/ 中存储网

Nutch相关框架安装使用最佳指南

2013-03-20 00:00:00 来源:中存储
Nutch相关框架安装使用最佳指南
    博客分类: 大数据 nutchhadooplucenesolrgoragangliahbasezookeeperaccumulocassandraavronagiossplunkpighive  一、nutch1.2
    二、nutch1.5.1
    三、nutch2.0
    四、配置SSH
    五、安装Hadoop Cluster(伪分布式运行模式)并运行Nutch
    六、安装Hadoop Cluster(分布式运行模式)并运行Nutch
    七、配置Ganglia监控Hadoop集群和HBase集群
    八、Hadoop配置Snappy压缩
    九、Hadoop配置Lzo压缩
    十、配置zookeeper集群以运行hbase
    十一、配置Hbase集群以运行nutch-2.1(Region Servers会因为内存的问题宕机)
    十二、配置Accumulo集群以运行nutch-2.1(gora存在BUG)
    十三、配置Cassandra 集群以运行nutch-2.1(Cassandra 采用去中心化结构)
    十四、配置MySQL 单机服务器以运行nutch-2.1
    十五、nutch2.1 使用DataFileAvroStore作为数据源
    十六、nutch2.1 使用AvroStore作为数据源
    十七、配置SOLR
    十八、Nagios监控
    十九、配置Splunk
    二十、配置Pig
    二十一、配置Hive
    二十二、配置Hadoop2.x集群


    一、nutch1.2
    步骤和二大同小异,在步骤 5、配置构建路径 中需要多两个操作:在左部Package Explorer的 nutch1.2文件夹上单击右键 > Build Path > Configure Build Path...   >  选中Source选项 > Default output folder:修改nutch1.2/bin为nutch1.2/_bin,在左部Package Explorer的 nutch1.2文件夹下的bin文件夹上单击右键 > Team > 还原
    二中黄色背景部分是版本号的差异,红色部分是1.2版本没有的,绿色部分是不一样的地方,如下:
    1、Add JARs... >  nutch1.2 > lib ,选中所有的.jar文件 > OK
    2、crawl-urlfilter.txt
    3、将crawl -urlfilter.txt.template改名为crawl -urlfilter.txt
    4、修改crawl-urlfilter.txt,将
    # accept hosts in MY.DOMAIN.NAME
    +^http://([a-z0-9]*.)*MY.DOMAIN.NAME/

    # skip everything else
    -.
    5、cd /home/ysc/workspace/nutch1.2
    nutch1.2是一个完整的搜索引擎,nutch1.5.1只是一个爬虫。nutch1.2可以把索引提交给SOLR,也可以直接生成LUCENE索引,nutch1.5.1则只能把索引提交给SOLR:
    1、cd /home/ysc
    2、wget http://mirrors.tuna.tsinghua.edu.cn/apache/tomcat/tomcat-7/v7.0.29/bin/apache-tomcat-7.0.29.tar.gz
    3、tar -xvf apache-tomcat-7.0.29.tar.gz
    4、在左部Package Explorer的 nutch1.2文件夹下的build.xml文件上单击右键 > Run As > Ant Build... > 选中war target > Run
    5、cd /home/ysc/workspace/nutch1.2/build
    6、unzip nutch-1.2.war -d nutch-1.2
    7、cp -r nutch-1.2 /home/ysc/apache-tomcat-7.0.29/webapps
    8、vi /home/ysc/apache-tomcat-7.0.29/webapps/nutch-1.2/WEB-INF/classes/nutch-site.xml
    加入以下配置:
    <property>
      <name>searcher.dir</name>
      <value>/home/ysc/workspace/nutch1.2/data</value>
      <description>
      Path to root of crawl.  This directory is searched (in
      order) for either the file search-servers.txt, containing a list of
      distributed search servers, or the directory "index" containing
      merged indexes, or the directory "segments" containing segment
      indexes.
      </description>
    </property>
    9、vi /home/ysc/apache-tomcat-7.0.29/conf/server.xml

    <Connector port="8080" protocol="HTTP/1.1"
                   connectionTimeout="20000"
                   redirectPort="8443"/>
    改为
    <Connector port="8080" protocol="HTTP/1.1"
                   connectionTimeout="20000"
                   redirectPort="8443" URIEncoding="utf-8"/>

    10、cd /home/ysc/apache-tomcat-7.0.29/bin
    11、./startup.sh
    12、访问:http://localhost:8080/nutch-1.2/

    关于nutch1.2更多的BUG修复及资料,请参看我在CSDN发布的资源:http://download.csdn.net/user/yangshangchuan

    二、nutch1.5.1
    1、下载并解压eclipse(集成开发环境)
    下载地址:http://www.eclipse.org/downloads/,下载Eclipse IDE for Java EE Developers
    2、安装Subclipse插件(SVN客户端)
    插件地址:http://subclipse.tigris.org/update_1.8.x,
    3、安装IvyDE插件(下载依赖Jar)
    插件地址:http://www.apache.org/dist/ant/ivyde/updatesite/
    4、签出代码
    File > New > Project > SVN > 从SVN 检出项目
    创建新的资源库位置 > URL:https://svn.apache.org/repos/asf/nutch/tags/release-1.5.1/ > 选中URL > Finish
    弹出New Project向导,选择Java Project > Next,输入Project name:nutch1.5.1 > Finish
    5、配置构建路径
    在左部Package Explorer的 nutch1.5.1文件夹上单击右键 > Build Path > Configure Build Path...  
    > 选中Source选项 > 选择src > Remove > Add Folder... > 选择src/bin, src/java, src/test 和 src/testresources(对于插件,需要选中src/plugin目录下的每一个插件目录下的src/java , src/test文件夹) > OK
    切换到Libraries选项 >
    Add Class Folder... > 选中nutch1.5.1/conf > OK
    Add JARs... >  需要选中src/plugin目录下的每一个插件目录下的lib目录下的jar文件 > OK
    Add Library... > IvyDE Managed Dependencies > Next > Main > Ivy File > Browse > ivy/ivy.xml > Finish
    切换到Order and Export选项>
    选中conf > Top
    6、执行ANT
    在左部Package Explorer的 nutch1.5.1文件夹下的build.xml文件上单击右键 > Run As > Ant Build
    在左部Package Explorer的 nutch1.5.1文件夹上单击右键 > Refresh
    在左部Package Explorer的 nutch1.5.1文件夹上单击右键 > Build Path > Configure Build Path...   >  选中Libraries选项 > Add Class Folder... >  选中build > OK
    7、修改配置文件nutch-site.xml 和regex-urlfilter.txt
    将nutch-site.xml.template改名为nutch-site.xml
    将regex-urlfilter.txt.template改名为regex-urlfilter.txt
    在左部Package Explorer的 nutch1.5.1文件夹上单击右键 > Refresh
    将如下配置项加入文件nutch-site.xml:
    <property>
      <name>http.agent.name</name>
      <value>nutch</value>
    </property>
    <property>
      <name>http.content.limit</name>
      <value>-1</value>
    </property>
    修改regex-urlfilter.txt,将
    # accept anything else
    +.
    替换为:
    +^http://([a-z0-9]*.)*news.163.com/
    -.
    8、开发调试
    在左部Package Explorer的 nutch1.5.1文件夹上单击右键 > New > Folder > Folder name: urls
    在刚新建的urls目录下新建一个文本文件url,文本内容为:http://news.163.com
    打开src/java下的org.apache.nutch.crawl.Crawl.java类,单击右键Run As > Run Configurations > Arguments > 在Program arguments输入框中输入: urls -dir data -depth 3 > Run
    在需要调试的地方打上断点Debug As > Java Applicaton
    9、查看结果
    查看segments目录:
    打开src/java下的org.apache.nutch.segment.SegmentReader.java类
    单击右键Run As > Java Applicaton,控制台会输出该命令的使用方法
    单击右键Run As > Run Configurations > Arguments > 在Program arguments输入框中输入: -dump data/segments/*  data/segments/dump
    用文本编辑器打开文件data/segments/dump/dump查看segments中存储的信息

