better error handling in casting relations to scalars
-----------------------------------------------------
Key: PIG-1550
URL: https://issues.apache.org/jira/browse/PIG-1550
Project: Pig
Issue Type: Bug
Reporter: Olga Natkovich
Assignee: Thejas M Nair
Fix For: 0.8.0
I ran the following script:
Input data:
joe 100
sam 20
bob 134
Script:
A = load 'user_clicks' as (user: chararray, clicks: int);
B = group A by user;
C = foreach A generate group, SUM(A.clicks);
D = foreach A generate clicks/(double)C.$1;
dump C;
Since C contains more than 1 tuple, I expected to get an error which I did.
However, the error was not very clear. When the job failed, I did see a valid
error (however it lacked the error code): 210630 [main] ERROR
org.apache.pig.tools.pigstats.PigStats - ERROR 0: Scalar has more than one row
in the output
However at the end of processing, I saw a misleading error:
210709 [main] ERROR org.apache.pig.tools.grunt.Grunt - ERROR 2088: Unable to
get results for:
hdfs://wilbur20.labs.corp.sp1.yahoo.com:9020/tmp/temp818551960/tmp1063730945:org.apache.pig.impl.io.InterStorage
10/08/19 17:16:22 ERROR grunt.Grunt: ERROR 2088: Unable to get results for:
hdfs://wilbur20.labs.corp.sp1.yahoo.com:9020/tmp/temp818551960/tmp1063730945:org.apache.pig.impl.io.InterStorage
--
This message is automatically generated by JIRA.
-
You can reply to this email to add a comment to the issue online.
