[PR #6534/a9fd9f89 backport][stable-6] added handling of zypper exitcode 102: ZYPPER_EXIT_INF_REBOOT_NEEDED

[patchback]

This is a backport of PR #6534 as merged into main (a9fd9f8).

added handling of zypper exitcode 102: ZYPPER_EXIT_INF_REBOOT_NEEDED - Returned after a successful installation of a patch which requires reboot of computer.

SUMMARY

This code was not treated at all by the module and therefore considered a faillure and the changed status was never set.

After my fix the exitcode 102 will be treated exactly like 0 by the module internally, and the changed status will be reported correctly.

However, since I preserve the rc 102 in the retvals to allow the playbook to react to the requested reboot, the task must still include a "register" and a "failed_when" property to pass as OK in this event.

Example:

  tasks:
    - name: Apply patches on OpenSUSE/SUSE Linux
      zypper:
        name: '*'
        state: latest
        type: patch
        extra_args_precommand: '--non-interactive-include-reboot-patches'
      register: zypper_cmd
      failed_when: zypper_cmd.rc not in [0, 102]

    - name: Reboot the SUSE/OpenSUSE box if kernel updated
      reboot:
        msg: "Reboot initiated by Ansible for kernel updates"
        connect_timeout: 5
        reboot_timeout: 300
        pre_reboot_delay: 10
        post_reboot_delay: 30
        test_command: uptime
      when: zypper_cmd.rc == 102

ISSUE TYPE

• Bugfix Pull Request

COMPONENT NAME

zypper

ADDITIONAL INFORMATION

When used to do updates of the type "patch" the zypper module failed in case the updates were successfull, but resulted in the system needing a reboot. In that case the changed status was not correctly set. After my fix the task will still be treated a faillure by ansible unless you treat the rc 102 in a failed_when statement as shown above, but the changed state will be correctly set.

[ansibullbot]

cc @AnderEnder @alxgu @andytom @commel @evrardjp @lrupp @sealor @toabctl
click here for bot help