It would be far more efficient I think for the nodes to keep their data for several cycles before finally sending it to the api. This would mean that it would capture the environment data several times before actually reporting to the api. This would mean far far less posts to the api. The posts that do happen however will be slightly larger but that is ok. It will be more efficient. The api server will then just aggregate the data to the rest autonomously.
It would be far more efficient I think for the nodes to keep their data for several cycles before finally sending it to the api. This would mean that it would capture the environment data several times before actually reporting to the api. This would mean far far less posts to the api. The posts that do happen however will be slightly larger but that is ok. It will be more efficient. The api server will then just aggregate the data to the rest autonomously.