    查看crawldb目录:
    打开src/java下的org.apache.nutch.crawl.CrawlDbReader.java类
    单击右键Run As > Java Applicaton,控制台会输出该命令的使用方法
    单击右键Run As > Run Configurations > Arguments > 在Program arguments输入框中输入: data/crawldb -stats
    控制台会输出 crawldb统计信息

    查看linkdb目录:
    打开src/java下的org.apache.nutch.crawl.LinkDbReader.java类
    单击右键Run As > Java Applicaton,控制台会输出该命令的使用方法
    单击右键Run As > Run Configurations > Arguments > 在Program arguments输入框中输入: data/linkdb -dump data/linkdb_dump
    用文本编辑器打开文件data/linkdb_dump/part-00000查看linkdb中存储的信息
    10、全网分步骤抓取
    在左部Package Explorer的 nutch1.5.1文件夹下的build.xml文件上单击右键 > Run As > Ant Build
    cd  /home/ysc/workspace/nutch1.5.1/runtime/local
    #准备URL列表
    wget http://rdf.dmoz.org/rdf/content.rdf.u8.gz
    gunzip content.rdf.u8.gz
    mkdir dmoz
    bin/nutch org.apache.nutch.tools.DmozParser content.rdf.u8 -subset 5000 > dmoz/url
    #注入URL
    bin/nutch inject crawl/crawldb dmoz
    #生成抓取列表
    bin/nutch generate crawl/crawldb crawl/segments
    #第一次抓取
    s1=`ls -d crawl/segments/2* | tail -1`
    echo $s1
    #抓取网页
    bin/nutch fetch $s1
    #解析网页
    bin/nutch parse $s1
    #更新URL状态
    bin/nutch updatedb crawl/crawldb $s1
    #第二次抓取
    bin/nutch generate crawl/crawldb crawl/segments -topN 1000
    s2=`ls -d crawl/segments/2* | tail -1`
    echo $s2
    bin/nutch fetch $s2
    bin/nutch parse $s2
    bin/nutch updatedb crawl/crawldb $s2
    #第三次抓取
    bin/nutch generate crawl/crawldb crawl/segments -topN 1000
    s3=`ls -d crawl/segments/2* | tail -1`
    echo $s3
    bin/nutch fetch $s3
    bin/nutch parse $s3
    bin/nutch updatedb crawl/crawldb $s3
    #生成反向链接库
    bin/nutch invertlinks crawl/linkdb -dir crawl/segments

    11、索引和搜索
    cd  /home/ysc/
    wget http://mirror.bjtu.edu.cn/apache/lucene/solr/3.6.1/apache-solr-3.6.1.tgz
    tar -xvf apache-solr-3.6.1.tgz
    cd apache-solr-3.6.1 /example

    NUTCH_RUNTIME_HOME=/home/ysc/workspace/nutch1.5.1/runtime/local
    APACHE_SOLR_HOME=/home/ysc/apache-solr-3.6.1

    cp ${NUTCH_RUNTIME_HOME}/conf/schema.xml ${APACHE_SOLR_HOME}/example/solr/conf/
    如果需要把网页内容存储到索引中,则修改 schema.xml文件中的
    <field name="content" type="text" stored="false" indexed="true"/>

    <field name="content" type="text" stored="true" indexed="true"/>

    修改${APACHE_SOLR_HOME}/example/solr/conf/solrconfig.xml,将里面的<str name="df">text</str>都替换为<str name="df">content</str>

    把${APACHE_SOLR_HOME}/example/solr/conf/schema.xml中的 <schema name="nutch" version="1.5.1">修改为<schema name="nutch" version="1.5">
    #启动SOLR服务器
    java -jar start.jar

    http://127.0.0.1:8983/solr/admin/
    http://127.0.0.1:8983/solr/admin/stats.jsp

    cd  /home/ysc/workspace/nutch1.5.1/runtime/local
    #提交索引
    bin/nutch solrindex http://127.0.0.1:8983/solr/ crawl/crawldb -linkdb crawl/linkdb crawl/segments/*

    执行完整crawl:
    bin/nutch crawl urls -dir data -depth 2 -topN 100 -solr http://127.0.0.1:8983/solr/

    使用以下命令分页查看所有索引的文档:
    http://127.0.0.1:8983/solr/select/?q=*%3A*&version=2.2&start=0&rows=10&indent=on
    标题包含“网易”的文档:
    http://127.0.0.1:8983/solr/select/?q=title%3A%E7%BD%91%E6%98%93&version=2.2&start=0&rows=10&indent=on

    12、查看索引信息
    cd  /home/ysc/
    wget http://luke.googlecode.com/files/lukeall-3.5.0.jar
    java -jar lukeall-3.5.0.jar
    Path: /home/ysc/apache-solr-3.6.1/example/solr/data

    13、配置SOLR的中文分词
    cd  /home/ysc/
    wget http://mmseg4j.googlecode.com/files/mmseg4j-1.8.5.zip
    unzip mmseg4j-1.8.5.zip -d  mmseg4j-1.8.5

    APACHE_SOLR_HOME=/home/ysc/apache-solr-3.6.1
    mkdir $APACHE_SOLR_HOME/example/solr/lib
    mkdir $APACHE_SOLR_HOME/example/solr/dic
    cp mmseg4j-1.8.5/mmseg4j-all-1.8.5.jar $APACHE_SOLR_HOME/example/solr/lib
    cp mmseg4j-1.8.5/data/*.dic $APACHE_SOLR_HOME/example/solr/dic

    将${APACHE_SOLR_HOME}/example/solr/conf/schema.xml文件中的
    <tokenizer class="solr.WhitespaceTokenizerFactory"/>

    <tokenizer class="solr.StandardTokenizerFactory"/>
    替换为
    <tokenizer class="com.chenlb.mmseg4j.solr.MMSegTokenizerFactory" mode="complex" dicPath="/home/ysc/apache-solr-3.6.1/example/solr/dic"/>

    #重新启动SOLR服务器
    java -jar start.jar

    #重建索引,演示在开发环境中如何操作
    打开src/java下的org.apache.nutch.indexer.solr.SolrIndexer.java类
    单击右键Run As > Java Applicaton,控制台会输出该命令的使用方法
    单击右键Run As > Run Configurations > Arguments > 在Program arguments输入框中输入: http://127.0.0.1:8983/solr/  data/crawldb -linkdb  data/linkdb  data/segments/*
    使用luke重新打开索引就会发现分词起作用了

    三、nutch2.0
    nutch2.0和二中的nutch1.5.1的步骤相同,但在8、开发调试之前需要做以下配置:
    在左部Package Explorer的 nutch2.0文件夹上单击右键 > New > Folder > Folder name: data并指定数据存储方式,选如下之一:
    1、使用mysql作为数据存储
    1)、在nutch2.0/conf/nutch-site.xml中加入如下配置:
    <property>
      <name>storage.data.store.class</name>
      <value>org.apache.gora.sql.store.SqlStore</value>
    </property>
    2)、将nutch2.0/conf/gora.properties文件中的
    gora.sqlstore.jdbc.driver=org.hsqldb.jdbc.JDBCDriver
    gora.sqlstore.jdbc.url=jdbc:hsqldb:hsql://localhost/nutchtest
    gora.sqlstore.jdbc.user=sa
    gora.sqlstore.jdbc.password=
    修改为
    gora.sqlstore.jdbc.driver=com.mysql.jdbc.Driver
    gora.sqlstore.jdbc.url=jdbc:mysql://127.0.0.1:3306/nutch2
    gora.sqlstore.jdbc.user=root
    gora.sqlstore.jdbc.password=ROOT
    3)、打开nutch2.0/ivy/ivy.xml中的mysql-connector-java依赖
    4)、sudo apt-get install mysql-server
    2、使用hbase作为数据存储
    1)、在nutch2.0/conf/nutch-site.xml中加入如下配置:
    <property>
      <name>storage.data.store.class</name>
      <value>org.apache.gora.hbase.store.HBaseStore</value>
    </property>
    2)、打开nutch2.0/ivy/ivy.xml中的gora-hbase依赖
    3)、cd /home/ysc
    4)、wget http://mirror.bit.edu.cn/apache/hbase/hbase-0.90.5/hbase-0.90.5.tar.gz
    5)、tar -xvf hbase-0.90.5.tar.gz
    6)、vi  hbase-0.90.5/conf/hbase-site.xml
    加入以下配置:
      <property>
        <name>hbase.rootdir</name>
        <value>file:///home/ysc/hbase-0.90.5-database</value>
      </property>
    7)、hbase-0.90.5/bin/start-hbase.sh
    8)、将/home/ysc/hbase-0.90.5/hbase-0.90.5.jar加入开发环境eclipse的build path

