ensure that workers are registered and have sufficient resources
spark-cluster启动的配置里配置了每个worker的内存，如果你submit的时候指定的资源大于这个，就会报上述错
