Sergey Shelukhin created HIVE-6190:
--------------------------------------
Summary: redundant columns in metastore schema for stats
Key: HIVE-6190
URL: https://issues.apache.org/jira/browse/HIVE-6190
Project: Hive
Issue Type: Bug
Components: Metastore
Reporter: Sergey Shelukhin
Assignee: Sergey Shelukhin
Priority: Minor
package.jdo has:
{noformat}
<field name ="dbName">
<column name="DB_NAME" length="128" jdbc-type="VARCHAR"
allows-null="false"/>
</field>
<field name="tableName">
<column name="TABLE_NAME" length="128" jdbc-type="VARCHAR"
allows-null="false"/>
</field>
<field name="partitionName">
<column name="PARTITION_NAME" length="767" jdbc-type="VARCHAR"
allows-null="false"/>
</field>
<field name="partition">
<column name="PART_ID"/>
</field>
{noformat}
Only PART_ID is enough, the other fields are unnecessary and may potentially
cause bugs; similarly for table stats. One could argue that they were intended
for perf (denormalization), but stats retrieval currently is very slow on much
deeper level so it's not really justified.
--
This message was sent by Atlassian JIRA
(v6.1.5#6160)
