BookmarkSubscribeRSS Feed
🔒 This topic is locked. We are no longer accepting replies to this topic. Need further help? Please sign in and ask a new question.
WWD
Obsidian | Level 7 WWD
Obsidian | Level 7

Within Module 2 of the Data Scientist Certification material (Loading Big Data), there are two demonstrations that use “proc DS2 ds2accel = yes’.  In hps04d04, the option set = gridhost = is not set.  In hps0602, the option set = gridhost = is used.

 

My question is by setting the gridhost value, do all calculations move from “in-database” to “in-memory”?

 

Thank you.

1 REPLY 1
DavidGhan
SAS Employee

I don't see a program hps06d02.sas. But I can speak generally to your questions.

 

For DS2 the use of options set=Gridhost is NOT needed in order to use the code accelerator to run the DS2 process in-database. It does appear in some demo programs but that was really an error on my part to include it there when demonstrating DS2 running in-database. Options set=gridhost is only used with the High Performance Analytics and LASR procedures and identifies the High Performance Grid location or the LASR Server location so it is needed only for those technologies. However, DS2 uses the Code Accelerator software for in-database processing which is a separate technology.