I am currently using Hadoop 0.20.2 because Mahout does not support 0.21.0 yet. To get started, download maven2 and subversion form the ubuntu repository using the following:
sudo aptitude install maven2
sudo aptitude install subversion
svn co http://svn.apache.org/repos/asf/mahout/trunk
Next, cd into the directory and install it using maven. This will download all the necessary dependencies and build the latest jar.
mvn install
Once you have the core files installed, make sure your environment variables are set correctly. For me, this is:
export JAVA_HOME=/usr/lib/jvm/java-6-sun/
export HADOOP_CONF_DIR=/home/hadoop/hadoop-0.20.2/conf
export HADOOP_HOME=/home/hadoop/hadoop-0.20.2
export MAHOUT_CONF_DIR=/home/hadoop/mahout/src/conf
Now my current home directory looks like this:
/home/hadoop/mahout
/home/hadoop/hadoop-0.20.2
At this point Mahout should be ready to run. Switch into the directory:
/home/hadoop/mahout/bin/
and execute the mahout shell script. If all goes well, you should see all of the available functions.
Hai Vincent Thanx 4 the tutorial.
ReplyDeleteI am getting the following error on running mahout. My .bashrc file is provided
hadoop@master:~/trunk/bin$ ./mahout
MAHOUT_LOCAL is set, so we don't add HADOOP_CONF_DIR to classpath.
hadoop binary is not in PATH,HADOOP_HOME/bin,HADOOP_PREFIX/bin, running locally
Error occurred during initialization of VM
Could not reserve enough space for object heap
Could not create the Java virtual machine.
export HADOOP_HOME=/home/hadoop/hadoop/bin/
export JAVA_HOME=/usr/lib/jvm/java-6-openjdk/
export HADOOP_CONF_DIR=/home/hadoop/hadoop/conf/
export MAHOUT_CONF_DIR=/home/hadoop/trunk/src/conf/
export MAHOUT_LOCAL=/home/hadoop/trunk/bin/
I have same problem, did you solve it?
ReplyDelete