jenkins would fail randomly for not enough disk space

[Kai Qiang Wu(Kennan)]

accroding to infra guy, if we boot two instances(for k8s bay), use flavor 10GB would be ok for either rackspace cloud or hp cloud.

So we need create special flavors in our side

[OpenStack Infra (hudson-openstack)]

Fix proposed to branch: master
Review: https://review.openstack.org/212954

[OpenStack Infra (hudson-openstack)]

Reviewed: https://review.openstack.org/212954
Committed: https://git.openstack.org/cgit/openstack/magnum/commit/?id=644cfdf8b509f0ba1f5e6df2029c4fae274b6cbc
Submitter: Jenkins
Branch: master

commit 644cfdf8b509f0ba1f5e6df2029c4fae274b6cbc
Author: Kai Qiang Wu(Kennan) <email address hidden>
Date: Fri Aug 14 03:19:19 2015 +0000

    Use magnum specific flavor

    As infra uses public cloud to run jenkins jobs, some cloud
    provider do not have large disk space, for example, only 20GB
    root disk, not fit for k8s bay instance space requirements.
    According to infra core said, if two instances are booted,
    each with 10GB, it could all meet in back end clouds.
    So we need to create specific flavor for magnum use. It is
    verified in local env 8GB can work, so we will use 8GB now.

    Closes-Bug: #1484772
    Change-Id: I159623f77ad7e3083a5296b0ccc7d65cfb57e001