Issues with nova-manage volume_attachment subcommand

[Amit Uniyal]

Downstream bug report from Red Hat Bugzilla against Train: https://bugzilla.redhat.com/show_bug.cgi?id=2161733

1 - fix handling of instance locking
    Add a context manager for locking and unlocking instance for volume attachement refresh cmd.

2 - disconnecting the volume from the correct host
    Verify if instance is attached to correct compute host before removing volume connection.

[Amit Uniyal (auniyal)]

https://review.opendev.org/c/openstack/nova/+/877446
https://review.opendev.org/c/openstack/nova/+/873648

[OpenStack Infra (hudson-openstack)]

Reviewed: https://review.opendev.org/c/openstack/nova/+/873648
Committed: https://opendev.org/openstack/nova/commit/211737d0ce23a8a9b1f3953f4816150a9a3271fa
Submitter: "Zuul (22348)"
Branch: master

commit 211737d0ce23a8a9b1f3953f4816150a9a3271fa
Author: Amit Uniyal <email address hidden>
Date: Mon Feb 13 21:06:09 2023 +0000

    Added context manager for instance lock

    Moved lock and unlock instance code to context manager.
    Updated _refresh volume attachment method to use instance
    lock context manager

    Now there will be a single request ID for the lock, refresh, and unlock
    actions. Earlier, the volume_attachment refresh operation used to have a
    unique req-id for each action.

    Related-Bug: #2012365
    Change-Id: I6588836c3484a26d67a5995710761f0f6b6a4c18

[OpenStack Infra (hudson-openstack)]

Reviewed: https://review.opendev.org/c/openstack/nova/+/877446
Committed: https://opendev.org/openstack/nova/commit/a8f81d5f08692c16d538af3197460d9426cc00a1
Submitter: "Zuul (22348)"
Branch: master

commit a8f81d5f08692c16d538af3197460d9426cc00a1
Author: Amit Uniyal <email address hidden>
Date: Wed Mar 15 05:07:28 2023 +0000

    Disconnecting volume from the compute host

    cmd nova-manage volume_attachment refresh vm-id vol-id connetor

    There were cases where the instance said to live in compute#1 but the
    connection_info in the BDM record was for compute#2, and when the script
    called `remote_volume_connection` then nova would call os-brick on
    compute#1 (the wrong node) and try to detach it.

    In some case os-brick would mistakenly think that the volume was
    attached (because the target and lun matched an existing volume on the
    host) and would try to disconnect, resulting in errors on the compute
    logs.

    - Added HostConflict exception
    - Fixes dedent in cmd/manange.py
    - Updates nova-mange doc

    Closes-Bug: #2012365
    Change-Id: I21109752ff1c56d3cefa58fcd36c68bf468e0a73

[OpenStack Infra (hudson-openstack)]

Related fix proposed to branch: stable/2023.2
Review: https://review.opendev.org/c/openstack/nova/+/911394

[OpenStack Infra (hudson-openstack)]

Fix proposed to branch: stable/2023.2
Review: https://review.opendev.org/c/openstack/nova/+/911396

[OpenStack Infra (hudson-openstack)]

Related fix proposed to branch: stable/2023.1
Review: https://review.opendev.org/c/openstack/nova/+/912511

[OpenStack Infra (hudson-openstack)]

Fix proposed to branch: stable/2023.1
Review: https://review.opendev.org/c/openstack/nova/+/912513

[OpenStack Infra (hudson-openstack)]

This issue was fixed in the openstack/nova 29.0.0.0rc1 release candidate.