Open ironiq opened 3 years ago
From #116 via @ironiq
Seems i've found the problem: the unit is installed via template. This notifies the "restart $unit" handler. This handler will do the daemon_reload, set the state, but will not enable the unit. The next task is to enable and start the unit, but on CentOS7 i don't know why, but it cannot start even if the "enabled: yes" is there. When i add the "enabled: yes" to the handler, it will do its job.
Can you post your ansible --version
from your deploy host and also cat /etc/os-release
(or cat /etc/lsb-release
) on your CentOS instance? And if possible, please post the full output of ANSIBLE_STDOUT_CALLBACK=debug ansible-playbook -vvi YOUR_INVENTORY YOUR_PLAYBOOK.yml
. CI does test CentOS 7 endpoints, and the most recent build doesn't seem to have this issue.
First of all sorry for the 2 things in the pull request #116 : I'm quite new to git.
While running this role i got a deprecation warning at installing psycopg2/selinux on CentOS7: 'Invoking "pip" only once while using a loop via squash_actions is deprecated.'
Also there is a missing daemon_reload between installing Netbox' systemd unit files and starting them.