Hudi表通过spark sql添加字段后，用Hive Sql查询原有数据发生列错位问题

[liucongjy]

Tips before filing an issue

• Have you gone through our FAQs?

• Join the mailing list to engage in conversations and get faster support at dev-subscribe@hudi.apache.org.

• If you have triaged this as a bug, then file an issue directly.

Describe the problem you faced

A clear and concise description of the problem.

To Reproduce

Steps to reproduce the behavior:

1. create table dll
   CREATE TABLE zdyj_ehr.ehr_etbj_csyxzm (
   wdsyh STRING ,
   hzjcsyh STRING ,
   jgdm STRING ,
   jgmc STRING ,
   csyxzmbh STRING ,
   xsexm STRING ,
   jlsj TIMESTAMP)
   using hudi location '/user/hive/warehouse/ws.db/ehr_etbj_csyxzm' TBLPROPERTIES(type='cow' ,primaryKey='wdsyh','hoodie.datasource.write.recordkey.field'='wdsyh' ,preCombineField='gxsj' ,'hoodie.table.partition.fields'='jlsj',
   'hoodie.table.keygenerator.class' = 'org.apache.hudi.keygen.TimestampBasedKeyGenerator',
   'hoodie.datasource.write.keygenerator.class' = 'org.apache.hudi.keygen.TimestampBasedKeyGenerator',
   'hoodie.keygen.timebased.timestamp.type' = 'DATE_STRING',
   'hoodie.keygen.timebased.timezone' = 'GMT+8:00',
   'hoodie.keygen.timebased.input.dateformat' = 'yyyy-MM-dd hh:mm:ss',
   'hoodie.keygen.timebased.output.dateformat' = 'yyyyMM',
   'hoodie.schema.on.read.enable' = 'true',
   'hoodie.datasource.write.keygenerator.consistent.logical.timestamp.enabled' = 'true')
   partitioned by(jlsj);

2.using spark sql add data to ehr_etbj_csyxzm
insert into ehr_etbj_csyxzm values('t001','h001','j01','zhongyiyuan','ttt','tom',current_timestamp());

3.query data using hive shell
select * from ehr_etbj_csyxzm
jlsj field value is '2018-08-14 00:00:00'

4.using spark sql alter table add new column
alter table ehr_etbj_csyxzm add columns(ext1 string comment '扩展字段1');

5.query data using hive shell
select * from ehr_etbj_csyxzm
新加字段ext1 值为'2018-08-14 00:00:00'，原有字段jlsj的值为空
正确结果应该是 字段ext1的值为空，jlsj值为'2018-08-14 00:00:00'

Expected behavior

A clear and concise description of what you expected to happen.

Environment Description

• Hudi version : 0.15

• Spark version : 3.3.0

• Hive version : 2.3.9

• Hadoop version : 3.0

• Storage (HDFS/S3/GCS..) :hdfs

• Running on Docker? (yes/no) :no

Additional context

Add any other context about the problem here.

Stacktrace

Add the stacktrace of the error.

[cshuo]

@liucongjy Hi, what about the result of other queries besides select *, such as select wdsyh, ext1, jlsj from zdyj_ehr.ehr_etbj_csyxzm

[rangareddy]

Hi @liucongjy,

After altering the table using Spark, you may not be able to query the Hive table data, and you might encounter a following exception.

hive> SELECT * FROM ehr_etbj_csyxzm;
OK
Failed with exception java.io.IOException:org.apache.hudi.exception.HoodieException: Field ext1 not found in log schema. Query cannot proceed! Derived Schema Fields: [_hoodie_commit_time, _hoodie_partition_path, jgmc, _hoodie_record_key, jgdm, csyxzmbh, jlsj, wdsyh, _hoodie_commit_seqno, hzjcsyh, _hoodie_file_name, xsexm]
Time taken: 23.787 seconds

The above issue is fixed in Hudi 1.0.1 and upstream jira is HUDI-8880