    四、配置SSH
    三台机器 devcluster01, devcluster02, devcluster03,分别在每一台机器上面执行如下操作:
    1、sudo vi /etc/hosts
    加入以下配置:
    192.168.1.1 devcluster01
    192.168.1.2 devcluster02
    192.168.1.3 devcluster03
    2、安装SSH服务:
    sudo apt-get install openssh-server
    3、(有提示的时候回车键确认)
    ssh-keygen -t rsa
    该命令会在用户主目录下创建 .ssh 目录,并在其中创建两个文件:id_rsa 私钥文件。是基于 RSA 算法创建。该私钥文件要妥善保管,不要泄漏。id_rsa.pub 公钥文件。和 id_rsa 文件是一对儿,该文件作为公钥文件,可以公开。
    4、cp .ssh/id_rsa.pub .ssh/authorized_keys
    把 三台机器 devcluster01, devcluster02, devcluster03 的文件/home/ysc/.ssh/authorized_keys的内容复制出来合并成一个文件并替换每一台机器上的/home/ysc/.ssh/authorized_keys文件
    在devcluster01上面执行时,以下两条命令的主机为02和03
    在devcluster02上面执行时,以下两条命令的主机为01和03
    在devcluster03上面执行时,以下两条命令的主机为01和02
    5、ssh-copy-id -i .ssh/id_rsa.pub ysc@ devcluster02
    6、ssh-copy-id -i .ssh/id_rsa.pub ysc@ devcluster03
    以上两条命令实际上是将 .ssh/id_rsa.pub 公钥文件追加到远程主机 server 的 user 主目录下的 .ssh/authorized_keys 文件中。

    五、安装Hadoop Cluster(伪分布式运行模式)并运行Nutch
    步骤和四大同小异,只需要1台机器 devcluster01,所以黄色背景部分全部设置为devcluster01,不需要第11步

    六、安装Hadoop Cluster(分布式运行模式)并运行Nutch
    三台机器 devcluster01, devcluster02, devcluster03(vi /etc/hostname)
    使用用户ysc登陆 devcluster01:
    1、cd /home/ysc
    2、wget http://mirrors.tuna.tsinghua.edu.cn/apache/hadoop/common/hadoop-1.1.1/hadoop-1.1.1-bin.tar.gz
    3、tar -xvf hadoop-1.1.1-bin.tar.gz
    4、cd  hadoop-1.1.1
    5、vi conf/masters
    替换内容为 :
    devcluster01
    6、vi conf/slaves
    替换内容为 :
    devcluster02
    devcluster03
    7、vi conf/core-site.xml
    加入配置:
      <property>
        <name>fs.default.name</name>
        <value>hdfs://devcluster01:9000</value>
        <description>
           Where to find the Hadoop Filesystem through the network.
           Note 9000 is not the default port.
           (This is slightly changed from previous versions which didnt have "hdfs")
        </description>
      </property>
        <property>
         <name>hadoop.security.authorization</name>
          <value>true</value>
        </property>
    编辑conf/hadoop-policy.xml
    8、vi conf/hdfs-site.xml
    加入配置:
    <property>
      <name>dfs.name.dir</name>
      <value>/home/ysc/dfs/filesystem/name</value>
    </property>

    <property>
      <name>dfs.data.dir</name>
      <value>/home/ysc/dfs/filesystem/data</value>
    </property>

    <property>
      <name>dfs.replication</name>
      <value>1</value>
    </property>

    <property>
      <name>dfs.block.size</name>
      <value>671088640</value>
      <description>The default block size for new files.</description>
    </property>
    9、vi conf/mapred-site.xml
    加入配置:
    <property>
      <name>mapred.job.tracker</name>
      <value>devcluster01:9001</value>
      <description>
        The host and port that the MapReduce job tracker runs at. If
        "local", then jobs are run in-process as a single map and
        reduce task.
        Note 9001 is not the default port.
      </description>
    </property>

    <property>
      <name>mapred.reduce.tasks.speculative.execution</name>
      <value>false</value>
      <description>If true, then multiple instances of some reduce tasks
                   may be executed in parallel.</description>
    </property>

    <property>
      <name>mapred.map.tasks.speculative.execution</name>
      <value>false</value>
      <description>If true, then multiple instances of some map tasks
                   may be executed in parallel.</description>
    </property>

    <property>
      <name>mapred.child.java.opts</name>
      <value>-Xmx2000m</value>
    </property>

    <property>
      <name>mapred.tasktracker.map.tasks.maximum</name>
      <value>4</value>
      <description>
        the core number of host
      </description>
    </property>

    <property>
      <name>mapred.map.tasks</name>
      <value>4</value>
    </property>

    <property>
      <name>mapred.tasktracker.reduce.tasks.maximum</name>
      <value>4</value>
        <description>
        define mapred.map tasks to be number of slave hosts.the best number is the  number of slave hosts plus the core numbers of per host
        </description>
    </property>

    <property>
      <name>mapred.reduce.tasks</name>
      <value>4</value>
      <description>
        define mapred.reduce tasks to be number of slave hosts.the best number is the  number of slave hosts plus the core numbers of per host
      </description>
    </property>

    <property>
      <name>mapred.output.compression.type</name>
      <value>BLOCK</value>
      <description>If the job outputs are to compressed as SequenceFiles, how should they be compressed? Should be one of NONE, RECORD or BLOCK.
      </description>
    </property>

    <property>
      <name>mapred.output.compress</name>
      <value>true</value>
      <description>Should the job outputs be compressed?
      </description>
    </property>

    <property>
      <name>mapred.compress.map.output</name>
      <value>true</value>
      <description>Should the outputs of the maps be compressed before being                sent across the network. Uses SequenceFile compression.
      </description>
    </property>

    <property>
      <name>mapred.system.dir</name>
      <value>/home/ysc/mapreduce/system</value>
    </property>

    <property>
      <name>mapred.local.dir</name>
      <value>/home/ysc/mapreduce/local</value>
    </property>
    10、vi conf/hadoop-env.sh
    追加:
    export JAVA_HOME=/home/ysc/jdk1.7.0_05
    export HADOOP_HEAPSIZE=2000
    #替换掉默认的垃圾回收器,因为默认的垃圾回收器在多线程环境下会有更多的wait等待
    export HADOOP_OPTS="-server -Xmn256m -XX:+UseParNewGC -XX:+UseConcMarkSweepGC -XX:CMSInitiatingOccupancyFraction=70"
    11、复制HADOOP文件
    scp -r /home/ysc/hadoop-1.1.1 ysc@devcluster02:/home/ysc/hadoop-1.1.1
    scp -r /home/ysc/hadoop-1.1.1 ysc@devcluster03:/home/ysc/hadoop-1.1.1
    12、sudo vi /etc/profile
    追加并重启系统:
    export PATH=/home/ysc/hadoop-1.1.1/bin:$PATH
    13、格式化名称节点并启动集群
    hadoop namenode -format
    start-all.sh
    14、cd /home/ysc/workspace/nutch1.5.1/runtime/deploy
    mkdir urls
    echo http://news.163.com > urls/url
    hadoop dfs -put urls urls
    bin/nutch crawl urls -dir data -depth 2 -topN 100
    15、访问 http://localhost:50030 可以查看 JobTracker 的运行状态。访问 http://localhost:50060 可以查看 TaskTracker 的运行状态。访问 http://localhost:50070 可以查看 NameNode 以及整个分布式文件系统的状态,浏览分布式文件系统中的文件以及 log 等
    16、通过stop-all.sh停止集群
    17、如果NameNode和SecondaryNameNode不在同一台机器上,则在SecondaryNameNode的conf/hdfs-site.xml文件中加入配置:
    <property>
      <name>dfs.http.address</name>
      <value>namenode:50070</value>
    </property>

