Comment 2 for bug 1427718

Revision history for this message
OpenStack Infra (hudson-openstack) wrote : Fix merged to fuel-astute (master)

Reviewed: https://review.openstack.org/160891
Committed: https://git.openstack.org/cgit/stackforge/fuel-astute/commit/?id=47b738431ed87af314d0477625a18cdd258a0148
Submitter: Jenkins
Branch: master

commit 47b738431ed87af314d0477625a18cdd258a0148
Author: Vladimir Sharshov (warpc) <email address hidden>
Date: Tue Mar 3 19:57:56 2015 +0300

    Prevent unexpected erasing nodes during image based provision

    Classic provisioning flow:

    bootstrap -> erase -> reboot -> provisioning (node do not respond via mcollective) -> reboot -> target

    If system is backed online via mcollective and return bootstrap status, it will be erased.

    Image flow before this change:

    bootstrap -> erase -> provisioning (node still can respond via mcollective) -> reboot -> target

    Sometimes we can ask nodes about status in case of image flow
    before they will be rebooted. And send them to erase if they are answered 'bootstrap'. To prevent such situation, we add
    new systemtype 'image'.

    Now workflow will be looked:

    bootstrap -> erase -> image -> provisioning (node still can respond via mcollective) -> reboot -> target

    We marked nodes with 'image' systemtype as not provisioned, but do not try to erase them.

    This new type should not block any other operations with node:

    * delete;
    * stop provisioning;
    * reprovisioning.

    Change-Id: I7fcc32c604526575d7805386ecfe3d0d40cbd12e
    Closes-Bug: #1427718