diff --git a/docs/running-on-yarn.md b/docs/running-on-yarn.md
index cd18808681ece7bbe3d37788746566421d67b18c..fe0221ce7c5b65822177998038fd64182e696a8a 100644
--- a/docs/running-on-yarn.md
+++ b/docs/running-on-yarn.md
@@ -559,6 +559,8 @@ pre-packaged distribution.
 1. In the `yarn-site.xml` on each node, add `spark_shuffle` to `yarn.nodemanager.aux-services`,
 then set `yarn.nodemanager.aux-services.spark_shuffle.class` to
 `org.apache.spark.network.yarn.YarnShuffleService`.
+1. Increase `NodeManager's` heap size by setting `YARN_HEAPSIZE` (1000 by default) in `etc/hadoop/yarn-env.sh` 
+to avoid garbage collection issues during shuffle. 
 1. Restart all `NodeManager`s in your cluster.
 
 The following extra configuration options are available when the shuffle service is running on YARN: