I put the SparkContext instance to RDD processed to get to the middle of the data, and the intermediate data needs to be Shared all nodes, and I should how to operate, can broadcast it out? Detailed point is, we according to the log file to read out all the frozen accounts, and these were frozen accounts need to be Shared, so good judgment,

Thank you, I a fledglings guard, also please feel free to guide,

CodePudding user response:

See how each machine is you want to use this data, such as his output to the HDFS, so each machine have to do is read from the HDFS, or need specific see you how to use the data

CodePudding user response:

Sc. Broadcast