DStream.print cost/overhead

[Roberto Coluccio <ro...@gmail.com>]

Hello everyone,

I wasn't able to find any detailed reference about the *DStream.print*
operation "cost" in terms of computation, network, storage, and more.
Having a (soft) real time Spark Streaming processing pipeline made of
several subsequent steps, can a DStream.print operation performed after
each stage create a significant overhead? Why?

Thanks in advance.

Roberto