hive改表结构的两个坑

坑一:改变字段类型后更新数据不成功

 

关于hive插入数据的一个小坑,今天插入一个表中数据,插入时写的是常数,比如0.01 ,表中的字段也是DECIMAL(5,2)

按照常理插入的应该是0.01,但是插入后查询是0,为甚!

 

就分析呀,看语句没问题啊,上网查,上hive官网查,呀~ 发现了原因哦

https://cwiki.apache.org/confluence/display/Hive/LanguageManual+Types#LanguageManualTypes-Decimals

 

发现在插入分区表时会出现这种情况,此时需要对之前的分区处理下~:

 

那就测试一下 按照官网的说法:

先建表:

CREATE TABLE `tb_dw_cu_mimsg_fbi_servdt_day1`(

`mimsg_serv` int COMMENT '微信服务量')

PARTITIONED BY (

`statis_date` varchar(8))

ROW FORMAT DELIMITED

FIELDS TERMINATED BY '|' ;

 

然后插入数据:

insert overwrite table tb_dw_cu_mimsg_fbi_servdt_day1 partition (statis_date=20160501) values(1.02);

 

然后查询:

hive> select * from tb_dw_cu_mimsg_fbi_servdt_day1;

OK

1 20160501

 

发现结果跟想象中的一样~

然后修改表字段:

ALTER TABLE tb_dw_cu_mimsg_fbi_servdt_day1 REPLACE COLUMNS (mimsg_serv DECIMAL(5,2))

 

然后再次插入数据:

insert overwrite table tb_dw_cu_mimsg_fbi_servdt_day1 partition (statis_date=20160501) values(1.02);

 

查询:

hive> select * from tb_dw_cu_mimsg_fbi_servdt_day1;

OK

1 20160501

 

发现有问题啦!

那么按照官网处理:

Determine what precision/scale you would like to set for the decimal column in the table.

For each decimal column in the table, update the column definition to the desired precision/scale using theALTER TABLEcommand:

ALTERTABLEfoo CHANGECOLUMNdec_column_name dec_column_nameDECIMAL(38,18);

If the table is not a partitioned table, then you are done. If the table has partitions, then go on to step 3.

If the table is a partitioned table, then find the list of partitions for the table:

SHOW PARTITIONS foo;

 

ds=2008-04-08/hr=11

ds=2008-04-08/hr=12

...

Each existing partition in the table must also have its DECIMAL column changed to add the desired precision/scale.

This can be done with a singleALTER TABLE CHANGE COLUMNby using dynamic partitioning (available for ALTER TABLE CHANGE COLUMN in Hive 0.14 or later, withHIVE-8411):

SEThive.exec.dynamic.partition =true;

 

-- hive.exec.dynamic.partition needs to be set to true to enable dynamic partitioning with ALTER PARTITION

-- This will alter all existing partitions of the table - be sure you know what you are doing!

ALTERTABLEfoo PARTITION (ds, hr) CHANGECOLUMNdec_column_name dec_column_nameDECIMAL(38,18);

 

这里对表作如下处理:

ALTER TABLE tb_dw_cu_mimsg_fbi_servdt_day1 PARTITION (statis_date) CHANGE COLUMN mimsg_serv mimsg_serv DECIMAL(5,2);

 

 

再次插入数据:

insert overwrite table tb_dw_cu_mimsg_fbi_servdt_day1 partition (statis_date=20160501) values(1.02);

 

然后查询:

hive> select * from tb_dw_cu_mimsg_fbi_servdt_day1;

OK

1.02 20160501

Time taken: 0.066 seconds, Fetched: 1 row(s)

 

发现结果跟想象中的一样了。这个坑算是过去了~

 

坑二:增加字段后更新数据不成功

还是上面的例子那张表,再增加一个字段:

alter table tb_dw_cu_mimsg_fbi_servdt_day1 add COLUMNS (prov_code varchar(5))

 

然后查询

hive> select * from tb_dw_cu_mimsg_fbi_servdt_day1;

OK

1.02 NULL 20160501

Time taken: 0.082 seconds, Fetched: 1 row(s)

 

发现新增的字段默认的值是NULL,现在我重新覆盖一下元数据,给增加的字段一个值:

insert overwrite table tb_dw_cu_mimsg_fbi_servdt_day1 partition(statis_date=20160501) values (2.01,0371);

 

然后查询:

hive> select * from tb_dw_cu_mimsg_fbi_servdt_day1;

OK

2.01 NULL 20160501

发现还是NULL

不是我们想象的结果,查看一下官方文档说明发现了问题的所在:

https://cwiki.apache.org/confluence/display/Hive/LanguageManual+DDL#LanguageManualDDL-Add/ReplaceColumns

ALTER TABLE ADD or REPLACE COLUMNS CASCADE will override the table partition's column metadata regardless of the table or partition's protection mode. Use with discretion.

 

那就这样处理:

alter table tb_dw_cu_mimsg_fbi_servdt_day1 replace COLUMNS (mimsg_serv decimal(5,2),prov_code varchar(5)) CASCADE;

 

直接查询,发现数据显示的数据已经发生了变化了~

hive> select * from tb_dw_cu_mimsg_fbi_servdt_day1;

OK

2.01 0371 20160501

 

综上发现,我们是按照Oracle这样的标准在考虑H-SQL,但是通过阅读官方文档发现二者之间还是有很大不同的,通过这次踩坑发现,仔细阅读官方文档的重要性!!!

猜你喜欢

转载自windyqin.iteye.com/blog/2331766