问题描述
我试图从Hive表中读取一行,并将其字段用作我的Spark应用程序其余部分中的常量.
I'm trying to read a single row from a Hive table and use its fields as constants in the rest of my Spark application.
object IniConstHive extends Serializable {
val techTbl: DataFrame = spark.table("my_db.my_conf_table").orderBy(col("my_date").desc)
val firstrow: Row = techTbl.head
val my_firstfield: Double = firstrow.getAs[java.math.BigDecimal](0).doubleValue
val my_secondfield: Double = firstrow.getAs[java.math.BigDecimal](1).doubleValue
}
不幸的是,我在调用IniConstHive.my_firstfield时遇到了此异常:
Unfortunately, I get this exception when I call IniConstHive.my_firstfield:
Caused by: java.util.NoSuchElementException: None.get
at scala.None$.get(Option.scala:347)
at scala.None$.get(Option.scala:345)
at org.apache.spark.storage.BlockInfoManager.releaseAllLocksForTask(BlockInfoManager.scala:343)
at org.apache.spark.storage.BlockManager.releaseAllLocksForTask(BlockManager.scala:676)
at org.apache.spark.executor.Executor$TaskRunner.run(Executor.scala:329)
at java.util.concurrent.ThreadPoolExecutor.runWorker(ThreadPoolExecutor.java:1149)
at java.util.concurrent.ThreadPoolExecutor$Worker.run(ThreadPoolExecutor.java:624)
at java.lang.Thread.run(Thread.java:748)
这是正确的方法吗?有没有更有效的方法来达到结果?
Is this the right approach to do this? Is there a more efficient way to acheive the result?
推荐答案
原来,问题出在IniConstHive对象中,而不是我调用它的地方.实际上,我试图在自定义UDAF中调用IniConstHive.my_firstfield,该自定义UDAF在执行程序内部运行.因此,我的应用程序尝试在驱动程序之外实例化多个SparkContext.
It turned out the problem was not in the IniConstHive object but the place where I called it. In fact, I was trying to call IniConstHive.my_firstfield inside a custom UDAF, which runs inside executors. For this reason my application tried to instantiate multiple SparkContext outside the driver.
我通过在驱动程序中调用IniConstHive.my_firstfield并将结果作为参数传递给UDAF构造函数来解决.
I solved by calling IniConstHive.my_firstfield in the driver and then passing the result as a parameter in the UDAF constructor.
这篇关于如何收集单行数据框并将字段用作常量的文章就介绍到这了,希望我们推荐的答案对大家有所帮助,也希望大家多多支持!