gistfile1.eclass

JVM_SIZE="-server -Xms18g -Xmx18g"

# New Generation Sizes (-XX:NewSize -XX:MaxNewSize)
JVM_SIZE_NEW="-XX:NewSize=2048m -XX:MaxNewSize=2048m"

# Type of Garbage Collector to use
JVM_GC_TYPE="-XX:+UseConcMarkSweepGC -XX:+UseParNewGC"

# Tuning options for the above garbage collector
JVM_GC_OPTS="-XX:CMSInitiatingOccupancyFraction=70 -XX:SurvivorRatio=2"

# JVM GC activity logging settings ($LOG_DIR set in the ctl script)  ## DME: PrintTenuringDistribution not set in master_conf.sh
JVM_GC_LOG="-XX:+PrintTenuringDistribution -XX:+PrintGCDetails -XX:+PrintGCTimeStamps -Xloggc:$LOG_DIR/gc.log"

You mentioned that these settings are for nodes with 32Gg RAM. Why are you allocating more than 18Gb heap? Are there diminishing returns due to GC?