Flawed volume delete behavior in teardown script

[Sameer Adhikari]

Teardown script has a serious flaw in the task "Remove all logical volumes".
It ignore the scenario where the admin may have created another volume outside of AIO, and AIO deletes that too.
I stumbled on this because we create a snapshot volume that we use to revert the system back to know good state.
This step is easy to create.
1. Create a snapshot volume (for example, lvcreate --size 50G -s -n root_lv /dev/lxc/root)
2. Run lvs (you will see you root_lv volume)
3. Run AIO
4. From /opt/openstack-ansible run scripts/teardown.sh
5. Run lvs (You will see that root_lv has disappeared, and so has your ability to revert the system to an earlier know good image)

Why doesn't teardown only delete the volumes created to host the containers?

[Byron McCollum (byron-mccollum)]

I think the LXC volume group was intended for exclusive use by the LXC containers managed by OpenStack-Ansible. If you need a LV for other purposes, as you have suggested, create a separate VG.

[Sameer Adhikari (tri2sing)]

The suggestion to use a separate VG make sense.

As I look through the script, it seems that not only does it destroy all volumes on lxc, it also destroys all containers irrespective of whether they were created by AIO or by some other application.

Is it an acceptable solution to tweak the message to at least state that the expected behavior is to destroy all containers and volumes, even those created outside of AIO.

[OpenStack Infra (hudson-openstack)]

Fix proposed to branch: master
Review: https://review.openstack.org/249325

[OpenStack Infra (hudson-openstack)]

Reviewed: https://review.openstack.org/249325
Committed: https://git.openstack.org/cgit/openstack/openstack-ansible/commit/?id=96b33d6d1c55e9d9bc723f415d1a3d8561823d13
Submitter: Jenkins
Branch: master

commit 96b33d6d1c55e9d9bc723f415d1a3d8561823d13
Author: Jean-Philippe Evrard <email address hidden>
Date: Tue Nov 24 17:57:27 2015 +0100

    Improve message of the teardown.sh

    Currently the teardown doesn't care about destroy other lxc
    containers and destroys everything.
    Same for logical volumes in the lxc volume group.

    This tweaks the message to make it more clear.

    Closes-Bug: #1518492

    Change-Id: If8bbd9c8b6cea4d18ad982ee676e0480a88a1a4e

[OpenStack Infra (hudson-openstack)]

Reviewed: https://review.openstack.org/250704
Committed: https://git.openstack.org/cgit/openstack/openstack-ansible/commit/?id=6736537eb2974df1f3d55bed1e2389563ad7311c
Submitter: Jenkins
Branch: liberty

commit 6736537eb2974df1f3d55bed1e2389563ad7311c
Author: Jean-Philippe Evrard <email address hidden>
Date: Tue Nov 24 17:57:27 2015 +0100

    Improve message of the teardown.sh

    Currently the teardown doesn't care about destroy other lxc
    containers and destroys everything.
    Same for logical volumes in the lxc volume group.

    This tweaks the message to make it more clear.

    Closes-Bug: #1518492

    Change-Id: If8bbd9c8b6cea4d18ad982ee676e0480a88a1a4e
    (cherry picked from commit 96b33d6d1c55e9d9bc723f415d1a3d8561823d13)

[Doug Hellmann (doug-hellmann)]

This issue was fixed in the openstack/openstack-ansible 12.0.8 release.

[Doug Hellmann (doug-hellmann)]

This issue was fixed in the openstack/openstack-ansible 12.0.9 release.

[Doug Hellmann (doug-hellmann)]

This issue was fixed in the openstack/openstack-ansible 13.0.0 release.

[Davanum Srinivas (DIMS) (dims-v)]

This issue was fixed in the openstack/openstack-ansible 13.0.0 release.

[Davanum Srinivas (DIMS) (dims-v)]

This issue was fixed in the openstack/openstack-ansible 12.0.11 release.