    七、配置Ganglia监控Hadoop集群和HBase集群
    1、服务器端(安装到master devcluster01上)
    1)、ssh devcluster01
    2)、useradd ganglia -g ganglia
    3)、sudo apt-get install  ganglia-monitor ganglia-webfront gmetad
    //补充:在Ubuntu10.04上,ganglia-webfront这个package名字叫ganglia-webfrontend
    //如果install出错,则运行sudo apt-get update,如果update出错,则删除出错路径
    4)、vi /etc/ganglia/gmond.conf
    先找到setuid = yes,改成setuid =no;
    在找到cluster块中的name,改成name =”hadoop-cluster”;
    5)、sudo apt-get install rrdtool
    6)、vi /etc/ganglia/gmetad.conf
    在这个配置文件中增加一些datasource,即其他2个被监控的节点,增加以下内容:
    data_source “hadoop-cluster” devcluster01:8649 devcluster02:8649 devcluster03:8649
    gridname "Hadoop"
    2、数据源端(安装到所有slaves上)
    1)、ssh devcluster02
    useradd ganglia -g ganglia
    sudo apt-get install  ganglia-monitor
    useradd ganglia -g ganglia
    2)、ssh devcluster03
    useradd ganglia -g ganglia
    sudo apt-get install  ganglia-monitor
    useradd ganglia -g ganglia
    3)、ssh devcluster01
    scp /etc/ganglia/gmond.conf devcluster02:/etc/ganglia/gmond.conf
    scp /etc/ganglia/gmond.conf devcluster03:/etc/ganglia/gmond.conf
    3、配置WEB
    1)、ssh devcluster01
    2)、sudo ln -s /usr/share/ganglia-webfrontend /var/www/ganglia
    3)、vi /etc/apache2/apache2.conf
    添加:
    ServerName devcluster01
    4、重启服务
    1)、ssh devcluster02
    sudo /etc/init.d/ganglia-monitor restart
    ssh devcluster03
    sudo /etc/init.d/ganglia-monitor restart
    2)、ssh devcluster01
    sudo /etc/init.d/ganglia-monitor restart
    sudo /etc/init.d/gmetad restart
    sudo /etc/init.d/apache2 restart
    5、访问页面
    http:// devcluster01/ganglia
    6、集成hadoop
    1)、ssh devcluster01
    2)、cd /home/ysc/hadoop-1.1.1
    3)、vi conf/hadoop-metrics2.properties
    # 大于0.20以后的版本用ganglia31 *.sink.ganglia.class=org.apache.hadoop.metrics2.sink.ganglia.GangliaSink31
    *.sink.ganglia.period=10
    # default for supportsparse is false
    *.sink.ganglia.supportsparse=true
    *.sink.ganglia.slope=jvm.metrics.gcCount=zero,jvm.metrics.memHeapUsedM=both
    *.sink.ganglia.dmax=jvm.metrics.threadsBlocked=70,jvm.metrics.memHeapUsedM=40
    #广播IP地址,这是缺省的,统一设该值(只能用组播地址239.2.11.71)
    namenode.sink.ganglia.servers=239.2.11.71:8649
    datanode.sink.ganglia.servers=239.2.11.71:8649
    jobtracker.sink.ganglia.servers=239.2.11.71:8649
    tasktracker.sink.ganglia.servers=239.2.11.71:8649
    maptask.sink.ganglia.servers=239.2.11.71:8649
    reducetask.sink.ganglia.servers=239.2.11.71:8649
    dfs.class=org.apache.hadoop.metrics.ganglia.GangliaContext31
    dfs.period=10
    dfs.servers=239.2.11.71:8649
    mapred.class=org.apache.hadoop.metrics.ganglia.GangliaContext31
    mapred.period=10
    mapred.servers=239.2.11.71:8649
    jvm.class=org.apache.hadoop.metrics.ganglia.GangliaContext31
    jvm.period=10
    jvm.servers=239.2.11.71:8649
    4)、scp conf/hadoop-metrics2.properties root@devcluster02:/home/ysc/hadoop-1.1.1/conf/hadoop-metrics2.properties
    5)、scp conf/hadoop-metrics2.properties root@devcluster03:/home/ysc/hadoop-1.1.1/conf/hadoop-metrics2.properties
    6)、stop-all.sh
    7)、start-all.sh
    7、集成hbase
    1)、ssh devcluster01
    2)、cd /home/ysc/hbase-0.92.2
    3)、vi conf/hadoop-metrics.properties(只能用组播地址239.2.11.71)
    hbase.extendedperiod = 3600
    hbase.class=org.apache.hadoop.metrics.ganglia.GangliaContext31
    hbase.period=10
    hbase.servers=239.2.11.71:8649
    jvm.class=org.apache.hadoop.metrics.ganglia.GangliaContext31
    jvm.period=10
    jvm.servers=239.2.11.71:8649
    rpc.class=org.apache.hadoop.metrics.ganglia.GangliaContext31
    rpc.period=10
    rpc.servers=239.2.11.71:8649
    4)、scp conf/hadoop-metrics.properties root@devcluster02:/home/ysc/ hbase-0.92.2/conf/hadoop-metrics.properties
    5)、scp conf/hadoop-metrics.properties root@devcluster03:/home/ysc/ hbase-0.92.2/conf/hadoop-metrics.properties
    6)、stop-hbase.sh
    7)、start-hbase.sh

    八、Hadoop配置Snappy压缩
    1、wget http://snappy.googlecode.com/files/snappy-1.0.5.tar.gz
    2、tar -xzvf snappy-1.0.5.tar.gz
    3、cd snappy-1.0.5
    4、./configure
    5、make
    6、make install
    7、scp /usr/local/lib/libsnappy* devcluster01:/home/ysc/hadoop-1.1.1/lib/native/Linux-amd64-64/
    scp /usr/local/lib/libsnappy* devcluster02:/home/ysc/hadoop-1.1.1/lib/native/Linux-amd64-64/
    scp /usr/local/lib/libsnappy* devcluster03:/home/ysc/hadoop-1.1.1/lib/native/Linux-amd64-64/
    8、vi /etc/profile
    追加:
    export LD_LIBRARY_PATH=/home/ysc/hadoop-1.1.1/lib/native/Linux-amd64-64
    9、修改mapred-site.xml
    <property>
      <name>mapred.output.compression.type</name>
      <value>BLOCK</value>
      <description>If the job outputs are to compressed as SequenceFiles, how should
       they be compressed? Should be one of NONE, RECORD or BLOCK.
      </description>
    </property>

    <property>
      <name>mapred.output.compress</name>
      <value>true</value>
      <description>Should the job outputs be compressed?
      </description>
    </property>

    <property>
      <name>mapred.compress.map.output</name>
      <value>true</value>
      <description>Should the outputs of the maps be compressed before being
       sent across the network. Uses SequenceFile compression.
      </description>
    </property>

    <property>
      <name>mapred.map.output.compression.codec</name>
      <value>org.apache.hadoop.io.compress.SnappyCodec</value>
      <description>If the map outputs are compressed, how should they be
       compressed?
      </description>
    </property>

    <property>
      <name>mapred.output.compression.codec</name>
      <value>org.apache.hadoop.io.compress.SnappyCodec</value>
      <description>If the job outputs are compressed, how should they be compressed?
      </description>
    </property>

