There is an issue with this patch. It's main purpose is the ability to calculate values of pg_stat_statements view for a time period between two samples without resetting pg_stat_statements being absolutely sure that the statement was not evicted. Such approach solves the problem for metrics with except of min and max time values. It seems that partial reset is needed here resetting min/max values during a sample. But overall min/max values will be lost in this case. Does addition of resettable min/max metrics to the pg_stat_statemets view seems reasonable here?