========================================
1、启动服务
start-all.sh
2、关闭安全模式
hadoop dfsadmin -safemode leave
3、
hive --service hiveserver2 &
========================================
CREATE database moves;
use moves;
create table movw(
movieId int,
name string,
types string,
score string,
years string
)
row format delimited
fields terminated by ',';
load data local inpath "/root/log.csv" overwrite into table movw;
SELECT * FROM movw;
create table suo(
movieId int,
name string,
types string,
score string
)
partitioned by (years string) #分区
clustered by (types) #分桶字段
sorted by (score DESC )#排序 into 4 buckets #分4桶
row format delimited
fields terminated by ',';
set hive.exec.dynamic.partition.mode=nonstrict; #设置动态分区
SET hive.exec.dynamic.partition = true; #开启分区
SET hive.enforce.bucketing = true; #开启分区
INSERT into table suo partition(years) SELECT movieId,name,types,score,years FROM movw cluster by (types);
=================================================================
insert into table moves SELECT movieid,SUBSTR(title, 1, LENGTH(title)-6) as title,substr(title,-5,4) as years,genres[0] as types,score from move;