How to Spark implement the interactive In-Memory Cache?

Question

I wonder if a program ends, the memory used for the piece of program are freed by GC.

And how to cache the data in Spark when I am in a interactive scala interpreter?

Is that mean one time of interpreter is in one process?

But much more often, I use terminal to run the code, not in the interpreter itself, in this case, how can I achieve In-Memory?

Answer 1

无论是使用解释器还是通过命令提示符，都可以使用rdd.cache()在内存中使用keep rdd。