Rescue doesn't use supplied admin_pass

[Johannes Erdfelt]

A recent change broke rescue when an admin_pass is supplied by the user.

The change that broke this was https://review.openstack.org/#/c/10432/7

The problem is that it tries to store admin_pass in the database. There is no admin_pass column in the database, so it gets silently dropped.

This is also complicated by the fact that the xenapi driver expects admin_pass to be a part of the instance dict and will fail with an exception during rescue.

[OpenStack Infra (hudson-openstack)]

Fix proposed to branch: master
Review: https://review.openstack.org/11080

[OpenStack Infra (hudson-openstack)]

Reviewed: https://review.openstack.org/11080
Committed: http://github.com/openstack/nova/commit/dd45bf735c073c9fb59d1d420d2e47713d28577d
Submitter: Jenkins
Branch: master

commit dd45bf735c073c9fb59d1d420d2e47713d28577d
Author: Johannes Erdfelt <email address hidden>
Date: Thu Aug 9 15:17:35 2012 +0000

    Fix setting admin_pass in rescue command

    Fixes bug 1034948

    The database doesn't have an admin_pass column, so the instance_update()
    call is a noop.

    Change-Id: I9f22de1e926ed48b8fd4561ddf2129db6a1e0b65