best way to write a UDF is write few test cases around it with few expected
datasets to capture the errors while developing itself rather than in a
hive session
On Tue, Jun 26, 2012 at 8:06 PM, Jan Dolinár <dolik....@gmail.com> wrote:
> Hi,
>
> Check the hadoop logs of the failed task. My best guess is that there is
> an uncaught exception thrown somewhere in your code. The logs will tell
> where and what caused the problem.
>
> Best regards,
> Jan
>
>
> On Tue, Jun 26, 2012 at 4:20 PM, Yue Guan <pipeha...@gmail.com> wrote:
>
>> Hi, hive users
>>
>> I have the following udf:
>>
>> package com.name.hadoop.hive.udf;
>>
>> import java.util.Set;
>>
>> import org.apache.commons.lang.StringUtils;
>>
>> import org.apache.hadoop.hive.ql.exec.UDF;
>> import org.apache.hadoop.io.Text;
>>
>>
>> public class MyUDF extends UDF {
>>
>> private Map<Long, Set<Long>> aMapping;
>> private final Text result = new Text();
>>
>> public MyUDF() throws Exception {
>> aMapping = someModuleSamePackage.getMapping();
>> }
>>
>> public Text evaluate(final Text o) throws Exception {
>> result.clear();
>>
>> if (o != null) {
>> Long id = new Long(o.toString());
>> Set<Long> ids = aMapping.get(id);
>> if (ids != null) {
>> String resultString = StringUtils.join(ids, ",");
>> result.set(resultString);
>> }
>> }
>>
>> return result;
>> }
>> }
>>
>> However, I always get FAILED: Execution Error, return code 2 from
>> org.apache.hadoop.hive.ql.exec.MapRedTask. Anyone suggests anything to
>> work around this? Thank you in advance.
>>
>
>
--
Nitin Pawar
