Discover Version Number Taks fails on RHEL

[Yojimbo108]

I get the below error running this on CentOS 7

TASK [neel.rundeck : discover version number] **********************************
fatal: [rundeck.aem.bt.com]: FAILED! => {"changed": true, "cmd": "dpkg -s rundeck | grep -i version", "delta": "0:00:00.002894", "end": "2017-03-24 11:17:55.840221", "failed": true, "rc": 1, "start": "2017-03-24 11:17:55.837327", "stderr": "/bin/sh: dpkg: command not found", "stdout": "", "stdout_lines": [], "warnings": []}

I think this is down to the below section in main.yml

# install rundeck
- include: install_debian.yml
  when: ansible_os_family == 'Debian'

- include: install_redhat.yml
  when: ansible_os_family == 'RedHat'

- name: discover version number
  shell: dpkg -s rundeck | grep -i version
  register: rundeck_targeting_version

Looks like you need to push that discover version number task up into install_debian.yml & install_redhat.yml and use something like rpm -qi rundeck on Redhat instead of dpkg.

Thanks

[mat-green]

Hello,

Thank you for this and I will do as you suggest.

M

[Yojimbo108]

Cool, I'm not sure if you want to use rpm -qi as it throws an ansible warning:

TASK [neel.rundeck : Redhat | discover version number] *************************
ok: [rundeck.aem.bt.com]
 [WARNING]: Consider using yum, dnf or zypper module rather than running rpm

so I'm not sure if there's something better to use!

Thanks

[mat-green]

I am considering this fixed.

[Russell-IO]

@mat-green I was having this bug too with tag 17.03.16 (just switched the rundeck base image from Ubuntu 16.04 => Centos 7.3) however, after checking out 17.04.17 it works as expected

[mat-green]

Yes, this was fixed in 17.04.17.