I did a couple tests in which I would retrieve all the data (multiple thousand of values) with the timestamp within last second from * and keep doing that every second afterwards. I would do a simple calculation and send the new data back to PI
A) 1st approach was to retrieve recorded values given a start and end time. (multi-threaded for each point)
B) 2nd approach was to use eventpipes and just get whatever was in there every second. (Multi-threaded for each point)
B seemed to work much faster than A but I would like an expert opinion on it. I also would like to know if the data always come chronologically ordered when I get it from the eventpipe?