Create a single stat panel by monitoring multiple end points

I am using blackbox to scrape the endpoints of 2 OCP cluster. The clusters are HA. I need to create a dashboard which gives me a availability of OCP over 30 days. Even if one endpoint is down the availability should not dip unless both are down.
my query is
max(avg_over_time(probe_success{instance=~“ |”}[30d]))*100.
I do not get the desired result. Can someone help me with this.