[
https://issues.apache.org/jira/browse/HIVE-2682?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanel&focusedCommentId=13176429#comment-13176429
]
Phabricator commented on HIVE-2682:
-----------------------------------
rajat has commented on the revision "HIVE-2682 [jira] Clean-up logs".
INLINE COMMENTS
ql/src/java/org/apache/hadoop/hive/ql/stats/jdbc/JDBCStatsPublisher.java:128
:P
Then I need to change this in multiple places.
ql/src/java/org/apache/hadoop/hive/ql/exec/JoinOperator.java:112-114 I think
this is related to skew joins. If we touch 1000 rows for a single key, there is
a chance of data being skewed for that key and it needs to be handled
specially. (I am just guessing, though).
REVISION DETAIL
https://reviews.facebook.net/D1035
> Clean-up logs
> -------------
>
> Key: HIVE-2682
> URL: https://issues.apache.org/jira/browse/HIVE-2682
> Project: Hive
> Issue Type: Wish
> Components: Logging
> Reporter: Rajat Goel
> Priority: Trivial
> Attachments: HIVE-2682.D1035.1.patch
>
> Original Estimate: 24h
> Remaining Estimate: 24h
>
> Just wanted to cleanup some logs being printed at wrong loglevel -
> 1. org.apache.hadoop.hive.ql.exec.CommonJoinOperator prints "table 0 has 1000
> rows for join key [...]" as WARNING. Is it really that?
> 2. org.apache.hadoop.hive.ql.exec.GroupByOperator prints "Hash Table
> completed flushed" and "Begin Hash Table flush at close: size = 21" as
> WARNING. It shouldn't be.
> 3. org.apache.hadoop.hive.ql.stats.jdbc.JDBCStatsPublisher prints "Warning.
> Invalid statistic." which looks fishy.
--
This message is automatically generated by JIRA.
If you think it was sent incorrectly, please contact your JIRA administrators:
https://issues.apache.org/jira/secure/ContactAdministrators!default.jspa
For more information on JIRA, see: http://www.atlassian.com/software/jira
