Ben Chuanlong Du's Blog

It is never too late to learn.

Spark Issue: High Disk and Memory Spill When Doing Shuffle

Things on this page are fragmentary and immature notes/thoughts of the author. Please read with your own judgement!

Symtom

High disk and memory spill when doing shuffle.

Cause

Insufficient executor memory (you can monitor this spill metrics from Spark UI).

Solution

  1. Increase executor memory.

    --executor-memory=4G
    
  2. For jobs that …