    九、Hadoop配置Lzo压缩
    1、wget http://www.oberhumer.com/opensource/lzo/download/lzo-2.06.tar.gz
    2、tar -zxvf lzo-2.06.tar.gz
    3、cd lzo-2.06
    4、./configure --enable-shared
    5、make
    6、make install
    7、scp /usr/local/lib/liblzo2.* devcluster01:/lib/x86_64-linux-gnu
    scp /usr/local/lib/liblzo2.* devcluster02:/lib/x86_64-linux-gnu
    scp /usr/local/lib/liblzo2.* devcluster03:/lib/x86_64-linux-gnu
    8、wget http://hadoop-gpl-compression.apache-extras.org.codespot.com/files/hadoop-gpl-compression-0.1.0-rc0.tar.gz
    9、tar -xzvf hadoop-gpl-compression-0.1.0-rc0.tar.gz
    10、cd hadoop-gpl-compression-0.1.0
    11、cp lib/native/Linux-amd64-64/* /home/ysc/hadoop-1.1.1/lib/native/Linux-amd64-64/
    12、cp hadoop-gpl-compression-0.1.0.jar /home/ysc/hadoop-1.1.1/lib/(这里hadoop集群的版本要和compression使用的版本一致)
    13、scp -r /home/ysc/hadoop-1.1.1/lib devcluster02:/home/ysc/hadoop-1.1.1/
    scp -r /home/ysc/hadoop-1.1.1/lib devcluster03:/home/ysc/hadoop-1.1.1/
    14、vi /etc/profile
    追加:
    export LD_LIBRARY_PATH=/home/ysc/hadoop-1.1.1/lib/native/Linux-amd64-64
    15、修改core-site.xml
    <property>
      <name>io.compression.codecs</name>
      <value>com.hadoop.compression.lzo.LzoCodec,org.apache.hadoop.io.compress.DefaultCodec,org.apache.hadoop.io.compress.GzipCodec,org.apache.hadoop.io.compress.BZip2Codec,org.apache.hadoop.io.compress.SnappyCodec</value>
      <description>A list of the compression codec classes that can be used
       for compression/decompression.</description>
    </property>

    <property>
      <name>io.compression.codec.lzo.class</name>
      <value>com.hadoop.compression.lzo.LzoCodec</value>
    </property>

    <property>
      <name>fs.trash.interval</name>
      <value>1440</value>
      <description>Number of minutes between trash checkpoints.
      If zero, the trash feature is disabled.
      </description>
    </property>
    16、修改mapred-site.xml
    <property>
      <name>mapred.output.compression.type</name>
      <value>BLOCK</value>
      <description>If the job outputs are to compressed as SequenceFiles, how should
       they be compressed? Should be one of NONE, RECORD or BLOCK.
      </description>
    </property>

    <property>
      <name>mapred.output.compress</name>
      <value>true</value>
      <description>Should the job outputs be compressed?
      </description>
    </property>

    <property>
      <name>mapred.compress.map.output</name>
      <value>true</value>
      <description>Should the outputs of the maps be compressed before being
       sent across the network. Uses SequenceFile compression.
      </description>
    </property>

    <property>
      <name>mapred.map.output.compression.codec</name>
      <value>com.hadoop.compression.lzo.LzoCodec</value>
      <description>If the map outputs are compressed, how should they be
       compressed?
      </description>
    </property>

    <property>
      <name>mapred.output.compression.codec</name>
      <value>com.hadoop.compression.lzo.LzoCodec</value>
      <description>If the job outputs are compressed, how should they be compressed?
      </description>
    </property>

    十、配置zookeeper集群以运行hbase
    1、ssh devcluster01
    2、cd /home/ysc
    3、wget http://mirror.bjtu.edu.cn/apache/zookeeper/stable/zookeeper-3.4.5.tar.gz
    4、tar -zxvf  zookeeper-3.4.5.tar.gz
    5、cd zookeeper-3.4.5
    6、cp conf/zoo_sample.cfg  conf/zoo.cfg
    7、vi conf/zoo.cfg
    修改:dataDir=/home/ysc/zookeeper
    添加:
    server.1=devcluster01:2888:3888
    server.2=devcluster02:2888:3888
    server.3=devcluster03:2888:3888
    maxClientCnxns=100
    8、scp -r  zookeeper-3.4.5  devcluster01:/home/ysc
    scp -r  zookeeper-3.4.5  devcluster02:/home/ysc
    scp -r  zookeeper-3.4.5  devcluster03:/home/ysc
    9、分别在三台机器上面执行:
    ssh devcluster01
    mkdir /home/ysc/zookeeper(注:dataDir是zookeeper的数据目录,需要手动创建)
    echo 1 > /home/ysc/zookeeper/myid
    ssh devcluster02
    mkdir /home/ysc/zookeeper
    echo 2 > /home/ysc/zookeeper/myid
    ssh devcluster03
    mkdir /home/ysc/zookeeper
    echo 3 > /home/ysc/zookeeper/myid
    10、分别在三台机器上面执行:
    cd /home/ysc/zookeeper-3.4.5
    bin/zkServer.sh start
    bin/zkCli.sh -server devcluster01:2181
    bin/zkServer.sh status

    十一、配置Hbase集群以运行nutch-2.1(Region Servers会因为内存的问题宕机)
    1、nutch-2.1使用gora-0.2.1, gora-0.2.1使用hbase-0.90.4,hbase-0.90.4和hadoop-1.1.1不兼容,hbase-0.94.4和gora-0.2.1不兼容,hbase-0.92.2没问题。hbase存在系统时间同步的问题,并且误差要再30s以内。
    sudo apt-get install ntp
    sudo ntpdate -u 210.72.145.44
    2、HBase是数据库,会在同一时间使用很多的文件句柄。大多数linux系统使用的默认值1024是不能满足的。还需要修改 hbase 用户的 nproc,在压力下,如果过低会造成 OutOfMemoryError异常。
    vi /etc/security/limits.conf
    添加:
    ysc soft nproc 32000
    ysc hard nproc 32000
    ysc soft nofile 32768
    ysc hard nofile 32768
    vi /etc/pam.d/common-session
    添加:
    session required  pam_limits.so
    3、登陆master,下载并解压hbase
    ssh devcluster01
    cd /home/ysc
    wget http://apache.etoak.com/hbase/hbase-0.92.2/hbase-0.92.2.tar.gz
    tar -zxvf hbase-0.92.2.tar.gz
    cd hbase-0.92.2
    4、修改配置文件hbase-env.sh
    vi conf/hbase-env.sh
    追加:
    export JAVA_HOME=/home/ysc/jdk1.7.0_05
    export HBASE_MANAGES_ZK=false
    export HBASE_HEAPSIZE=10000
    #替换掉默认的垃圾回收器,因为默认的垃圾回收器在多线程环境下会有更多的wait等待
    export HBASE_OPTS="-server -Xmn256m -XX:+UseParNewGC -XX:+UseConcMarkSweepGC -XX:CMSInitiatingOccupancyFraction=70"
    5、修改配置文件hbase-site.xml
    vi conf/hbase-site.xml
    <property> 
    <name>hbase.rootdir</name> 
    <value>hdfs://devcluster01:9000/hbase</value>    
    </property>
    <property> 
    <name>hbase.cluster.distributed</name> 
    <value>true</value> 
    </property> 
    <property>  
    <name>hbase.zookeeper.quorum</name>       
    <value>devcluster01,devcluster02,devcluster03</value>  
    </property>
    <property>
    <name>hfile.block.cache.size</name>
    <value>0.25</value>
    <description>
    Percentage of maximum heap (-Xmx setting) to allocate to block cache
    used by HFile/StoreFile. Default of 0.25 means allocate 25%.
    Set to 0 to disable but it's not recommended.
    </description>
    </property>
    <property>
    <name>hbase.regionserver.global.memstore.upperLimit</name>
    <value>0.4</value>
    <description>Maximum size of all memstores in a region server before new
      updates are blocked and flushes are forced. Defaults to 40% of heap
    </description>
    </property>
       <property>
    <name>hbase.regionserver.global.memstore.lowerLimit</name>
    <value>0.35</value>
    <description>When memstores are being forced to flush to make room in
      memory, keep flushing until we hit this mark. Defaults to 35% of heap.
      This value equal to hbase.regionserver.global.memstore.upperLimit causes
      the minimum possible flushing to occur when updates are blocked due to
      memstore limiting.
    </description>
       </property>
    <property>
    <name>hbase.hregion.majorcompaction</name>
    <value>0</value>
    <description>The time (in miliseconds) between 'major' compactions of all
    HStoreFiles in a region.  Default: 1 day.
    Set to 0 to disable automated major compactions.
    </description>
    </property>
    6、修改配置文件regionservers
    vi conf/regionservers
    devcluster01
    devcluster02
    devcluster03
    7、因为HBase建立在Hadoop之上,Hadoop使用的hadoop*.jar和HBase使用的 必须 一致。所以要将 HBase lib 目录下的hadoop*.jar替换成Hadoop里面的那个,防止版本冲突。
    cp  /home/ysc/hadoop-1.1.1/hadoop-core-1.1.1.jar  /home/ysc/hbase-0.92.2/lib
    rm  /home/ysc/hbase-0.92.2/lib/hadoop-core-1.0.3.jar
    8、复制文件到regionservers
    scp -r /home/ysc/hbase-0.92.2 devcluster01:/home/ysc
    scp -r /home/ysc/hbase-0.92.2 devcluster02:/home/ysc
    scp -r /home/ysc/hbase-0.92.2 devcluster03:/home/ysc
    9、启动hadoop并创建目录
    hadoop fs -mkdir /hbase
    10、管理HBase集群:
    启动初始 HBase 集群:
    bin/start-hbase.sh
    停止HBase 集群:
    bin/stop-hbase.sh
    启动额外备份主服务器,可以启动到 9 个备份服务器 (总数10 个):
    bin/local-master-backup.sh start 1
    bin/local-master-backup.sh start 2 3
    启动更多 regionservers, 支持到 99 个额外regionservers (总100个):
    bin/local-regionservers.sh start 1
    bin/local-regionservers.sh start 2 3 4 5
    停止备份主服务器:
    cat /tmp/hbase-ysc-1-master.pid |xargs kill -9
    停止单独 regionserver:
    bin/local-regionservers.sh stop 1
    使用HBase命令行模式:
    bin/hbase shell
    11、web界面
    http://devcluster01:60010
    http://devcluster01:60030
    12、如运行nutch2.1则方法一:
    cp conf/hbase-site.xml /home/ysc/nutch-2.1/conf
    cd /home/ysc/nutch-2.1
    ant
    cd runtime/deploy
    unzip -d apache-nutch-2.1 apache-nutch-2.1.job
    rm  apache-nutch-2.1.job
    cd apache-nutch-2.1
    rm lib/hbase-0.90.4.jar
    cp /home/ysc/hbase-0.92.2/hbase-0.92.2.jar  lib
    zip -r ../apache-nutch-2.1.job ./*
    cd ..
    rm -r apache-nutch-2.1
    13、如运行nutch2.1则方法二:
    cp conf/hbase-site.xml /home/ysc/nutch-2.1/conf
    cd /home/ysc/nutch-2.1
    cp /home/ysc/hbase-0.92.2/hbase-0.92.2.jar  lib
    ant
    cd runtime/deploy
    zip -d apache-nutch-2.1.job lib/hbase-0.90.4.jar

