hive-user mailing list archives

Site index · List index
Message view « Date » · « Thread »
Top « Date » · « Thread »
From Dmitry Tolpeko <dmtolp...@gmail.com>
Subject Re: Run HPL/SQL on CDH 5.7 with Hive 1.1
Date Wed, 30 Nov 2016 17:51:14 GMT
Looks like a Hive issue executing truncate. Can you please add --trace
option to command line, maybe it shows more details. Did you try to execute
truncate on this table in Hive CLI? Does it work? Is it managed or external
table?

Thanks,
Dmitry

On Wed, Nov 30, 2016 at 5:46 PM, Su Changfeng <changfeng.su@outlook.com>
wrote:

> Hi Dmitry Tolpeko and Hive Experts
>
> My company  wants to migrate some ETL processes written in Oracle PL/SQL
> to Hive or something else.
>
> Because we don't have any component running on our Hadoop cluster
> which supports PL/SQL, wa wanna try  HPL/SQL for our PL/SQL migration.
>
> Our Hadoop is running on CDH express and the version is 5.8.1 with Hive
> 1.1, and the HPL/SQL I used is the latest version 0.3.17 (choose this
> version is just because it is the only version supports "Truncate Table"
> DDL operation, and our ETLs have lot of "Truncate Table" statements.
>
> But unfortunately "Truncate Table" does not work as expected and it runs
> into exception (SQL is as below)
>
> *REATE OR REPLACE Procedure p_M_TJB_BAL_INF()*
>
> *BEGIN*
>
> *  PRINT 'DELETE_TABLES excuting' ;*
>
> * Truncate table TEST_DB.TEST5;*
>
> * INSERT INTO TEST_DB.TEST5 VALUES ('1','C'),('2','D'),*
> * ('3','C'),('4','D');*
>
> * select * from TEST_DB.test5;*
>
> *END;*
>
> *call p_M_TJB_BAL_INF('2016-11-30');*
>
>
> ...
>
> pen connection: jdbc:hive2://localhost:10000 (287 ms)
>
> Starting SQL statement
>
> Unhandled exception in HPL/SQL
>
> java.sql.SQLException: Error while processing statement: FAILED: Execution
> Error, return code 1 from org.apache.hadoop.hive.ql.exec.DDLTask.
> Exception while processing
>
> at org.apache.hive.jdbc.HiveStatement.execute(HiveStatement.java:279)
> ...
>
>
> But after I changing the "Truncate Table" into drop/re-create table DDL,
> it works very well.
>
> *drop table if exists test_db.test5;*
>
> *create table test_db.test5(id int ,names string )*
>
> *clustered by (id) into 2 buckets Stored as textfile;*
>
> Do you have any idea or input? Any comment is appreciated.
>
> --
>
> Thanks & Regards,
>
> Kevin
>
>

Mime
View raw message