[ 
https://issues.apache.org/jira/browse/HIVE-7999?page=com.atlassian.jira.plugin.system.issuetabpanels:all-tabpanel
 ]

Chu Tong updated HIVE-7999:
---------------------------
    Description: 
When it comes to selecting * on a Hive table with large number of partitions, a 
query like this:

SELECT PARTITIONS.PART_ID, SDS.SD_ID, SDS.CD_ID, SERDES.SERDE_ID, 
PARTITIONS.CREATE_TIME, ... SERDES.SLIB from PARTITIONS LEFT OUTER JOIN SDS ON 
PARTITIONS.SD_ID = SDS.SD_ID LEFT OUTER JOIN SERDES ON SDS.SERDE_ID = 
SERDES.SERDE_ID where PART_ID in (1,2,3,4 ... 10000000 ...

is generated and executed on metastore, however, due the query lists all the 
partitions in it, SQL server is unable to compile/execute such a long query and 
causing the whole query to fail.


  was:
When it comes to selecting * on a Hive table with large number of partitions, a 
query like this:

SELECT PARTITIONS.PART_ID, SDS.SD_ID, SDS.CD_ID, SERDES.SERDE_ID, 
PARTITIONS.CREATE_TIME, ... SERDES.SLIB from PARTITIONS LEFT OUTER JOIN SDS ON 
PARTITIONS.SD_ID = SDS.SD_ID LEFT OUTER JOIN SERDES ON SDS.SERDE_ID = 
SERDES.SERDE_ID where PART_ID in (1,2,3,4 ... 10000000 ...

is generated and executed on metastore, however, due the query lists all the 
partitions in it, SQL DB is unable to compile/execute such a long query and 
causing the whole query to fail.



> Hive metastore query too long when select * on table with large number of 
> partitions
> -------------------------------------------------------------------------------------
>
>                 Key: HIVE-7999
>                 URL: https://issues.apache.org/jira/browse/HIVE-7999
>             Project: Hive
>          Issue Type: Bug
>          Components: Metastore
>            Reporter: Chu Tong
>
> When it comes to selecting * on a Hive table with large number of partitions, 
> a query like this:
> SELECT PARTITIONS.PART_ID, SDS.SD_ID, SDS.CD_ID, SERDES.SERDE_ID, 
> PARTITIONS.CREATE_TIME, ... SERDES.SLIB from PARTITIONS LEFT OUTER JOIN SDS 
> ON PARTITIONS.SD_ID = SDS.SD_ID LEFT OUTER JOIN SERDES ON SDS.SERDE_ID = 
> SERDES.SERDE_ID where PART_ID in (1,2,3,4 ... 10000000 ...
> is generated and executed on metastore, however, due the query lists all the 
> partitions in it, SQL server is unable to compile/execute such a long query 
> and causing the whole query to fail.



--
This message was sent by Atlassian JIRA
(v6.3.4#6332)

Reply via email to