[ 
https://issues.apache.org/jira/browse/HIVE-964?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanel&focusedCommentId=12801981#action_12801981
 ] 

John Sichi commented on HIVE-964:
---------------------------------

Regarding the "Output: defa...@t1" diff:

I noticed something similar while I was developing negative tests for CREATE 
VIEW.  I think the problem is that if the same object name (like "t1") is used 
in more than one negative test script, then you'll get different results for 
DROP TABLE depending on whether the test is run by itself vs as part of a 
suite. The problem is that negative tests don't get a chance to clean up after 
themselves, so the execution of the pre-drop in the next test is going to 
depend on whether a previous test ran leaving behind objects.

I worked around it by changing my scripts to use unique object names.

A less brittle approach (which I've used on similar test frameworks in the 
past) is to automate the "pre-drop" logic to make it generic at the beginning 
of each test.  The test framework scans the catalog and issues a DROP for each 
existing object (in silent mode).  This also means test authors can skip 
writing both pre-drop and post-drop (regardless of whether it is a positive or 
negative test), which is nice.


> handle skewed keys for a join in a separate job
> -----------------------------------------------
>
>                 Key: HIVE-964
>                 URL: https://issues.apache.org/jira/browse/HIVE-964
>             Project: Hadoop Hive
>          Issue Type: Improvement
>          Components: Query Processor
>            Reporter: Namit Jain
>            Assignee: He Yongqiang
>             Fix For: 0.6.0
>
>         Attachments: hive-964-2009-12-17.txt, hive-964-2009-12-28-2.patch, 
> hive-964-2009-12-29-4.patch, hive-964-2010-01-08.patch, 
> hive-964-2010-01-13-2.patch, hive-964-2010-01-14-3.patch, 
> hive-964-2010-01-15-4.patch
>
>
> The skewed keys can be written to a temporary table or file, and a followup 
> conditional task can be used to perform the join on those keys.
> As a first step, JDBM can be used for those keys

-- 
This message is automatically generated by JIRA.
-
You can reply to this email to add a comment to the issue online.

Reply via email to