sparksql读取数据库问题

spark jdbc方式读取数据库，每个excutor都会去拉数据？

说明：该问题来源spark技术分享朋友群

spark.read("jdbc")
  .option("url", url)
  .option("dbtable", "pets")
  .option("user", user)
  .option("password", password)
  .option("numPartitions", 10)
  .option("partitionColumn", "owner_id")
  .option("lowerBound", 1)
  .option("upperBound", 10000)
  .load()

这种一般会转换为

SELECT * FROM pets WHERE owner_id >= 1 and owner_id < 1000
SELECT * FROM pets WHERE owner_id >= 1000 and owner_id < 2000
SELECT * FROM pets WHERE owner_id >= 2000 and owner_id < 3000
...