Closed bpleines closed 6 years ago
@bpleines thanks for the submission.
@springdo @pcarney8 @oybed I'm keen for your feedback here, particularly using ansible.cfg
. I'm generally in favor of less command line arguments, but we have had a running debate here on whether that obscures too much from folks.
@sherl0cks I'm all for less command line args. Considering this always runs against localhost and uses whatever your oc
client is logged into, makes me even more in favor of this. If we leave it in there, it promotes confusion due to the nature of Ansible as a tool to SSH into remote boxes, in my opinion.
Personally I rarely run with an inventory in a location called inventory
. I also feel that using the ansible.cfg to drive the inventory location hides too much and I would not be in favor of this for other repos. However, considering that this is Labs CI/CD and it already "hides quite a bit" (think run.sh
) I suppose yet one more hidden feature may not be too bad. One ask is to get this added to the README, however - i.e.: outline how a user can point to a custom inventory if they don't use the inventory
location.
Merging - please see #213.
Having inventory specified in ansible.cfg means that playbook invocations don't have to run -i inventory --- (Issue: Add ansible.cfg #11) ***note: the lab documentation playbook commands could be edited as well if this is merged
Mainly is important for customers because this is how the Ansible Docs do it. Include_role and import_role are typically reserved for edge cases like looping over roles --- (Call roles directly instead of using include_role #12)