You are viewing a plain text version of this content. The canonical link for it is here.
Posted to dev@hive.apache.org by "Bobeff (JIRA)" <ji...@apache.org> on 2016/04/18 12:46:25 UTC
[jira] [Created] (HIVE-13534) xception when trying to access
TIMESTAMP columns into parquet file using hive external table
Bobeff created HIVE-13534:
-----------------------------
Summary: xception when trying to access TIMESTAMP columns into parquet file using hive external table
Key: HIVE-13534
URL: https://issues.apache.org/jira/browse/HIVE-13534
Project: Hive
Issue Type: Bug
Components: File Formats, Hive, Import/Export, JDBC
Reporter: Bobeff
Assignee: Sushanth Sowmyan
Priority: Critical
Imported data was stored from a netezza datasource using a sqoop import command like this
SQL DDL creation script of imported table looks like this
CREATE TABLE "ADMIN"."MIS_AUX_ITR" (
"DDEBVAL" DATE,
"DFINVAL" DATE,
"NAUX" VARCHAR(6),
"CDMNITR" VARCHAR(3),
"CDERIMG" VARCHAR(1),
"DDERIMG" DATE
);
Import sqoop job is the following
sqoop job
--create import-name
-- import
--connect jdbc:netezza://server:port/database
--username user
--password pwd
--table MIS_AUX_ITR
--as-parquetfile
--target-dir hdfs:///prod/ZA/dee/MIS_AUX_ITR
-m 1
After import parquet file schema is the following
> yarn jar /tmp/parquet-tools-1.6.0.jar schema /prod/ZA/dee/MIS_AUX_ITR/2cf3e971-4c2c-408f-bd86-5d3cf3bd4fa5.parquet
message MIS_AUX_ITR {
optional int64 DDEBVAL;
optional int64 DFINVAL;
optional binary NAUX (UTF8);
optional binary CDMNITR (UTF8);
optional binary CDERIMG (UTF8);
optional int64 DDERIMG;
}
In order to access data stored into the parquet file we created the external table below
CREATE EXTERNAL TABLE za_dee.MIS_AUX_ITR
(
`DDEBVAL` DATE,
`DFINVAL` DATE,
`NAUX` VARCHAR(6),
`CDMNITR` VARCHAR(3),
`CDERIMG` VARCHAR(1),
`DDERIMG` DATE
)
COMMENT 'Table DEE MIS_AUX_ITR'
STORED AS PARQUET
LOCATION
'/prod/ZA/dee/MIS_AUX_ITR';
But when we try to list data from external table above we get the following exception
hive> CREATE EXTERNAL TABLE za_dee.MIS_AUX_ITR_V_PPROD
> (
> `DDEBVAL`DATE,
> `DFINVAL`DATE,
> `NAUX`VARCHAR(6),
> `CDMNITR`VARCHAR(3),
> `CDERIMG`VARCHAR(1),
> `DDERIMG`DATE
> )
> COMMENT 'Table DEE MIS_AUX_ITR_V_PROD'
> STORED AS PARQUET
> LOCATION
> '/prod/ZA/dee/MIS_AUX_ITR_V_PPROD';
OK
Time taken: 0.196 seconds
hive> select * from za_dee.MIS_AUX_ITR_V_PPROD limit 100;
OK
SLF4J: Failed to load class "org.slf4j.impl.StaticLoggerBinder".
SLF4J: Defaulting to no-operation (NOP) logger implementation
SLF4J: See http://www.slf4j.org/codes.html#StaticLoggerBinder for further details.
