[TOC] ## 語法結構 利用查詢語句，將查詢結果插入新的表 ~~~ INSERT OVERWRITE [INTO] TABLE tablename1 [PARTITION (partcol1=val1, partcol2=val2 ...)] select_statement1 FROM from_statement ~~~ ## 覆蓋插入 ~~~ insert overwrite table t_1 partition(city='japan') select id,name,age from stu_partition; ~~~ ## 數據插入本地 要寫local ~~~ insert overwrite local directory '/root/hivedata/data' select * from t_1; ~~~ 格式化導入到本地 ~~~ insert overwrite local directory '/root/hivedata/data' row format delimited fields terminated by '\t' select * from t_1; ~~~ ## 數據插入hdfs ~~~ insert overwrite directory '/root/hivedata/data' select * from t_1; ~~~ ## 插入一條數據 ~~~ INSERT INTO TABLE VALUES(XX,YY,ZZ); ~~~ 插入數據到分區 ~~~ insert into table student partition(month='201702') values(1, 'jdxia'); ~~~ ## Multi Inserts多重插入 ~~~ FROM from_statement INSERT OVERWRITE TABLE tablename1 [PARTITION (partcol1=val1, partcol2=val2 ...)] select statement1 [INSERT OVERWRITE TABLE tablename2 [PARTITION ...] Select statement2] ... ~~~ 意思就是里面的多行sql的數據都來自第一個from_statement這個表  Dynamic partition inserts動態分區插入: 不會手動指定分區,會根據查詢的值來指定分區 partition(寫列名) ~~~ INSERT OVERWRITE TABLE tablename PARTITION (partcol1[=val1], partcol2[=val2] ...) select_statement FROM from_statement ~~~ 使用了非嚴格模式 ~~~ set hive.exec.dynamic.partition.mode=nonstrict ~~~  這邊分區我們沒有指定,是查詢的時候,插入動態插入的