networking. In case of any other information is required please let me know through comments, i will attach that in the description – Vaibhav Jain Dec 05 '16 at 14:47
  • Looks like an issue with creating a directory (look at the `Error` column). Issue with write permissions, filesystem, etc...? Title is probably a bit misleading as it doesn't look like an error specific to overlay networking. – johnharris85 Dec 05 '16 at 15:04
  • If cluster is created without specifying the network than the cluster is formed. Only when we want that cluster to work with our defined n/w than it is not getting up. If there would be any permission constrain than cluster might not work at all. – Vaibhav Jain Dec 06 '16 at 07:06

1 Answers1

-1

They're definitely not failing because of the network. Whenever you see a container in a "rejected" state, it's normally because the config is valid, but it couldn't be implemented. Most common is overlay networks defined with an existing network. Creates an asynchronous routing scenario and Docker can't implement it.

It looks like it can't write to the host in /var/lib/docker (which is where aufs layers are stored)

Can you check the hosts and make sure they haven't run out of space, or that docker can create files in that directory?

Dockstar
  • 1,005
  • 11
  • 15