Open Stack

+1 :)

I'm also +1 on amending our FFE rules so that the PTL can get a FFE on one thing of their desire, regardless of anyone disagreeing, as long as they have two cores that are willing to review. As a small thank-you for being PTL! :D (I'm serious even though I just thought of this.)

--ruby

On 2018-01-23, 5:23 AM, "Dmitry Tantsur" <dtantsur at redhat.com> wrote:

    Hi all,
    
    I'm writing to request an FFE for the classic drivers deprecation work [1][2]. 
    This is a part of the driver composition reform [3] - the effort started in 
    Ocata to revamp bare metal drivers.
    
    The following changes are in scope of this FFE:
    1. Provide an automatic migration to hardware types as part of 'ironic-dbsync 
    online_data_migrations'
    2. Update the CI to use hardware types
    3. Issue a deprecation warning when loading classic drivers, and deprecate 
    enabled_drivers option.
    
    Finishing it in Queens will allow us to stick to our schedule (outlined in [1]) 
    to remove classic drivers in Rocky. Keeping two methods of loading drivers is a 
    maintenance burden. Even worse, two sets of mostly equivalent drivers confuse 
    users, and the confusion well increase as we introduce features (like rescue) 
    that are only available for nodes using the new-style drivers.
    
    The downside of this work is that it introduces a non-trivial data migration 
    close to the end of the cycle. Thus, it is designed [1][2] to not fail if the 
    migration cannot fully succeed due to environmental reasons.
    
    rloo and stendulker were so kind to agree to review this work during the feature 
    freeze window, if it gets an exception.
    
    Dmitry
    
    [1] 
    http://specs.openstack.org/openstack/ironic-specs/specs/not-implemented/classic-drivers-future.html
    [2] https://review.openstack.org/536298
    [3] 
    http://specs.openstack.org/openstack/ironic-specs/specs/7.0/driver-composition-reform.html
    
    __________________________________________________________________________
    OpenStack Development Mailing List (not for usage questions)
    Unsubscribe: OpenStack-dev-request at lists.openstack.org?subject:unsubscribe
    http://lists.openstack.org/cgi-bin/mailman/listinfo/openstack-dev