Reply
Highlighted
New Contributor
Posts: 1
Registered: ‎07-01-2014

Data is inserted as Null for few columns in Hive Table using Parquet Format

Hi ,

 
I am evaluating the Kite SDK for our project  .
As a start i was looking into the examples available in Kite Data Module.
 
My requirement is I want to load the data in Hive table in Parquet Format.
For this I slightly modified the example CreateHCatalogUserDatasetGeneric.java.
 
While creating the DatasetDescriptor , i associated the format with it also like this.
 
DatasetDescriptor descriptor = new DatasetDescriptor.Builder().schemaUri("resource:user.avsc").format(Formats.PARQUET).build();
 
I am using the same avro schema file available in the example.
 
Now when I executed this example , it is inserting the 100 rows(i.e expected) in the table but the value for creationDate and favoriteColor is inserting as null.
 
I tried to create different schema also but the behavior is same , few of the columns are always inserting as null.
 
Kite Version :0.13.0
hadoop-client : 2.3.0-cdh5.0.0
hive-hcatalog-core : 0.12.0-cdh5.0.0
CDH Version : 5.0.0
 
Any suggestions ?
Announcements
The Kite SDK is a collection of docs, sample code, APIs, and tools to make Hadoop application development faster. Learn more at http://kitesdk.org.