    启用snappy压缩:
    1、vi conf/gora-hbase-mapping.xml
    在family上面添加属性:compression="SNAPPY"
    2、mkdir /home/ysc/hbase-0.92.2/lib/native/Linux-amd64-64
    3、cp /home/ysc/hadoop-1.1.1/lib/native/Linux-amd64-64/* /home/ysc/hbase-0.92.2/lib/native/Linux-amd64-64
    4、vi /home/ysc/hbase-0.92.2/conf/hbase-site.xml
    增加:
                    <property>
                            <name>hbase.regionserver.codecs</name>
                            <value>snappy</value>
                    </property>



    十二、配置Accumulo集群以运行nutch-2.1(gora存在BUG)
    1、wget http://apache.etoak.com/accumulo/1.4.2/accumulo-1.4.2-dist.tar.gz
    2、tar -xzvf accumulo-1.4.2-dist.tar.gz
    3、cd accumulo-1.4.2
    4、cp conf/examples/3GB/standalone/* conf
    5、vi conf/accumulo-env.sh
    export HADOOP_HOME=/home/ysc/cluster3
    export ZOOKEEPER_HOME=/home/ysc/zookeeper-3.4.5
    export JAVA_HOME=/home/jdk1.7.0_01
    export ACCUMULO_HOME=/home/ysc/accumulo-1.4.2
    6、vi conf/slaves
    devcluster01
    devcluster02
    devcluster03
    7、vi conf/masters
    devcluster01
    8、vi conf/accumulo-site.xml
    <property>
      <name>instance.zookeeper.host</name>
      <value>host6:2181,host8:2181</value>
      <description>comma separated list of zookeeper servers</description>
    </property>

    <property>
      <name>logger.dir.walog</name>
      <value>walogs</value>
      <description>The directory used to store write-ahead logs on the local filesystem. It is possible to specify a comma-separated list of directories.</description>
    </property>

    <property>
      <name>instance.secret</name>
      <value>ysc</value>
      <description>A secret unique to a given instance that all servers must know in order to communicate with one another.
       Change it before initialization. To change it later use ./bin/accumulo org.apache.accumulo.server.util.ChangeSecret [oldpasswd] [newpasswd],
       and then update this file.
      </description>
    </property>

    <property>
      <name>tserver.memory.maps.max</name>
      <value>3G</value>
    </property>

    <property>
      <name>tserver.cache.data.size</name>
      <value>50M</value>
    </property>

    <property>
      <name>tserver.cache.index.size</name>
      <value>512M</value>
    </property>

    <property>
      <name>trace.password</name>
      <!--
    change this to the root user's password, and/or change the user below
       -->
      <value>ysc</value>
    </property>

    <property>
      <name>trace.user</name>
      <value>root</value>
    </property>
    9、bin/accumulo init
    10、bin/start-all.sh
    11、bin/stop-all.sh
    12、web访问:http://devcluster01:50095/

    修改nutch2.1:
    1、cd  /home/ysc/nutch-2.1
    2、vi  conf/gora.properties
    增加:
    gora.datastore.default=org.apache.gora.accumulo.store.AccumuloStore
    gora.datastore.accumulo.mock=false
    gora.datastore.accumulo.instance=accumulo
    gora.datastore.accumulo.zookeepers=host6,host8
    gora.datastore.accumulo.user=root
    gora.datastore.accumulo.password=ysc
    3、vi  conf/nutch-site.xml
    增加:
    <property>
      <name>storage.data.store.class</name>
      <value>org.apache.gora.accumulo.store.AccumuloStore</value>
    </property>
    4、vi ivy/ivy.xml
    增加:
    <dependency org="org.apache.gora" name="gora-accumulo" rev="0.2.1" conf="*->default" />
    5、升级accumulo
    cp /home/ysc/accumulo-1.4.2/lib/accumulo-core-1.4.2.jar  /home/ysc/nutch-2.1/lib
    cp /home/ysc/accumulo-1.4.2/lib/accumulo-start-1.4.2.jar  /home/ysc/nutch-2.1/lib
    cp /home/ysc/accumulo-1.4.2/lib/cloudtrace-1.4.2.jar  /home/ysc/nutch-2.1/lib
    6、ant
    7、cd runtime/deploy
    8、删除旧jar
    zip -d apache-nutch-2.1.job lib/accumulo-core-1.4.0.jar
    zip -d apache-nutch-2.1.job lib/accumulo-start-1.4.0.jar
    zip -d apache-nutch-2.1.job lib/cloudtrace-1.4.2.jar

    十三、配置Cassandra 集群以运行nutch-2.1(Cassandra 采用去中心化结构)
    1、vi /etc/hosts(注意:需要登录到每一台机器上面,将localhost解析到实际地址)
    192.168.1.1       localhost
    2、wget http://labs.mop.com/apache-mirror/cassandra/1.2.0/apache-cassandra-1.2.0-bin.tar.gz
    3、tar -xzvf  apache-cassandra-1.2.0-bin.tar.gz
    4、cd apache-cassandra-1.2.0
    5、vi conf/cassandra-env.sh
    增加:
    MAX_HEAP_SIZE="4G"
    HEAP_NEWSIZE="800M"
    6、vi conf/log4j-server.properties
    修改:
    log4j.appender.R.File=/home/ysc/cassandra/system.log
    7、vi conf/cassandra.yaml
    修改:
    cluster_name: 'Cassandra  Cluster'
    data_file_directories:
        - /home/ysc/cassandra/data
    commitlog_directory: /home/ysc/cassandra/commitlog
    saved_caches_directory: /home/ysc/cassandra/saved_caches

