Hive知识点（六） --Hive表的导入|导出

load data [local] inpath ‘filepath’ [overwrite] into table 表名 [partition(partiton=value,…)]

1.导入Hive表

1.1加载本地（Linux系统）文件到Hive表

load data local inpath '/home/hadoop/tmp/data.txt' into table hive2;

1.2加载HDFS文件到Hive表

hdfs dfs -mkidr  /user/data      #在HDFS中创建目录
hdfs dfs -put '/home/hadoop/tmp/data.txt' /user/data    #上传到HDFS中
load data inpath '/user/data/data.txt' into table hive2;   #加载到Hive表中
select * from hive2;    #查看

1.3加载HDFS文件并覆盖Hive表之前的数据

load data local inpath '/tmp/data.txt' overwrite into table hive2;

1.4创建Hive表并通过insert加载

create table hive3 like hive2;
insert into table hive3 select * from hive2;    #这句会执行MapReduce

2.导出Hive表

2.1Hive表===》Linuxs本地

1.hive的shell窗口进行导出

insert overwrite local '/home/hadoop/tmp/out1' select * from hive1.hive1;    #默认格式导出
insert overwrite local '/home/hadoop/tmp/out2' row format delimited fields terminated by '\t' collection items terminated by '\n';    #指定数据数据的格式

2.linuxs命令行窗口进行导出

cd app/hive
bin/hive -e "select * from hive1.hive1;" > /home/hadoop/tmp/out3

2.2Hive表===》HDFS

insert overwrite directory '/user/hive/tmp/out' row format delimited fields terminated by '\t' collection items terminated by '\n' select * from hive1.hive1;