[
https://issues.apache.org/jira/browse/PIG-4277?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanel&focusedCommentId=14500678#comment-14500678
]
Mohit Sabharwal commented on PIG-4277:
--------------------------------------
All failures except the following are due to "group by" results not being
ordered by group key in Spark (but is ordered in MR).
{{testCrossAfterGroupAll}} expects "auto-parallelism" to be 2 or higher.:
{code}
// auto-parallelism is 2 in MR, 20 in Tez, so check >=2
Assert.assertTrue(partFiles.length >= 2);
{code}
> Enable unit test "TestEvalPipeline2" for spark
> ----------------------------------------------
>
> Key: PIG-4277
> URL: https://issues.apache.org/jira/browse/PIG-4277
> Project: Pig
> Issue Type: Sub-task
> Components: spark
> Reporter: liyunzhang_intel
> Assignee: Mohit Sabharwal
> Fix For: spark-branch
>
> Attachments: TEST-org.apache.pig.test.TestEvalPipeline2.txt
>
>
> error log is attached
--
This message was sent by Atlassian JIRA
(v6.3.4#6332)
