This is how it works, I think: sc.parallelize(..) takes the variable inside the (..) and returns a “distributable equivalent” of that variable. That is, an RDD is returned.
This RDD can be worked on by multiple workers threads in _parallel_. The parallelize(..) has to be done on the driver running on the Master(and not on workers), so that the resulting RDD’s parts can be sent across to the workers who have enough resources. Each worker processes a part of the RDD. So it is the “sc.paralleize (..) operation is performed in the driver program” that enables scalability. Horizontal scalablity. From: Akhil Das [mailto:ak...@sigmoidanalytics.com] Sent: 26 June 2015 13:33 To: shahab Cc: user@spark.apache.org Subject: Re: Performing sc.paralleize (..) in workers not in the driver program Why do you want to do that? Thanks Best Regards On Thu, Jun 25, 2015 at 10:16 PM, shahab <shahab.mok...@gmail.com<mailto:shahab.mok...@gmail.com>> wrote: Hi, Apparently, sc.paralleize (..) operation is performed in the driver program not in the workers ! Is it possible to do this in worker process for the sake of scalability? best /Shahab The information contained in this electronic message and any attachments to this message are intended for the exclusive use of the addressee(s) and may contain proprietary, confidential or privileged information. If you are not the intended recipient, you should not disseminate, distribute or copy this e-mail. Please notify the sender immediately and destroy all copies of this message and any attachments. WARNING: Computer viruses can be transmitted via email. The recipient should check this email and any attachments for the presence of viruses. The company accepts no liability for any damage caused by any virus transmitted by this email. www.wipro.com