Question about parallelism of saveAsNewHadoopFile

[chi zhang]

Hi,

I'm testing my spark program recently, in which I have 4 RDD of different data, after a groupby and some map operations, I call the saveAsNewAPIHadoopFile to save my data to HDFS.

I saw that the groupby and  saveAsNewAPIHadoopFile operations were processed one by one in spark web monitor. I want to know that why are these 4 saveAsNewAPIHadoopFile operations not concurrently? 

thanks:)