I have a playbook, the task consistently flags one user as ‘ok’ and one user as ‘changed’.
See this gist: https://gist.github.com/bdunbar/ec35fb02d4147ad1f5b5
Why is this? Clearly the user already exists, so it should be ‘ok’ and not ‘changed’.
I’m doing something wrong, and I’d like to know what that is.
Thanks!
~brian
How are the systems different?
I’m running the playbook against the same host, ‘sftp’ each time.
‘aws-stage-sftp’ is Ubuntu 12.04.4 LTS
inventory.yml
(snip)
aws-stage-sftp ansible_ssh_host=11.11.11.11
(snip)
[sftp]
aws-stage-sftp
Just a guess, but since both users are pointing at the same home dir, but have different UID/GIDs, is it possible that ansible is chowning ‘/home/sftp’ ? Although if that were the case, I’d expect both the tasks in subsequent runs to report as ‘changed’…
Bar the chowning, that makes sense. They both have the same homedir because they’re jailed ‘demi-root’ users.
I should probably did into the users module in the code and see what it’s actually doing in there.