I’m stuck on what the approach should be for caching serialized objects as part of search queries. Right now, the serializers are performing extra queries and even with prefetch_related, a typical query is taking around 40 seconds to complete.

            trips = (
                Trip.objects.filter(filter_query)
                .prefetch_related(
                    "trip_values",
                    "trip_values__operator_trip_field",
                    "trip_values__operator_trip_field__trip_field",
                    "operator_service_day",
                    "trip_occurrences",
                    "trip_locations",
                    "assigned_vehicle_formation",
                )
                .all()
            
            query_time = time.time() - search_start
            print("Query time: ", query_time)
            serialization_start = time.time()
            serialized = TripListSerializer(trips, many=True)
            serialization_time = time.time() - serialization_start
            print("Serialization time: ", serialization_time)

            return Response(serialized.data)

So, if I wanted to cache each object instead, I would have to perform the same query as above, but I would then need to loop over each id and fetch it’s cache and then combine the cache. So if i have 200+ search results, that’s 200+ cache lookups.

Is that really the approach I’m suppose to go or is there a better alternative.