xenapi: stop using get_all_vdis_in_sr to get cached images

[John Garbutt]

get_all_vdis_in_sr is an expensive multiple call, lets try remove some uses of it.

For example, in _find_cached_images, lets consider querying xapi directly for the vdis with the given image-id.

[John Garbutt (johngarbutt)]

replacing with get_all_records_where, see example in get_sr_path

[OpenStack Infra (hudson-openstack)]

Fix proposed to branch: master
Review: https://review.openstack.org/46706

[OpenStack Infra (hudson-openstack)]

Reviewed: https://review.openstack.org/46706
Committed: http://github.com/openstack/nova/commit/ce8e95c7a3fe9d94d6b03ee70c147e5e92aa3b5b
Submitter: Jenkins
Branch: master

commit ce8e95c7a3fe9d94d6b03ee70c147e5e92aa3b5b
Author: John Garbutt <email address hidden>
Date: Fri Sep 13 17:04:06 2013 +0100

    xenapi: stop using get_all_vdis_in_sr in spawn

    Currently when trying to find a cached image, the very expensive call to
    get_all_vdis_in_sr is used, where we could instead fetch the VDI
    directly using a more targeted query.

    Now we are fetching the VDI by name_label we must ensure to clear the
    name_label on newly created VDIs to ensure they do not get picked up by
    later calls to _find_cached_image().

    Fixes bug 1221292

    The above code that checks for _find_cached_images has race conditions
    where its possible to end up with two VDIs returned. To stop this
    happening the code to create the cached images is now synchronized on
    the image being fetched.

    Fixes bug 1226073

    Change-Id: I534fb8f42b00b5d39dc17dd5fee297144b5f379a

[Debasish Das (dskdas)]

Because of the change to _find_cached_image, where it directly gets VDI, there is an issue with changing storage repository.

The issue can be reproduced with the following sequence:
1. Create a VM
2. Switch storage repository (got a new flash card and want to use it instead of spindle disk).
3. Create a VM with the same glance image
4. The VM still gets created on old storage repository, even though _find_sr returns the new storage repository

The only way, I could find, to get the VM to spin up on the new storage repository is to delete all VMs from old storage repository (referring to the glance image) and then create a new one.