Bug 1514928

Description of problem:

After removing a container provider which had configured with Prometheus alerts, 
if other provider set with the same name
the old alerts will show in the alerts screen


Version-Release number of selected component (if applicable):


How reproducible:

1. define an alert on OCP for "too many pods on node"
2. define alert and alert policy on CFME, assign a policy to the enterprise
3. add the provider to CFME
4. verify that the alerts appear in alerts dashboard
5. remove the specific provider
6. remove the old alert definition from OCP
6. define an alert on OCP for "node up"
7. remove Prometheus pod (to load the new alert configuration)
8. define alert and alert police on CFME, assign a policy to the enterprise
8. add the provider to CFME with the same name
9. old alerts will show up in the alerts' dashboard

Actual results:
Unrelevant (old) alerts display in the dashboard screen

Expected results:
Only new alerts have to be shown in the alert dashboard

Additional info: