hive实践_01

本地一份包含有中文的文本文件在上传到hive前，需要先转化为UTF-8格式，否则会出现乱码。(notepad++ 格式>>>转化UTF-8编码格式)

 

--------------------------------------------------------------------------------------------------不带分区----------------------------------------------------------------------------------------------------------

不带分区的表创建：(换行符 做行分割，制表符 做列分割) 

drop table if exists SCAN.ZTO_SCAN_COME_TEST_WL; 

 create table SCAN.ZTO_SCAN_COME_TEST_WL

(

  USERNAME      string,

  USERID  string

)

ROW FORMAT DELIMITED  FIELDS TERMINATED BY ' '  LINES TERMINATED BY ' '  STORED AS TEXTFILE;

 

删除表数据： insert overwrite table ZTO_SCAN_COME_TEST_WL  select * from ZTO_SCAN_COME_TEST_WL where 1=0;

 

全量导入数据方式一：覆盖

insert overwrite table ZTO_SCAN_COME_TEST_WL

select '李磊'  as username,'1006' as userid from ZTO_SCAN_COME_MANSUM  limit 1

 

全量导入数据方式二：覆盖

 

load data inpath '/user/hive/HQL_SOURCE/date_test_one.txt'

overwrite into  table ZTO_SCAN_COME_TEST_WL

 

增量导入数据方式一：添加

insert into table ZTO_SCAN_COME_TEST_WL

select '李磊1'  as username,'1007' as userid from ZTO_SCAN_COME_TEST_WL  limit 1

 

增量导入数据方式二：添加

 

load data inpath '/user/hive/HQL_SOURCE/date_test_one.txt'

into  table ZTO_SCAN_COME_TEST_WL

 

 

--------------------------------------------------------------------------------------------------带分区----------------------------------------------------------------------------------------------------------------

带分区的表创建：(换行符 做行分割，制表符 做列分割) 

drop table if exists SCAN.ZTO_SCAN_COME_TEST_WL; 

create table SCAN.ZTO_SCAN_COME_TEST_WL

(

  USERNAME      string,

  USERID  string

)

partitioned by(ds1 string,ds2 string)

ROW FORMAT DELIMITED  FIELDS TERMINATED BY ' '  LINES TERMINATED BY ' '  STORED AS TEXTFILE; 

 

删除某个分区数据：

ALTER TABLE ZTO_SCAN_COME_TEST_WL DROP IF EXISTS PARTITION (ds2='南京');

ALTER TABLE ZTO_SCAN_COME_TEST_WL DROP IF EXISTS PARTITION (ds1='20160902');

alter table ZTO_SCAN_COME_TEST_WL drop if  exists partition(ds1='20160902',ds2='北京');

 

注意：

1、带有分区的表数据插入时，必须指定到每个分区，一个不能少

2、带有分区的表数据删除时，至少指定一个分区即可

 

全量导入数据方式一：覆盖

insert overwrite table ZTO_SCAN_COME_TEST_WL  partition (ds1='20160901',ds2='上海') 

select '李磊'  as username,'1006' as userid from ZTO_SCAN_COME_MANSUM  limit 1;

 

insert overwrite table ZTO_SCAN_COME_TEST_WL  partition (ds1='20160901',ds2='南京') 

select '李磊'  as username,'1007' as userid from ZTO_SCAN_COME_MANSUM  limit 1;

 

全量导入数据方式二：覆盖

load data inpath '/user/hive/HQL_SOURCE/date_test_one.txt'

overwrite into  table ZTO_SCAN_COME_TEST_WL partition (ds1='20160902',ds2='南京') 

 

load data inpath '/user/hive/HQL_SOURCE/date_test_one.txt'

overwrite into  table ZTO_SCAN_COME_TEST_WL partition (ds1='20160902',ds2='泰州') 

 

增量导入方式一：添加

insert into table ZTO_SCAN_COME_TEST_WL  partition (ds1='20160901',ds2='上海') 

select '李磊'  as username,'1006' as userid from ZTO_SCAN_COME_MANSUM  limit 1;

 

增量导入方式二：添加

load data inpath '/user/hive/HQL_SOURCE/date_test_one.txt'

into  table ZTO_SCAN_COME_TEST_WL partition (ds1='20160902',ds2='南京')