我有 2 个配置单元表。
源表有以下列:
correspondence_id | decimal(22,0) | |
| template_id | decimal(18,0) | |
| language_cd | varchar(6) | |
| delivery_channel_cd | varchar(20) | |
| job_id | decimal(18,0) | |
| correspondence_content | string | |
| create_user_id | varchar(40) | |
| create_ts | timestamp | |
| last_updt_user_id | varchar(40) | |
| last_updt_ts | timestamp | |
| data_src_id | decimal(18,0) | |
| src_app_resource_cd | varchar(50)
目的地有以下列:
correspondence_id | decimal(22,0) | |
| template_id | decimal(18,0) | |
| template_cd | varchar(20) | |
| template_type_cd | varchar(40) | |
| category_cd | varchar(20) | |
| language_cd | varchar(6) | |
| delivery_channel_cd | varchar(20) | |
| job_id | decimal(18,0) | |
| correspondence_content | string | |
| create_user_id | varchar(40) | |
| create_ts | timestamp | |
| last_updt_user_id | varchar(40) | |
| last_updt_ts | timestamp | |
| data_src_id | decimal(18,0) | |
| src_app_resource_cd | varchar(50) | |
| part_create_year_num | int | |
| part_create_month_num | int | |
| | NULL | NULL |
| # Partition Information | NULL | NULL |
| # col_name | data_type | comment |
| | NULL | NULL |
| part_create_year_num | int | |
| part_create_month_num | int
我使用以下查询来传输数据:
FROM source_table cc insert overwrite table
destination_table partition
(part_create_year_num=2016, part_create_month_num=9 )
select cc.correspondence_id, cc.template_id, cc.language_cd, cc.delivery_channel_cd, cc.job_id,
cc.correspondence_content, cc.create_user_id, cc.create_ts, cc.last_updt_user_id, cc.last_updt_ts,
cc.data_src_id, cc.src_app_resource_cd
但是当我运行这个查询时，我得到以下错误
rror: Error while compiling statement: FAILED: SemanticException [Error 10044]: Line 1:79 Cannot insert into target table because column number/types are different '9': Table insclause-0 has 15 columns, but query has 12 columns. (state=42000,code=10044)
org.apache.hive.service.cli.HiveSQLException: Error while compiling statement: FAILED: SemanticException [Error 10044]: Line 1:79 Cannot insert into target table because column number/types are different '9': Table insclause-0 has 15 columns, but query has 12 columns.
显然源表和目标表是不同的，但是我怎样才能使这个查询工作，我试过放入占位符值，但也没有用。
最佳答案
目标表似乎有五个额外的列
普通列: 1 个模板_cd |变种(20) | 2 模板类型CD |变种(40) | | 3 类别_cd |变种(20) | |
分区列 4. part_create_year_num |整数 | | 5. part_create_month_num |整数
查询应该是
Insert overwrite table Destination_table partition(part_create_year_num=2016, part_create_month_num=9 ) select
correspondence_id,
template_id,
'' as template_cd,
'' as template_type_cd,
'' as category_cd,
language_cd,
delivery_channel_cd,
job_id,
correspondence_content,
create_user_id,
create_ts,
last_updt_user_id,
last_updt_ts,
data_src_id,
src_app_resource_cd
from source_table
关于sql - Hive Insert 从一个表覆盖到另一个具有不同列数的表，我们在Stack Overflow上找到一个类似的问题： https://stackoverflow.com/questions/36703948/