topic SAS Hadoop connection with EP (Embedded process) for Base SAS Codes in New SAS User

SAS Hadoop connection with EP (Embedded process) for Base SAS Codes

shrriam — Thu, 03 Oct 2019 06:27:24 GMT

Can i run my Base SAS codes for Hadoop using SAS EP (Embedded Process) ?

Or do i have to first convert those Base Codes to DS2 with the help PROC DS2 and only then can i use SAS EP?

Re: SAS Hadoop connection with EP (Embedded process) for Base SAS Codes

AnandVyas — Thu, 03 Oct 2019 06:51:59 GMT

Hi @shrriam ,

This document lists out requirements and restrictions for Data step in Hadoop.

Re: SAS Hadoop connection with EP (Embedded process) for Base SAS Codes

shrriam — Thu, 03 Oct 2019 07:04:56 GMT

Thanks for the response !!

The document says , " If you have installed and activated the SAS Embedded Process on a Hadoop cluster, it is possible for DATA step code to be executed in parallel against the input data residing on the HDFS file system. "

Does it mean , It allows the normal Base SAS codes to run with SAS EP ? I was confirmed from SAS that EP is widely for DS2 performance enhancements , we may have to convert the codes to DS2 in order to leverage the performance .

Re: SAS Hadoop connection with EP (Embedded process) for Base SAS Codes

AnandVyas — Thu, 03 Oct 2019 07:42:09 GMT

Yes, your base sas code has to ref to the libref pointing to HDFS as input and output library. In a general scenario, you don't have to convert your code to DS2. Below link gives you an overview on when to use DS2.

https://documentation.sas.com/?docsetId=ds2ref&docsetTarget=p1shb3eskw9xh9n1orhpgt608it7.htm&docsetVersion=9.4&locale=en

It's true that EP is used for performance enhancements using in-database processing or parllel loading of Data if you have SAS High performance analytics cluster. It would vary from case to case basis. If SAS has recommended to convert to DS2 for performance improvement, then you should give it a shot and test it out.