Failed with exception java.io.IOException:org.apache.hadoop.hive.ql.metadata.HiveException: java.lang.ClassCastException: org.apache.hadoop.io.LongWritable cannot be cast to org.apache.hadoop.hive.serde2.io.DateWritable
Time taken: 0.529 seconds
hive>
We also tried with the following external table
CREATE EXTERNAL TABLE za_dee.MIS_AUX_ITR_V_PPROD_BI
(
`DDEBVAL` BIGINT,
`DFINVAL` BIGINT,
`NAUX` VARCHAR(6),
`CDMNITR` VARCHAR(3),
`CDERIMG` VARCHAR(1),
`DDERIMG` BIGINT
)
COMMENT 'Table DEE MIS_AUX_ITR_V_PROD_BI'
STORED AS PARQUET
LOCATION '/prod/ZA/dee/MIS_AUX_ITR_V_PPROD';
Then the “Date” columns are shown as “timestamp” values as below
hive> select DDEBVAL from za_dee.MIS_AUX_ITR_V_PPROD_BI limit 5;
OK
1080770400000
1080770400000
1080770400000
1080770400000
1080770400000
Time taken: 0.081 seconds, Fetched: 5 row(s)
hive>
However “Date” values can be listed by casting as Timestamp
hive> select cast(DDEBVAL as Timestamp) from za_dee.MIS_AUX_ITR_V_PPROD_BI limit 5;
OK
2004-04-01 00:00:00
2004-04-01 00:00:00
2004-04-01 00:00:00
2004-04-01 00:00:00
2004-04-01 00:00:00
Time taken: 0.087 seconds, Fetched: 5 row(s)
hive>
We also have tested with an external table using TIMESTAMP type as shown below
CREATE EXTERNAL TABLE za_dee.MIS_AUX_ITR
(
`DDEBVAL` TIMESTAMP,
`DFINVAL` TIMESTAMP,
`NAUX` VARCHAR(6),
`CDMNITR` VARCHAR(3),
`CDERIMG` VARCHAR(1),
`DDERIMG` TIMESTAMP
)
COMMENT 'Table DEE MIS_AUX_ITR'
STORED AS PARQUET
LOCATION
'/prod/ZA/dee/MIS_AUX_ITR';
But we got the same behavior: an exception when trying to access data from an Oracle DB.
I tried this
CREATE EXTERNAL TABLE za_dee.MIS_AUX_ITR_V_PPROD_TS
(
`DDEBVAL` TIMESTAMP,
`DFINVAL` TIMESTAMP,
`NAUX` VARCHAR(6),
`CDMNITR` VARCHAR(3),
`CDERIMG` VARCHAR(1),
`DDERIMG` TIMESTAMP
)
COMMENT 'Table DEE MIS_AUX_ITR_V_PROD_TS'
STORED AS PARQUET
LOCATION
'/prod/ZA/dee/MIS_AUX_ITR_V_PPROD';
and then i created and launched the sqoop job below
sqoop job --create import-za_dee-MIS_AUX_ITR_V-full-default-import-PPROD -- import
--connect jdbc:netezza:/<server>:<port>/db
--username <user>
--password <password>
--table MIS_AUX_ITR_V
--as-parquetfile
--hive-import
--hive-overwrite
--hive-database za_dee
--hive-table MIS_AUX_ITR_V_PPROD_TS
-m 1
sqoop job --exec import-za_dee-MIS_AUX_ITR_V-full-default-import-PPROD
the raising error is the following
16/04/11 17:15:09 INFO orm.CompilationManager: Writing jar file: /tmp/sqoop-sqoop/compile/3533e18a81a65fb8eb88ec9cef2f4688/codegen_MIS_AUX_ITR_V.jar
16/04/11 17:15:09 WARN manager.NetezzaManager: It looks like you are importing from Netezza.
16/04/11 17:15:09 WARN manager.NetezzaManager: This transfer can be faster! Use the --direct
16/04/11 17:15:09 WARN manager.NetezzaManager: option to exercise a Netezza-specific fast path.
16/04/11 17:15:17 INFO mapreduce.ImportJobBase: Beginning import of MIS_AUX_ITR_V
16/04/11 17:15:17 INFO manager.SqlManager: Executing SQL statement: SELECT t.* FROM "MIS_AUX_ITR_V" AS t WHERE 1=0
16/04/11 17:15:18 INFO manager.SqlManager: Executing SQL statement: SELECT t.* FROM "MIS_AUX_ITR_V" AS t WHERE 1=0
16/04/11 17:15:18 INFO manager.SqlManager: Executing SQL statement: SELECT t.* FROM "MIS_AUX_ITR_V" AS t WHERE 1=0
16/04/11 17:15:19 INFO hive.metastore: Trying to connect to metastore with URI thrift://slhdm007.maif.local:9083
16/04/11 17:15:19 INFO hive.metastore: Connected to metastore.
16/04/11 17:15:19 ERROR tool.ImportTool: Imported Failed: Cannot convert unsupported type: timestamp
--
This message was sent by Atlassian JIRA
(v6.3.4#6332)