This bug was initially created as a copy of Bug #1774459

Customer thinks that RHOSP13 needs to have the "octavia failover circuit breaker"
feature as follows:

~~~
When a disaster (e.g. loss of amphora boot network reachability) happens,
Octavia health manager try to execute failover for all amphorae and these failovers will fail.

We need to prepare the disaster situations before the situation actually happens.

As discussed in this story https://storyboard.openstack.org/#!/story/2005604,
failover circuit breaker is required for production octavia service.

RFE: https://storyboard.openstack.org/#!/story/2005604
Patch: https://review.opendev.org/#/c/656811/

Step to Reproduce:
For example, drop all heartbeat from all amphorae in controller nodes.

Actual Results:
Health manager tries to execute failover for all existing amphorae

Expected Results:
Health manager stops some failovers based on the number of simultaneous failovers.

Version-Release number of selected component:
Red Hat OpenStack Platform Version Number: OSP13
Release Number:
Related Package Version:
openstack-octavia-*-2.1.0-1.el7ost.noarch.rpm
Related Middleware/Application:
Underlying RHEL Release Number: 7.6
Underlying RHEL Architecture: x86_64
Underlying RHEL Kernel Version: 3.10.0-957.1.3.el7.x86_64

~~~