yahoo-eng-team team mailing list archive

[Mark Doffman <mjdoffma@xxxxxxxxxx>]

Belive that for now this is invalid. There is code that is superficially
similar between the 'detach' code in manager.py and the block_device
attach function, but there are subtle differences. The code in
manager.py calls roll_detach on failure, which I believe is
inappropriate for the block_device.py attach function. There isn't an
easy way to re-use this code without a much larger re-factor.

** Changed in: nova
       Status: Triaged => Invalid

-- 
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/1497343

Title:
  Need to consolidate duplicated volume detach code between compute
  manager and block_device

Status in OpenStack Compute (nova):
  Invalid

Bug description:
  In this change:

  https://review.openstack.org/#/c/186742/11/nova/virt/block_device.py

  It was pointed out that the change is adding volume detach code that
  is duplicated with what's also in the _shutdown_instance method in
  nova.compute.manager.

  We wanted to get that bug fix into liberty before rc1 but we should
  consolidate this duplicate volume detach code int the
  nova.virt.block_device module and then have the compute manager call
  that.

  This bug is just tracking the reminder to clean this up.

To manage notifications about this bug go to:
https://bugs.launchpad.net/nova/+bug/1497343/+subscriptions