Watch this demo on dynamic partition pruning. Spark takes any query and converts the query into logical plan and later physical plan of the query. Dynamic partition pruning reduces the scanned data limiting it to only needed data and avoids scanning or reading of unwanted data. Using this feature will reduce the number of I/O reads and increase the speed and time of queries. This optimization is applied at run time and works when two or more fact tables are joined to refer dimension tables. The partitions are pruned at execution time when the filters are applied on the dimension table. The examples below shows how the tables are created and later joined.
Learn more here - https://blog.cloudera.com/build-on-yo...
Смотрите видео Demo: Spark 3 Dynamic Partition Pruning онлайн без регистрации, длительностью часов минут секунд в хорошем качестве. Это видео добавил пользователь Cloudera, Inc. 11 Сентябрь 2020, не забудьте поделиться им ссылкой с друзьями и знакомыми, на нашем сайте его посмотрели 890 раз и оно понравилось 5 людям.