shelve offloading an instance always results in an error being logged

[Matt Riedemann]

This is a recent regression:

https://review.openstack.org/#/c/493834/3/nova/compute/resource_tracker.py

http://logstash.openstack.org/#dashboard/file/logstash.json?query=message%3A%5C%22Failed%20to%20clean%20allocation%20of%20a%20shelve%20offloaded%5C%22%20AND%20tags%3A%5C%22screen-n-cpu.txt%5C%22&from=7d

http://logs.openstack.org/58/496858/1/check/gate-tempest-dsvm-neutron-scenario-multinode-lvm-multibackend-ubuntu-xenial-nv/eea4ffd/logs/screen-n-cpu.txt#_Aug_23_21_05_33_438007

Aug 23 21:05:33.438007 ubuntu-xenial-2-node-rax-iad-10587240 nova-compute[10028]: ERROR nova.compute.resource_tracker [None req-51d8a00f-35e1-489a-b070-edaad0b4c401 tempest-test_creds-1988617175 tempest-test_creds-1988617175] [instance: 1fb82ba9-6157-4c66-84f8-b97ed5a4fd84] Failed to clean allocation of a shelve offloaded instance

This is because the delete_allocation_for_instance method doesn't return anything, and "if not None" is always True.

[OpenStack Infra (hudson-openstack)]

Fix proposed to branch: master
Review: https://review.openstack.org/496930

[OpenStack Infra (hudson-openstack)]

Fix proposed to branch: stable/pike
Review: https://review.openstack.org/497237

[OpenStack Infra (hudson-openstack)]

Reviewed: https://review.openstack.org/496930
Committed: https://git.openstack.org/cgit/openstack/nova/commit/?id=e208cbb1f122b948af91a8d11c0f45dcec5bbc76
Submitter: Jenkins
Branch: master

commit e208cbb1f122b948af91a8d11c0f45dcec5bbc76
Author: Matt Riedemann <email address hidden>
Date: Wed Aug 23 17:27:05 2017 -0400

    How about not logging errors every time we shelve offload?

    The delete_allocation_for_instance method does not return
    anything and "if not None" is always True, so we're always
    logging an error when we shelve offload an instance, which
    is not the most fun thing in the world.

    This was probably caused by some confusion with how
    remove_provider_from_instance_allocation works which does
    return a boolean result.

    Change-Id: I30b716b0649f6d367ea747149da5c5f921ae19c5
    Closes-Bug: #1712689

[OpenStack Infra (hudson-openstack)]

Reviewed: https://review.openstack.org/497237
Committed: https://git.openstack.org/cgit/openstack/nova/commit/?id=df6ea0412f462815cf57560e96f33f4e2ff9d072
Submitter: Jenkins
Branch: stable/pike

commit df6ea0412f462815cf57560e96f33f4e2ff9d072
Author: Matt Riedemann <email address hidden>
Date: Wed Aug 23 17:27:05 2017 -0400

    How about not logging errors every time we shelve offload?

    The delete_allocation_for_instance method does not return
    anything and "if not None" is always True, so we're always
    logging an error when we shelve offload an instance, which
    is not the most fun thing in the world.

    This was probably caused by some confusion with how
    remove_provider_from_instance_allocation works which does
    return a boolean result.

    Change-Id: I30b716b0649f6d367ea747149da5c5f921ae19c5
    Closes-Bug: #1712689
    (cherry picked from commit e208cbb1f122b948af91a8d11c0f45dcec5bbc76)

[OpenStack Infra (hudson-openstack)]

This issue was fixed in the openstack/nova 16.0.0.0rc2 release candidate.

[OpenStack Infra (hudson-openstack)]

This issue was fixed in the openstack/nova 17.0.0.0b1 development milestone.