    - seeds: "192.168.1.1"
    listen_address: 192.168.1.1
    rpc_address: 192.168.1.1

    thrift_framed_transport_size_in_mb: 1023
    thrift_max_message_length_in_mb: 1024
    8、vi bin/stop-server
    增加:
    user=`whoami`
    pgrep -u $user -f cassandra | xargs kill -9
    9、复制cassandra到其他节点:
    cd ..
    scp -r apache-cassandra-1.2.0 devcluster02:/home/ysc
    scp -r apache-cassandra-1.2.0 devcluster03:/home/ysc
    分别在devcluster02和devcluster03上面修改:
    vi conf/cassandra.yaml
    listen_address: 192.168.1.2
    rpc_address: 192.168.1.2
    vi conf/cassandra.yaml
    listen_address: 192.168.1.3
    rpc_address: 192.168.1.3
    10、分别在3个节点上面运行
    bin/cassandra
    bin/cassandra -f   参数 -f 的作用是让 Cassandra 以前端程序方式运行,这样有利于调试和观察日志信息,而在实际生产环境中这个参数是不需要的(即 Cassandra 会以 daemon 方式运行)
    11、bin/nodetool -host devcluster01 ring
           bin/nodetool -host devcluster01 info
    12、bin/stop-server
    13、bin/cassandra-cli

    修改nutch2.1:
    1、cd  /home/ysc/nutch-2.1
    2、vi  conf/gora.properties
    增加:
    gora.cassandrastore.servers=host2:9160,host6:9160,host8:9160
    3、vi  conf/nutch-site.xml
    增加:
    <property>
      <name>storage.data.store.class</name>
      <value>org.apache.gora.cassandra.store.CassandraStore</value>
    </property>
    4、vi ivy/ivy.xml
    增加:
    <dependency org="org.apache.gora" name="gora-cassandra" rev="0.2.1" conf="*->default" />
    5、升级cassandra
    cp /home/ysc/apache-cassandra-1.2.0/lib/apache-cassandra-1.2.0.jar  /home/ysc/nutch-2.1/lib
    cp /home/ysc/apache-cassandra-1.2.0/lib/apache-cassandra-thrift-1.2.0.jar  /home/ysc/nutch-2.1/lib
    cp /home/ysc/apache-cassandra-1.2.0/lib/jline-1.0.jar  /home/ysc/nutch-2.1/lib
    6、ant
    7、cd runtime/deploy
    8、删除旧jar
    zip -d apache-nutch-2.1.job lib/cassandra-thrift-1.1.2.jar
    zip -d apache-nutch-2.1.job lib/jline-0.9.1.jar

    十四、配置MySQL 单机服务器以运行nutch-2.1
    1、apt-get install mysql-server mysql-client
    2、vi /etc/mysql/my.cnf
    修改:
    bind-address            = 221.194.43.2
    在[client]下增加:
    default-character-set=utf8
    在[mysqld]下增加:
    default-character-set=utf8
    3、mysql –uroot –pysc
    SHOW VARIABLES LIKE '%character%';
    4、service mysql restart
    5、mysql –uroot –pysc
    GRANT ALL PRIVILEGES ON *.* TO root@"%" IDENTIFIED BY "ysc";
    6、vi conf/gora-sql-mapping.xml
    修改字段的长度
    <primarykey column="id" length="333"/>
    <field name="content" column="content" />
    <field name="text" column="text" length="19892"/>
    7、启动nutch之后登陆mysql
    ALTER TABLE webpage MODIFY COLUMN content MEDIUMBLOB;
    ALTER TABLE webpage MODIFY COLUMN text MEDIUMTEXT;
    ALTER TABLE webpage MODIFY COLUMN title MEDIUMTEXT;
    ALTER TABLE webpage MODIFY COLUMN reprUrl MEDIUMTEXT;
    ALTER TABLE webpage MODIFY COLUMN baseUrl MEDIUMTEXT;
    ALTER TABLE webpage MODIFY COLUMN typ MEDIUMTEXT;
    ALTER TABLE webpage MODIFY COLUMN inlinks MEDIUMBLOB;
    ALTER TABLE webpage MODIFY COLUMN outlinks MEDIUMBLOB;

    修改nutch2.1:
    1、cd  /home/ysc/nutch-2.1
    2、vi  conf/gora.properties
    增加:
    gora.sqlstore.jdbc.driver=com.mysql.jdbc.Driver
    gora.sqlstore.jdbc.url=jdbc:mysql://host2:3306/nutch?createDatabaseIfNotExist=true&useUnicode=true&characterEncoding=utf8
    gora.sqlstore.jdbc.user=root
    gora.sqlstore.jdbc.password=ysc
    3、vi  conf/nutch-site.xml
    增加:
    <property>
      <name>storage.data.store.class</name>
      <value>org.apache.gora.sql.store.SqlStore </value>
    </property>

    <property>
      <name>encodingdetector.charset.min.confidence</name>
      <value>1</value>
      <description>A integer between 0-100 indicating minimum confidence value
      for charset auto-detection. Any negative value disables auto-detection.
      </description>
    </property>
    4、vi ivy/ivy.xml
    增加:
    <dependency org="mysql" name="mysql-connector-java" rev="5.1.18" conf="*->default"/>

    十五、nutch2.1 使用DataFileAvroStore作为数据源
    1、cd  /home/ysc/nutch-2.1
    2、vi  conf/gora.properties
    增加:
    gora.datafileavrostore.output.path=datafileavrostore
    gora.datafileavrostore.input.path=datafileavrostore
    3、vi  conf/nutch-site.xml
    增加:
    <property>
      <name>storage.data.store.class</name>
      <value>org.apache.gora.avro.store.DataFileAvroStore</value>
    </property>

    <property>
      <name>encodingdetector.charset.min.confidence</name>
      <value>1</value>
      <description>A integer between 0-100 indicating minimum confidence value
      for charset auto-detection. Any negative value disables auto-detection.
      </description>
    </property>



    十六、nutch2.1 使用AvroStore作为数据源
    1、cd  /home/ysc/nutch-2.1
    2、vi  conf/gora.properties
    增加:
    gora.avrostore.codec.type=BINARY
    gora.avrostore.input.path=avrostore
    gora.avrostore.output.path=avrostore
    3、vi  conf/nutch-site.xml
    增加:
    <property>
      <name>storage.data.store.class</name>
      <value>org.apache.gora.avro.store.AvroStore</value>
    </property>

    <property>
      <name>encodingdetector.charset.min.confidence</name>
      <value>1</value>
      <description>A integer between 0-100 indicating minimum confidence value
      for charset auto-detection. Any negative value disables auto-detection.
      </description>
    </property>



    十七、配置SOLR
    配置tomcat:
    1、wget http://www.fayea.com/apache-mirror/tomcat/tomcat-7/v7.0.35/bin/apache-tomcat-7.0.35.tar.gz
    2、tar -xzvf apache-tomcat-7.0.35.tar.gz
    3、cd apache-tomcat-7.0.35
    4、vi conf/server.xml
    增加URIEncoding="UTF-8":
    <Connector port="8080" protocol="HTTP/1.1"
       connectionTimeout="20000"
       redirectPort="8443" URIEncoding="UTF-8"/>
    5、mkdir conf/Catalina
    6、mkdir conf/Catalina/localhost
    7、vi conf/Catalina/localhost/solr.xml
    增加:
    <Context path="/solr">
    <Environment name="solr/home" type="java.lang.String" value="/home/ysc/solr/configuration/" override="false"/>
    </Context>
    8、cd ..

