There’s no control you can have on it. And I don’t understand why you would control it… Cache is implemented everywhere in Weather Station and this seems to me relatively well balanced (between freshness of data, capacity of services and stations and API quotas).
To me more specific:
– for Netatmo, Weather Underground and located stations (via OpenWeatherMap), the polling is done every 10 minutes (cached in database);
– for clientraw.txt and realtime.txt stations, the polling is done every 5 minutes (cached in database);
– for current observations (via OpenWeatherMap) added to all stations, the polling is done every 15 minutes (cached in database);
– for pollution data (via OpenWeatherMap) added to all stations, the polling is done every 30 minutes (cached in database);
– for computations (indexes and astronomical) it is done when new data is store (on a station by station basis) and cached in database;
– for stations pulling (to share data with PWS, MET Office and Wunderground), when it is activated, it is done every 10 minutes (which is the recommended value) – of course pulling have no cache;
– for backend and frontend sql queries, the cache is set on 30 minutes with invalidation policy (transient caching);
– for backend rendering, the “duplicate queries” cache is set based on the complete cycle for generating a page (cached in memory).
Obviously, only polling cache have effect on reducing API calls (see the first 4 lines).
I’m sorry if this answer is not what you are waiting for, I doesn’t understand what is really your question.