Home | About | Sematext search-lucene.com search-hadoop.com
 Search Hadoop and all its subprojects:

Switch to Threaded View
MapReduce >> mail # dev >> Building and Deploying MRv2


Copy link to this message
-
Re: Building and Deploying MRv2
Tom,

I downgraded maven and also changed from open-jdk to sun-jdk and there
is not any progress. I am using Ubuntu 11.04 and could not find
sun-java5-jdk in the Ubuntu repositories, so I installed sun-java6-jdk.

praveensripati@praveensripati:~$ java -version
java version "1.6.0_24"
Java(TM) SE Runtime Environment (build 1.6.0_24-b07)

praveensripati@praveensripati:~$ mvn -version
Apache Maven 2.2.1 (r801777; 2009-08-07 00:46:01+0530)

Thanks,
Praveen
On Friday 17 June 2011 02:04 AM, Thomas Graves wrote:
> I know at one time maven 3.x didn't work so I've been using maven 2.x.
>
> Well I've never tried using java6 for java5 home but I would think it
> wouldn't work.  I thought it was forrest that required java5. I would
> suggest using java5.
>
> Tom
>
>
> On 6/16/11 12:24 PM, "Praveen Sripati"<[EMAIL PROTECTED]>  wrote:
>
>> Tom,
>>
>>>> Note, it looks like your java5.home is pointing to java6?
>> I have java6 on my laptop and pointed java5.home variable to java6. The
>> hadoop doc says "Java 1.6.x - preferable from Sun". Is this the problem?
>>
>>>> What version of protobufs are you using?
>> I have protobuf 2.4.1.
>>
>>>> What about mvn version?
>> Apache Maven 3.0.3 (r1075438; 2011-02-28 23:01:09+0530)
>>
>>>> So you had both common and hdfs built before doing mapreduce and
>> common built before building hdfs? Or was common failing with the error
>> you mention  below? If you haven't already you might simply try
>> veryclean on everything and go again in order.
>> I tried common first and there were some errors related to fop, but the
>> common jars were created, so I started with hdfs and it was successful.
>> Then I started the yarn build which led to the
>> java_generate_equals_and_hash error.
>>
>> Thanks,
>> Praveen
>>
>>
>> On Thursday 16 June 2011 09:54 PM, Thomas Graves wrote:
>>> Note, it looks like your java5.home is pointing to java6?
>>>
>>> I've never seen this particular error. The java_generate_equals_and_hash
>>> option seems to have been added in protobuf2.4.0. What version of protobufs
>>> are you using?  The instructions say to use atleast 2.4.0a, I'm using 2.4.1
>>> right now.
>>>
>>> You need to define the following  (I use a build.properties file). These are
>>> the version I'm currently using.  All of these are just downloaded from the
>>> corresponding website.  Some links to those can be found here:
>>> http://yahoo.github.com/hadoop-common/installing.html
>>>
>>> java5.home=/home/tgraves/hadoop/jdk1.5.0_22/
>>> forrest.home=/home/tgraves/hadoop/apache-forrest-0.8
>>> ant.home=/home/tgraves/hadoop/apache-ant-1.8.2
>>> xercescroot=/home/tgraves/hadoop/xerces-c-src_2_8_0
>>> eclipse.home=/home/tgraves/hadoop/eclipse
>>> findbugs.home=/home/tgraves/hadoop/findbugs-1.3.9
>>>
>>> I thought this was the same as for trunk but perhaps I'm mistaken.
>>>
>>> What about mvn version?
>>> /home/y/libexec/maven/bin/mvn --version
>>> Apache Maven 2.2.1 (r801777; 2009-08-06 19:16:01+0000)
>>>
>>> So you had both common and hdfs built before doing mapreduce and common
>>> built before building hdfs? Or was common failing with the error you mention
>>> below?   If you haven't already you might simply try veryclean on everything
>>> and go again in order.
>>>
>>> Tom
>>>
>>>
>>> On 6/16/11 8:10 AM, "Praveen Sripati"<[EMAIL PROTECTED]>   wrote:
>>>
>>>> Hi,
>>>>
>>>> The hdfs build was successful after including the -Dforrest.home
>>>> property to the ant command.
>>>>
>>>> ***********
>>>>
>>>> When  I started the mapreduce build to get the below error.
>>>>
>>>> mvn clean install assembly:assembly
>>>>
>>>> Downloaded:
>>>> http://repo1.maven.org/maven2/org/apache/commons/commons-exec/1.0.1/commons-
>>>> ex
>>>> ec-1.0.1.jar
>>>> (49 KB at 24.4 KB/sec)
>>>> yarn_protos.proto:4:8: Option "java_generate_equals_and_hash" unknown.
>>>> [INFO]
>>>>
>>>> [INFO]
>>>> ------------------------------------------------------------------------
>>>> [INFO] Skipping hadoop-mapreduce
>>>> [INFO] This project has been banned from the build due to previous failures.