High Performance Spark Quotes

“Co-partitioning is related to but distinct from partition co-location. We say that multiple RDDs are co-partitioned if they are partitioned by the same known partitioner.”
― Holden Karau, High Performance Spark: Best Practices for Scaling and Optimizing Apache Spark

“Co-located RDDs are RDDs with the same partitioner that reside in the same physical location in memory.”
― Holden Karau, High Performance Spark: Best Practices for Scaling and Optimizing Apache Spark

“Beyond being less likely to run out of memory than groupByKey, the following four functions — reduceByKey, treeAggregate, aggregateByKey, and foldByKey — are implemented to use map-side combinations, meaning that records with the same key are combined”
― Holden Karau, High Performance Spark: Best Practices for Scaling and Optimizing Apache Spark