运行mapreduce出现java heap space

运行后，map的0%都没完成，直接
org.apache.hadoop.mapred.YarnChild: Error running child : java.lang.OutOfMemoryError: Java heap space

受到这句话的启发
For us to see why your job is running out of memory we would probably need to see your code. Perhaps you are creating memory-intensive objects every map() that could instead be created once in setup() and re-used every map()?

去检查了自己的代码，果然，有个HashSet放在map中初始化了。改在setup阶段初始化，map使用前clear，问题解决。