> For the complete documentation index, see [llms.txt](https://nag-9-s.gitbook.io/spark-notes/llms.txt). Markdown versions of documentation pages are available by appending `.md` to page URLs; this page is available as [Markdown](https://nag-9-s.gitbook.io/spark-notes/resilient-distributed-dataset-rdd/pair-rdds.md).

# Pair RDDs

Pair RDDs

Pair RDDs behave pretty similar to basic RDDs. The main difference is that RDD elements are key value pairs, and key value pairs are a natural fit for distributed computing problems.

Pair RDDs are heavily used to perform various kinds of aggregations and initial Extract Transform Load procedure steps

Performance-wise there’s one important, rarely mentioned fact about

them: Pair RDDs don’t spill on disk. Only basic RDDs can spill on disk.

means that a single Pair RDD must fit into computer memory. If Pair RDD content is larger than the size of the smallest amount of RAM in the cluster, it can’t be processed.