0

I'm getting errors when running PROC DS2 code to read a large data set residing on a different server. The LIBNAME statement uses a remote engine. SAS currently doesn't have a driver that can read remote data sets. PROC DS2 uses drivers not engines. This is not available yet in base sas as the documentation claims. SAS told me they will add a defect for their R&D. If anyone knows a work around on how to run DS2 code on a remote server please share your code. I use a remote server because the data is huge and wanted to use DS2 to minimize job run time. 

proc ds2;
data test201/overwrite=yes;
method run();
set {select * from randlib.Tax_prd};
end;
enddata;
run;

These are the errors. LIBNAME randlib '/username/sasfile' server = XXXX; NOTE: Libref RANDLIB was successfully assigned as follows: Engine: REMOTE Physical Name: /username/sasfile LIBNAME randlib '/username/sasfile' server = XXXX; NOTE: Libref RANDLIB was successfully assigned as follows: Engine: REMOTE Physical Name: /username/sasfile ERROR: Compilation error. ERROR: BASE driver, schema name RANDLIB was not found for this connection ERROR: Table "RANDLIB.TAX_PRD" does not exist or cannot be accessed ERROR: Line 215: Unable to prepare SELECT statement for table

user601828
  • 499
  • 3
  • 7
  • 17
  • What do you mean by a remote server? Do you mean you used SAS/Connect to connect your primary SAS session to another SAS session running on a different server? Provide more details about what you are trying to do and examples of the code you are trying to use. – Tom Dec 04 '19 at 19:04
  • Is your remote server sql, oracle? I'm guessing it's not a SAS db system, which is the issue? – Reeza Dec 04 '19 at 19:30
  • Show the errors in context in the question text (not comments). Are you running DS2 code in your local session or the remote session? Can you access the remote dataset using a simple data step instead of a ds2 step? – Tom Dec 04 '19 at 19:54
  • My remote server is a space allocation in unix. The data is in Oracle. I'm running the DS2 code in my remote session. I can't run it locally because the data is over a billion records. I use rsubmit to sign on to the remote server. Proc sql works it just runs for a few days. I was trying to use DS2 since I read it speeds up processing time. SAS confirmed that this is a defect because DS2 doesn't work on a remote server. I was trying for a few days but found out yesterday that I found a defect. – user601828 Dec 04 '19 at 20:38

0 Answers0