I recently turned on High Packet Loss monitoring, threshold set to >20%. However, after getting several alerts per day, I've realized that NPM must be basing it's measurement on some small number of pings, like 5 -- and thus is triggering 20% loss off just a single spurious packet drop.
Is there a way to make this alert more useful? For example, (a) increase the number of pings to make the results more granular, or (b) tell S.W. to follow up a single detected packet drop with a more proper test -- like 100+ pings?