Hi, I'm trying to use SAS Studio for data preparation within hadoop. I'm using the libname statement or pass-through querries as 'proc sql'. From my understanding SAS generates hiveql code that does the work within hadoop and simply delivers the answer back to SAS Studio. For 'smaller' tables it works just fine. For 'bigger' datasets I need to do the job directly in HIVE or use a work around. e.g. libname myhive hadoop subprotocol=hive2 port=10000 host="myhost" schema=default user=&u_name. pw=&u_pass.; data test; set myhive.bigtable; *the table has around 4m rows and 40 columns, which isn't that big really; run; For this statement I get the following error: ERROR: Prepare error: Error while processing statement: FAILED: Execution Error, return code 1 from org.apache.hadoop.hive.ql.exec.mr.MapRedTask SQL statement: SELECT * FROM `bigtable` I guess its something to do with the size of the table. If I use an obs=1000000 statement and read-in the table in 4 datasets á 1m each it works. I get the same error if I use any proc-step or proc-sql step on that data table. If I e.g. want to get some simple proc freq counts I get the same error. Any help, ideas, work arounds would be highly appreciated!
... View more
