spark Could not write all entries

使用 spark 將 dataFrame 儲存到 elasticsearch 出現以下報錯：

Caused by: org.elasticsearch.hadoop.EsHadoopException: Could not write all entries [1/1] (Maybe ES was overloaded?). Error sample (first [1] error messages):
	rejected execution of org.elasticsearch.transport.TransportService$4@7d5f91de on EsThreadPoolExecutor[bulk, queue capacity = 50, org.elasticsearch.common.util.concurrent.EsThreadPoolExecutor@3447703a[Running, pool size = 32, active threads = 32, queued tasks = 68, completed tasks = 9151096]]
Bailing out...

這個沒法查到定位到報錯位置，因此在新建 spark 的時候進行以下配置：

val masterUrl = "local"
val appName = "ttyb"
val sparkConf = new SparkConf()
  .setMaster(masterUrl)
  .setAppName(appName)
  .set("es.nodes", "172.16.14.21")
  .set("es.port", "9200")
  //Bailing out...錯誤
  .set("es.batch.size.entries", "1")
  //插入失敗後無限重複插數據
  .set("es.batch.write.retry.count", "-1")
  //查數據等待時間
  .set("es.batch.write.retry.wait", "100")
val Spark = SparkSession.builder().config(sparkConf).getOrCreate()

獲得新的錯誤：

org.elasticsearch.hadoop.rest.EsHadoopInvalidRequest: 
null

報錯顯示：

ES 負載太高，須要從新修復

本想重啓 ES ，發現是機器 磁盤空間已滿 ，查錯成功