On Sat, Sep 3, 2011 at 12:38 PM, arun k <arunk...@gmail.com> wrote: > Hi ALL ! > > I am trying to configure hadoop on eclipse for development purpose. > I am interested in seeing and understanding the working of Hadoop code : say > Jobtracker , Tasktracker, Namenode working,etc. > I have some problem with git and mvn on my system(as i am behind a proxy) so > i didn't want use the QuickStart in > http://wiki.apache.org/hadoopEclipseEnvironment<http://lucene.472066.n3.nabble.com/HELP-configuring-hadoop-on-ECLIPSE-td1086829.html#a3306100>
Git, Ant and Maven all support letting you set a web proxy. For git, simply set $http_proxy on your environment to the appropriate host:port For maven, follow this ~/.m2/settings.xml configuration tweak: https://wiki.duraspace.org/display/DSPACE/Set+Maven+Web+Proxy+Server+Settings For ant, follow the -Dhttp.proxyHost and -Dhttp.proxyPort instructions: http://ant.apache.org/manual/proxy.html (Set ANT_OPTS, for perma-soln.) After all this is ready, building the version of Hadoop which has the components you need to look into, is a piece of cake: $ git clone https://github.com/apache/hadoop-common.git hadoop $ cd hadoop $ git checkout branch-0.22 $ cd common; ant compile eclipse; cd .. $ cd hdfs; ant compile eclipse; cd .. $ cd mapreduce; ant compile eclipse; cd.. (Note: I've not run the above, but that's what it looks like you need to do to get to read JT/NN/DN/TT sources. Alternatively, you may checkout one of the 0.20.20x branch, but I can't tell you which one cause the whole strategy has changed there and I find it confusing and wish it were a simple mainline branch-0.20 to point you at.) -- Harsh J
