handlers notify restart haproxy container multiple times

[Xing Zhang]

**Bug Report**

What happened:
When haproxy config changed, notify in deploy, reconfigure or upgrade will notify and run handlers multiple times for haproxy:
RUNNING HANDLER [haproxy-config : Restart haproxy container]

What you expected to happen:
Restart haproxy container once.

How to reproduce it (minimal and precise):
Deploy for the first time, reconfigure when haproxy config changed or upgrade
see CI logs like kolla-ansible-centos8-binary, such as:
https://fb6b99d3b00179205d1b-88bfc8d18e6df19b9cb6dbfc3489a6ab.ssl.cf2.rackcdn.com/708213/29/check/kolla-ansible-centos8-binary/e5b11ca/primary/logs/ansible/deploy
NOTE:
grep
'NOTIFIED HANDLER haproxy-config : Restart haproxy container for primary'
and
'RUNNING HANDLER [haproxy-config : Restart haproxy container]'

Affected Versions:
Refactor Haproxy from train in https://review.opendev.org/#/c/601648/