The CLUSTERED BY and SORTED BY creation commands do not affect how data is inserted into a table – only how it is read. This means that users must be careful to insert data correctly by specifying the number of reducers to be equal to the number of buckets, and using CLUSTER BY and SORT BY commands in their query.
In general, distributing rows based on the hash will give you a even distribution(均匀分布) in the buckets.
set mapred.reduce.tasks = 3;
set hive.enforce.bucketing = true;
CREATE TABLE user_info_bucketed(user_id BIGINT, firstname STRING, lastname STRING)
COMMENT 'A bucketed copy of user_info'
PARTITIONED BY(ds STRING)
CLUSTERED BY(user_id) INTO 3 BUCKETS;
INSERT into TABLE user_info_bucketed
PARTITION (ds='2015-07-25')
values
(100,'python','postgresql'), (101,'python','postgresql'), (102,'python','postgresql'), (103,'python','postgresql'), (104,'python','postgresql'), (105,'python','postgresql'), (106,'python','postgresql'), (107,'python','postgresql'), (108,'python','postgresql'), (109,'python','postgresql'), (111,'python','postgresql'), (112,'python','postgresql'), (113,'python','postgresql'), (114,'python','postgresql'), (115,'python','postgresql'), (116,'python','postgresql'), (117,'python','postgresql'), (118,'python','postgresql'), (119,'python','postgresql'), (120,'python','postgresql'), (121,'python','postgresql'), (122,'python','postgresql'), (2000,'R','Oracle'), (2001,'R','Oracle'), (2002,'R','Oracle'), (2003,'R','Oracle'), (2004,'R','Oracle'), (2005,'R','Oracle'), (2006,'R','Oracle'), (2007,'R','Oracle'), (2008,'R','Oracle'), (2009,'R','Oracle'), (2010,'R','Oracle'), (2011,'R','Oracle'), (2012,'R','Oracle'), (2013,'R','Oracle'), (2014,'R','Oracle'), (2015,'R','Oracle'), (2016,'R','Oracle'), (2017,'R','Oracle'), (2018,'R','Oracle'), (2019,'R','Oracle'), (2020,'R','Oracle'), (2030,'R','Oracle'), (2040,'R','Oracle'), (2050,'R','Oracle');
[spark01 ~]$ hadoop fs -ls -R /user/hive/warehouse/test.db/user_info_bucketed
drwxrwxrwx - huai supergroup 0 2015-07-20 22:46 /user/hive/warehouse/test.db/user_info_bucketed/ds=2015-07-25
-rwxrwxrwx 3 huai supergroup 266 2015-07-20 22:46 /user/hive/warehouse/test.db/user_info_bucketed/ds=2015-07-25/000000_0
-rwxrwxrwx 3 huai supergroup 288 2015-07-20 22:46 /user/hive/warehouse/test.db/user_info_bucketed/ds=2015-07-25/000001_0
-rwxrwxrwx 3 huai supergroup 266 2015-07-20 22:46 /user/hive/warehouse/test.db/user_info_bucketed/ds=2015-07-25/000002_0
[spark01 ~]$ hadoop fs -cat /user/hive/warehouse/test.db/user_info_bucketed/ds=2015-07-25/000000_0 |sort
102pythonpostgresql
105pythonpostgresql
108pythonpostgresql
111pythonpostgresql
114pythonpostgresql
117pythonpostgresql
120pythonpostgresql
2001ROracle
2004ROracle
2007ROracle
2010ROracle
2013ROracle
2016ROracle
2019ROracle
2040ROracle
[spark01 ~]$ hadoop fs -cat /user/hive/warehouse/test.db/user_info_bucketed/ds=2015-07-25/000001_0 |sort
100pythonpostgresql
103pythonpostgresql
106pythonpostgresql
109pythonpostgresql
112pythonpostgresql
115pythonpostgresql
118pythonpostgresql
121pythonpostgresql
2002ROracle
2005ROracle
2008ROracle
2011ROracle
2014ROracle
2017ROracle
2020ROracle
2050ROracle
[spark01 ~]$ hadoop fs -cat /user/hive/warehouse/test.db/user_info_bucketed/ds=2015-07-25/000002_0 |sort
101pythonpostgresql
104pythonpostgresql
107pythonpostgresql
113pythonpostgresql
116pythonpostgresql
119pythonpostgresql
122pythonpostgresql
2000ROracle
2003ROracle
2006ROracle
2009ROracle
2012ROracle
2015ROracle
2018ROracle
2030ROracle