https://community.cloudera.com/t5/Support-Questions/Spark-driver-memory-keeps-growing/m-p/165011#M127378 <P>The job is coded in scala and uses default serialization parameters.</P><P>The object has no reference to the RDD itself.</P><P>What do you mean by "broadcast join"?</P><P>Here is the current DAG:</P><P><span class="lia-inline-image-display-wrapper lia-image-align-inline" image-alt="6472-dag.png" style="width: 1138px;"><img src="https://community.cloudera.com/t5/image/serverpage/image-id/20423i503D4F8F99666E97/image-size/medium?v=v2&amp;px=400" role="button" title="6472-dag.png" alt="6472-dag.png" /></span></P> Sun, 18 Aug 2019 11:23:43 GMT pvillard 2019-08-18T11:23:43Z