[
https://issues.apache.org/jira/browse/PIG-3975?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanel&focusedCommentId=14039348#comment-14039348
]
Koji Noguchi commented on PIG-3975:
-----------------------------------
Had a loong chat with Rohini on learning what is happening. This is my patch
but Rohini knew much more on what actually is happening. (as always)
So as I understand, my testcase in the patch caught another bug introduced
after PIG-3757.
On trunk, with my testcase but without my patch, testcase fails at
{noformat}
junit.framework.AssertionFailedError: Unexpected number of mapreduce job.
Missing concat job? expected:<4> but was:<5>
at
org.apache.pig.test.TestFRJoin2.testSoftLinkDependencyWithMultipleScalarReferences(TestFRJoin2.java:125)
{noformat}
This is even before my check on the missing dependency. It's failing due to
each scalar *reference* creating a new concat jobs. By having extra "
seen.add(newSpec);", it can avoid the repetitive concat creation attempt.
+1 on both the patches.
> Multiple Scalar reference calls leading to missing records
> ----------------------------------------------------------
>
> Key: PIG-3975
> URL: https://issues.apache.org/jira/browse/PIG-3975
> Project: Pig
> Issue Type: Bug
> Affects Versions: 0.8.1, 0.9.2, 0.10.1, 0.11.1, 0.12.2
> Reporter: Koji Noguchi
> Assignee: Koji Noguchi
> Priority: Critical
> Fix For: 0.13.0
>
> Attachments: pig-3975-v01_withouttest.patch,
> pig-3975-v02_withtests.patch, pig-3975-v03-additional-fix.patch,
> pig-3975-v03-trunk.patch, pig-3975-v03.patch
>
>
> We noticed that multiple pig runs with same input were producing different
> outputs.
> Simplified script looked like this.
> {noformat}
> A = load 'input1' as (a1:int);
> B = group A by a1 parallel 200;
> C = load 'input2' as (c1:int);
> D = foreach C generate B.$0;
> store D into '/tmp/deletemeD';
> E = load 'input3' as (c1:int);
> F = foreach E generate B.$0;
> store F into '/tmp/deletemeF';
> {noformat}
--
This message was sent by Atlassian JIRA
(v6.2#6252)
