yahoo-eng-team team mailing list archive

[OpenStack Infra <1784387@xxxxxxxxxxxxxxxxxx>]

Reviewed:  https://review.openstack.org/588931
Committed: https://git.openstack.org/cgit/openstack/nova/commit/?id=330c95090881a21eddf9f1483762e453330d6e45
Submitter: Zuul
Branch:    master

commit 330c95090881a21eddf9f1483762e453330d6e45
Author: Takashi NATSUME <natsume.takashi@xxxxxxxxxxxxx>
Date:   Sun Aug 5 02:39:47 2018 +0900

    api-ref: Add descriptions for rebuild
    
    Add descriptions about specifying a new image and
    asynchronous postconditions in the rebuild operation
    in the compute API reference.
    
    Change-Id: I0fd136d07dffc2be845b4b9330fae98c7115789b
    Closes-Bug: #1784387


** Changed in: nova
       Status: In Progress => Fix Released

-- 
You received this bug notification because you are a member of Yahoo!
Engineering Team, which is subscribed to OpenStack Compute (nova).
https://bugs.launchpad.net/bugs/1784387

Title:
  rebuild API reference should discuss implications of specifying a new
  image

Status in OpenStack Compute (nova):
  Fix Released

Bug description:
  The imageRef parameter description does not describe what happens
  during rebuild if a new image is provided, nor is there any
  asynchronous post-condition section for what the user should expect:

  https://developer.openstack.org/api-ref/compute/#rebuild-server-
  rebuild-action

  There are a few implications of specifying a new imageRef:

  1. You cannot specify a new image for a volume-backed server, it will result in an error.
  2. For non-volume-backed servers, specifying a new image will result in validating that the image is acceptable for the current compute host on which the server is running. If the new image is not valid, the server will go into "ERROR" status and can be rebuilt with a different image.

  The description of the API does not really mention what it does, i.e.
  destroys and re-creates the server including the root disk (if not on
  shared storage - which the user won't know beforehand) while
  preserving attached volumes and ports.

  As for asynchronous post-conditions, if the server was in status
  SHUTOFF before the rebuild, it will be stopped and in status SHUTOFF
  after the rebuild, otherwise it will be ACTIVE if the rebuild was
  successful or ERROR if the rebuild failed.

To manage notifications about this bug go to:
https://bugs.launchpad.net/nova/+bug/1784387/+subscriptions