What is the best way to keep updated table containing top n records for the last let\'s say 60 minutes for a stream using Spark Structured Streaming? I only want to update m
