You are viewing a plain text version of this content. The canonical link for it is here.
Posted to dev@hive.apache.org by "Deepak Jaiswal (JIRA)" <ji...@apache.org> on 2017/10/19 20:14:01 UTC

[jira] [Created] (HIVE-17851) Bucket Map Join : Pick correct number of buckets

Deepak Jaiswal created HIVE-17851:
-------------------------------------

             Summary: Bucket Map Join : Pick correct number of buckets
                 Key: HIVE-17851
                 URL: https://issues.apache.org/jira/browse/HIVE-17851
             Project: Hive
          Issue Type: Bug
            Reporter: Deepak Jaiswal
            Assignee: Deepak Jaiswal


CREATE TABLE tab_part (key int, value string) PARTITIONED BY(ds STRING) CLUSTERED BY (key) INTO 2 BUCKETS STORED AS TEXTFILE;
CREATE TABLE tab(key int, value string) PARTITIONED BY(ds STRING) CLUSTERED BY (key) INTO 4 BUCKETS STORED AS TEXTFILE;

select a.key, a.value, b.value
from tab a join tab_part b on a.key = b.key;

In above case, if tab_part is bigger then it should be the streaming side and the smaller side should create two hash tables. However, currently, it blindly picks 4 as number of buckets as it is the maximum number of buckets among all the tables involved in the join and create 4 hash tables.



--
This message was sent by Atlassian JIRA
(v6.4.14#64029)