Comment 1 for bug 1999679

Reviewed: https://review.opendev.org/c/starlingx/config/+/867525
Committed: https://opendev.org/starlingx/config/commit/ca6bc490b7a7df646827e80395e4a5f0b71d6301
Submitter: "Zuul (22348)"
Branch: master

commit ca6bc490b7a7df646827e80395e4a5f0b71d6301
Author: Heitor Matsui <email address hidden>
Date: Tue Dec 13 12:34:39 2022 -0300

    Use device_path to determine if PV is found

    After upgrading a host we observed a rare disk enumeration
    issue, that would later duplicate PVs, since it will report
    a different device_node and thus not match condition [1].

    This occurs when the same persistent device name used between
    CentOS and Debian points to different kernel derived device
    nodes. This is a unique scenario not previously handled by the
    conductor logic and due to a much later version of systemd/udev
    used in Debian vs. Centos.

    This commit adds logic to fetch device_path earlier and then
    use it to determine if PV is found at the upgraded system,
    leading the PV to being updated instead of created again.

    [1] https://opendev.org/starlingx/config/src/commit/748afd7f5b7d3fc5e958f7173ff1a19c946c73b4/sysinv/sysinv/sysinv/sysinv/conductor/manager.py#L5052

    Test Plan
    PASS: fresh install/bootstrap/unlock
    PASS: host lock/unlock
    PASS: upgrade AIO-DX
    PASS: force the enumeration issue via database and observe that
          existing PV is updated instead of duplicated after agent
          reports back to conductor

    Closes-bug: 1999679

    Change-Id: I43ae44f088c84b45a7a23c46d1ffca4568673e39
    Signed-off-by: Heitor Matsui <email address hidden>