I'm trying to deploy a Docker service in a swarm, but always ending up with a running container on my local host (the one I use as the docker roaming manager) and no service
Here is my setup:
I have 3 node Docker (v.1.12.1) swarm, which includes one host working as a manager, and two work nodes, all of which work in CentOS 7. On the manager node (localhost) launched Ansible (v. 2.1.1.0) playbook, and the swarm is already configured and working
Swarm: active
NodeID: d9h5xa832ax7wzeq8q44fjld3
Is Manager: true
ClusterID: 9cztoin3gy2ntbwehsmrkjuxi
Managers: 1
Nodes: 3
Orchestration:
Task History Retention Limit: 5
Raft:
Snapshot Interval: 10000
Heartbeat Tick: 1
Election Tick: 3
Dispatcher:
Heartbeat Period: 5 seconds
CA Configuration:
Expiry Duration: 3 months
Node Address: 10.25.190.209
Starting with a tutorial having this code
- hosts: localhost
name: Run JMeter test
vars_files:
- user.config.yml
vars:
execute_tpcds_test : "{{ run_tpcds_test }}"
roles:
- { role: run_jmeter, when: execute_tpcds_test is defined and execute_tpcds_test ==1 }
What causes this role:
- name: Deploy tpcds_tpg service to swarm
docker_service:
project_name: tpcds-tpg
definition:
version: '2'
services:
run_tests:
image: 'pbench/tpcds_tpg'
volumes:
- /opt/pbench/run_output/
command: ./run_jmeter.sh "{{jmeter_output_dir}}"
register: output
- debug: var=output
When I start ansible-playbook ./site.yml, I end up with a running container. Execution docker ps -adisplays
[pdo@sdl02133 tpcds-tpg]$ docker ps -a
CONTAINER ID IMAGE COMMAND CREATED STATUS PORTS NAMES
fef245b41365 pbench/tpcds_tpg "./run_jmeter.sh /opt" 21 seconds ago Up 20 seconds tpcdstpg_run_tests_1
docker service ls , , docker_service ,
, , , , Ansible docker_service, . , , , .., , . , , !