🐛 [BUG]: Should the install_container_service playbook pass when the status is down?

[paddyroddy]

Bug Description

In https://github.com/UCL-MIRSG/UCLH-MPBE-SRR-XNAT, I have run install_xnat playbook, followed by install_container_service playbook. The playbook ends with

PLAY RECAP ************************************************************************************************************
xnat_cserv                 : ok=47   changed=3    unreachable=0    failed=0    skipped=7    rescued=0    ignored=0
xnat_web                   : ok=6    changed=0    unreachable=0    failed=0    skipped=0    rescued=0    ignored=0

I would interpret this as passing. However, when I go to the portal I can see

which clearly indicates that it did not work.

From looking at xnat-cserv /storage/xnat/data/xnat/home/logs I can see an error along the lines of

2024-01-25 12:04:09,874 [http-nio-8080-exec-9] ERROR org.nrg.containers.api.DockerControlApi - Failed to list images. jakarta.ws.rs.ProcessingException: java.net.UnknownHostException: xnat-cserv
org.mandas.docker.client.exceptions.DockerException: jakarta.ws.rs.ProcessingException: java.net.UnknownHostException: xnat-cserv

Error aside, it begs the question whether the playbook should have some kind of check to see it actually worked. By the time I got install_xnat to finish, the portal worked. So naively, I would expect the same here.

Steps to Reproduce

Acceptance Criteria

Related Work Items

Beta Give feedback

No tasks being tracked yet.

Tasks

Beta Give feedback

No tasks being tracked yet.

[p-j-smith]

closing as a duplicate of #70