wiki:jazz/DRBL_Hadoop

Deploy Hadoop to PC Classroom using DRBL

  • java is required for Hadoop, so you need to install java runtime or jdk first. 
    ~$ echo "deb http://free.nchc.org.tw/debian/ etch non-free" > /tmp/etch-non-free.list
~$ sudo mv /tmp/etch-non-free.list /etc/apt/sources.list.d/.
~$ sudo apt-get update
~$ sudo apt-get install sun-java5-jdk
  • download Hadoop 0.18.2 
    ~$ wget http://ftp.twaren.net/Unix/Web/apache/hadoop/core/hadoop-0.18.2/hadoop-0.18.2.tar.gz
~$ tar zxvf hadoop-0.18.2.tar.gz
  • setup JAVA_HOME environment variable 
    ~$ echo "export JAVA_HOME=/usr/lib/jvm/java-1.5.0-sun" >> ~/.bash_profile
~$ source ~/.bash_profile
  • edit hadoop-0.18.2/conf/hadoop-env.sh

    • hadoop-0.18.2/conf/hadoop-env.sh

      old new  
      66# remote nodes.
      77
      88# The java implementation to use.  Required.
      9 # export JAVA_HOME=/usr/lib/j2sdk1.5-sun
       9export JAVA_HOME=/usr/lib/jvm/java-1.5.0-sun
       10export HADOOP_HOME=/home/jazz/hadoop-0.18.2
       11export HADOOP_CONF_DIR=$HADOOP_HOME/conf
      1012
      1113# Extra Java CLASSPATH elements.  Optional.
      1214# export HADOOP_CLASSPATH=
  • here is current DRBL setup 
    ******************************************************
          NIC    NIC IP                    Clients
+------------------------------+
|         DRBL SERVER          |
|                              |
|    +-- [eth0] X.X.X.X        +- to WAN
|                              |
|    +-- [eth1] 192.168.61.254 +- to clients group 1 [ 16 clients, their IP
|                              |             from 192.168.61.1 - 192.168.61.16]
+------------------------------+
******************************************************
Total clients: 16
******************************************************
  • Hadoop will use ssh connections for internal connection, thus we have to do SSH key exchange. 
    ~$ ssh-keygen
~$ cp .ssh/id_rsa.pub .ssh/authorized_keys
~$ sudo apt-get install dsh
~$ mkdir -p .dsh
~$ for ((i=1;i<=16;i++)); do echo "192.168.61.$i" >> .dsh/machines.list; done
  • edit hadoop-0.18.2/conf/hadoop-site.xml

    • hadoop-0.18.2/conf/hadoop-site.xml

      old new  
      44<!-- Put site-specific property overrides in this file. -->
      55
      66<configuration>
      7 
       7  <property>
       8    <name>fs.default.name</name>
       9    <value>hdfs://192.168.61.254:9000/</value>
       10    <description>
       11      The name of the default file system. Either the literal string
       12      "local" or a host:port for NDFS.
       13    </description>
       14  </property>
       15  <property>
       16    <name>mapred.job.tracker</name>
       17    <value>192.168.61.254:9001</value>
       18    <description>
       19      The host and port that the MapReduce job tracker runs at. If
       20      "local", then jobs are run in-process as a single map and
       21      reduce task.
       22    </description>
       23  </property>
      824</configuration>
  • edit /etc/rc.local for DRBL Server as Hadoop namenode

    • /etc/rc.local

      old new  
      1111#
      1212# By default this script does nothing.
      1313
       14echo 3 > /proc/sys/vm/drop_caches
       15/home/jazz/hadoop-0.18.2/bin/hadoop namenode -format
       16/home/jazz/hadoop-0.18.2/bin/hadoop-daemon.sh start namenode
       17/home/jazz/hadoop-0.18.2/bin/hadoop-daemon.sh start jobtracker
       18/home/jazz/hadoop-0.18.2/bin/hadoop-daemon.sh start tasktracker
      1419exit 0
  • edit hadoop_datanode for DRBL client as datanode 
    ~$ cat > hadoop_datanode << EOF

    #! /bin/sh
set -e

# /etc/init.d/hadoop_datanode: start and stop Hadoop DFS datanode for DRBL Client

export PATH="${PATH:+$PATH:}/usr/sbin:/sbin"

case "\$1" in
  start)
        echo -n "starting datanode:"
        /home/jazz/hadoop-0.18.2/bin/hadoop-daemon.sh start datanode
        echo "[OK]"
        ;;
  stop)
        echo -n "stoping datanode:"
        /home/jazz/hadoop-0.18.2/bin/hadoop-daemon.sh stop datanode
        echo "[OK]"
        ;;

  *)
        echo "Usage: /etc/init.d/hadoop_datanode {start|stop}"
        exit 1
esac

exit 0
EOF

    ~$ chmod a+x hadoop_datanode
~$ sudo /opt/drbl/sbin/drbl-cp-host hadoop_datanode /etc/init.d/
~$ sudo /opt/drbl/bin/drbl-doit update-rc.d hadoop_datanode defaults 99
  • shutdown DRBL clients
  • reboot DRBL server
  • use "Wake on LAN" for DRBL clients
  • browse ​http://192.168.61.254:50070 for DFS status

Known Issue

  • 如果要讓 DRBL Clients 都掛起硬碟來存資料，hadoop-site.xml 必須加入 
      <property>
    <name>dfs.data.dir</name>
    <value>預採用的路徑(Ex:/home/jazz/hadoop)</value>
    <final>true</final>
  </property>
  • 記得 namenode 要 format 否則會無法連接上 namenode
  • 有掛載硬碟時，當 datanode 第二次啟動，會出現錯誤訊息 
    2009-03-21 01:03:49,814 INFO org.apache.hadoop.dfs.Storage: Cannot lock storage /home/jazz/hadoop. The directory is already locked.
2009-03-21 01:03:49,915 ERROR org.apache.hadoop.dfs.DataNode: java.io.IOException: Cannot lock storage /home/jazz/hadoop. The directory is already locked.
        at org.apache.hadoop.dfs.Storage$StorageDirectory.lock(Storage.java:447)
        at org.apache.hadoop.dfs.Storage$StorageDirectory.analyzeStorage(Storage.java:300)
        at org.apache.hadoop.dfs.DataStorage.recoverTransitionRead(DataStorage.java:105)
        at org.apache.hadoop.dfs.DataNode.startDataNode(DataNode.java:306)
        at org.apache.hadoop.dfs.DataNode.<init>(DataNode.java:223)
        at org.apache.hadoop.dfs.DataNode.makeInstance(DataNode.java:3071)
        at org.apache.hadoop.dfs.DataNode.instantiateDataNode(DataNode.java:3026)
        at org.apache.hadoop.dfs.DataNode.createDataNode(DataNode.java:3034)
        at org.apache.hadoop.dfs.DataNode.main(DataNode.java:3156)

2009-03-21 01:03:49,915 INFO org.apache.hadoop.dfs.DataNode: SHUTDOWN_MSG:
/************************************************************
SHUTDOWN_MSG: Shutting down DataNode at pc144/192.168.100.44
************************************************************/
Last modified 16 years ago Last modified on Mar 20, 2009, 6:24:14 PM