lbaas: delete_health_monitor() should rely upon cascade operations

[Roman Podoliaka]

In LBaaS DB schema we have a Pool <--> HealthMonitor many-to-many association which is done
via the PoolMonitorAssociation model.

A HealthMonitor instance may be deleted by calling the "delete_health_monitor()" method
which is defined in quantum/db/loadbalancer/loadbalancer_db module.

Currently, this method performs a "manual" deletion of a HealthMonitor instance given its id:
it gets a list of all PoolMonitorAssociation instances filtered by a HealthMonitor id, for each of those
instances it retrieves a corresponding Pool instance and finally removes the association instance
from the "monitors" collection of a Pool instance.

We can simplify the deletion of HealthMonitor instances by relying upon cascade operations
defined for SQLAlchemy relationships (which is already done for Pool model).

[OpenStack Infra (hudson-openstack)]

Fix proposed to branch: master
Review: https://review.openstack.org/26959

[OpenStack Infra (hudson-openstack)]

Reviewed: https://review.openstack.org/26959
Committed: http://github.com/openstack/quantum/commit/ee50162b27a3e0bf310f8e4f273b69cd0660cf69
Submitter: Jenkins
Branch: master

commit ee50162b27a3e0bf310f8e4f273b69cd0660cf69
Author: Roman Podolyaka <email address hidden>
Date: Mon Apr 15 18:10:07 2013 +0300

    Simplify delete_health_monitor() using cascades

    Currently delete_health_monitor() emulates behaviour of
    cascade deletion: when a HealthMonitor instance is deleted,
    all corresponding PoolMonitorAssociations are queried and
    deleted one by one. This can be done automatically by means
    of SQLAlchemy if we set proper cascade flags on the relationship
    between HealthMonitor and PoolMonitorAssociations models.

    Fixes bug 1169107.

    Change-Id: I674c381192719a56433f67be53a32203c36ebe2b