We are seeing an issue when using Summaries2 to pull data for a certain period of time (start timestamp, end timestamp, sampling interval), specifically that when running the exact same query (using PI SDK in MATLAB) where in certain instances, we will have one fewer datapoint than expected (out of 100-1000 points). This happens intermittently as far as we can tell.
This consistency issue is not isolated to a particular set of start/end timestamps, nor is it isolated to a particular parameter.
Although we have a workaround (where we ask for more data, and then perform trimming of datasets in MATLAB), it is somewhat concerning that even when calling the exact same function with the exact same parameters, we can get (ever so slightly) different results, as far as number of datapoints. (As far as we can tell, the values themselves are the same from query to query, just the fact that there is one fewer datapoint on occasion.)
If anyone else has seen this (and/or know why we see this issue from time to time), feel free to let me know.