    下载SOLR:
    1、wget http://mirrors.tuna.tsinghua.edu.cn/apache/lucene/solr/4.1.0/solr-4.1.0.tgz
    2、tar -xzvf solr-4.1.0.tgz

    复制资源:
    1、mkdir /home/ysc/solr
    2、cp -r solr-4.1.0/example/solr  /home/ysc/solr/configuration
    3、unzip solr-4.1.0/example/webapps/solr.war -d /home/ysc/apache-tomcat-7.0.35/webapps/solr

    配置nutch:
    1、复制schema:
    cp /home/ysc/nutch-1.6/conf/schema-solr4.xml /home/ysc/solr/configuration/collection1/conf/schema.xml
    2、vi /home/ysc/solr/configuration/collection1/conf/schema.xml
    在<fields>下增加:
    <field name="_version_" type="long" indexed="true" stored="true"/>

    配置中文分词:
    1、wget http://mmseg4j.googlecode.com/files/mmseg4j-1.9.1.v20130120-SNAPSHOT.zip
    2、unzip mmseg4j-1.9.1.v20130120-SNAPSHOT.zip
    3、cp mmseg4j-1.9.1-SNAPSHOT/dist/* /home/ysc/apache-tomcat-7.0.35/webapps/solr/WEB-INF/lib
    4、unzip mmseg4j-1.9.1-SNAPSHOT/dist/mmseg4j-core-1.9.1-SNAPSHOT.jar -d  mmseg4j-1.9.1-SNAPSHOT/dist/mmseg4j-core-1.9.1-SNAPSHOT
    5、mkdir /home/ysc/dic
    6、cp   mmseg4j-1.9.1-SNAPSHOT/dist/mmseg4j-core-1.9.1-SNAPSHOT/data/* /home/ysc/dic
    7、vi /home/ysc/solr/configuration/collection1/conf/schema.xml
    将文件中的
    <tokenizer class="solr.WhitespaceTokenizerFactory"/>

    <tokenizer class="solr.StandardTokenizerFactory"/>
    替换为
    <tokenizer class="com.chenlb.mmseg4j.solr.MMSegTokenizerFactory" mode="complex" dicPath="/home/ysc/dic"/>

    配置tomcat本地库:
    1、wget http://apache.spd.co.il/apr/apr-1.4.6.tar.gz
    2、tar -xzvf apr-1.4.6.tar.gz
    3、cd apr-1.4.6
    4、./configure
    5、make
    6、make  install

    1、wget http://mirror.bjtu.edu.cn/apache/apr/apr-util-1.5.1.tar.gz
    2、tar -xzvf apr-util-1.5.1.tar.gz
    3、cd apr-util-1.5.1
    4、./configure --with-apr=/usr/local/apr
    5、make
    6、make  install

    1、wget http://mirror.bjtu.edu.cn/apache//tomcat/tomcat-connectors/native/1.1.24/source/tomcat-native-1.1.24-src.tar.gz
    2、tar -zxvf tomcat-native-1.1.24-src.tar.gz
    3、cd tomcat-native-1.1.24-src/jni/native
    4、./configure --with-apr=/usr/local/apr
                    --with-java-home=/home/ysc/jdk1.7.0_01
                    --with-ssl=no
                    --prefix=/home/ysc/apache-tomcat-7.0.35
    5、make
    6、make  install
    7、vi /etc/profile
    增加:
    export LD_LIBRARY_PATH=$LD_LIBRARY_PATH:/home/ysc/apache-tomcat-7.0.35/lib:/usr/local/apr/lib
    8、source /etc/profile

    启动tomcat:
    cd apache-tomcat-7.0.35
    bin/catalina.sh start
    http://devcluster01:8080/solr/

    十八、Nagios监控
    服务端:
    1、apt-get install apache2 nagios3 nagios-nrpe-plugin
    输入密码:nagiosadmin
    2、apt-get install nagios3-doc
    3、vi /etc/nagios3/conf.d/hostgroups_nagios2.cfg
    define hostgroup {
    hostgroup_name  nagios-servers
    alias           nagios servers
    members         devcluster01,devcluster02,devcluster03
    }
    4、cp  /etc/nagios3/conf.d/localhost_nagios2.cfg /etc/nagios3/conf.d/devcluster01_nagios2.cfg
    vi /etc/nagios3/conf.d/devcluster01_nagios2.cfg
    替换:
    g/localhost/s//devcluster01/g
    g/127.0.0.1/s//192.168.1.1/g
    5、cp  /etc/nagios3/conf.d/localhost_nagios2.cfg /etc/nagios3/conf.d/devcluster02_nagios2.cfg
    vi /etc/nagios3/conf.d/devcluster02_nagios2.cfg
    替换:
    g/localhost/s//devcluster02/g
    g/127.0.0.1/s//192.168.1.2/g
    6、cp  /etc/nagios3/conf.d/localhost_nagios2.cfg /etc/nagios3/conf.d/devcluster03_nagios2.cfg
    vi /etc/nagios3/conf.d/devcluster03_nagios2.cfg
    替换:
    g/localhost/s//devcluster03/g
    g/127.0.0.1/s//192.168.1.3/g

    7、vi /etc/nagios3/conf.d/services_nagios2.cfg
    将hostgroup_name改为nagios-servers
    增加:
    # check that web services are running
    define service {
    hostgroup_name                  nagios-servers
    service_description             HTTP
    check_command                   check_http
    use                             generic-service
    notification_interval           0 ; set > 0 if you want to be renotified
    }

    # check that ssh services are running
    define service {
    hostgroup_name                  nagios-servers
    service_description             SSH
    check_command                   check_ssh
    use                             generic-service
    notification_interval           0 ; set > 0 if you want to be renotified
    }
    8、vi /etc/nagios3/conf.d/extinfo_nagios2.cfg
    将hostgroup_name改为nagios-servers
    增加:
    define hostextinfo{
    hostgroup_name   nagios-servers
    notes            nagios-servers
    #       notes_url        http://webserver.localhost.localdomain/hostinfo.pl?host=netware1
    icon_image       base/debian.png
    icon_image_alt   Debian GNU/Linux
    vrml_image       debian.png
    statusmap_image  base/debian.gd2
    }
    9、sudo /etc/init.d/nagios3 restart
    10、访问http://devcluster01/nagios3/
    用户名:nagiosadmin密码:nagiosadmin

    监控端:
    1、apt-get install nagios-nrpe-server
    2、vi /etc/nagios/nrpe.cfg
    替换:
    g/127.0.0.1/s//192.168.1.1/g
    3、sudo /etc/init.d/nagios-nrpe-server restart

    十九、配置Splunk
    1、wget http://download.splunk.com/releases/5.0.2/splunk/linux/splunk-5.0.2-149561-Linux-x86_64.tgz
    2、tar -zxvf splunk-5.0.2-149561-Linux-x86_64.tgz
    3、cd splunk
    4、bin/splunk start --answer-yes --no-prompt --accept-license
    5、访问http://devcluster01:8000
    用户名:admin 密码:changeme
    6、添加数据 -> 从 UDP 端口 -> UDP 端口 *: 1688 -> 来源类型 从列表 log4j -> 保存
    7、配置hadoop
    vi /home/ysc/hadoop-1.1.1/conf/log4j.properties
    修改:
    log4j.rootLogger=${hadoop.root.logger}, EventCounter, SYSLOG
    增加:
    log4j.appender.SYSLOG=org.apache.log4j.net.SyslogAppender 
    log4j.appender.SYSLOG.facility=local1 
    log4j.appender.SYSLOG.layout=org.apache.log4j.PatternLayout 
    log4j.appender.SYSLOG.layout.ConversionPattern=%p %c{2}: %m%n 
    log4j.appender.SYSLOG.SyslogHost=host6:1688
    log4j.appender.SYSLOG.threshold=INFO 
    log4j.appender.SYSLOG.Header=true
    log4j.appender.SYSLOG.FacilityPrinting=true 
    8、配置hbase
    vi /home/ysc/hbase-0.92.2/conf/log4j.properties
    修改:
    log4j.rootLogger=${hbase.root.logger},SYSLOG
    增加:
    log4j.appender.SYSLOG=org.apache.log4j.net.SyslogAppender 
    log4j.appender.SYSLOG.facility=local1 
    log4j.appender.SYSLOG.layout=org.apache.log4j.PatternLayout 
    log4j.appender.SYSLOG.layout.ConversionPattern=%p %c{2}: %m%n 
    log4j.appender.SYSLOG.SyslogHost=host6:1688
    log4j.appender.SYSLOG.threshold=INFO 
    log4j.appender.SYSLOG.Header=true
    log4j.appender.SYSLOG.FacilityPrinting=true
    9、配置nutch
    vi /home/lanke/ysc/nutch-2.1-hbase/conf/log4j.properties
    修改:
    log4j.rootLogger=INFO,DRFA,SYSLOG
    增加:
    log4j.appender.SYSLOG=org.apache.log4j.net.SyslogAppender 
    log4j.appender.SYSLOG.facility=local1 
    log4j.appender.SYSLOG.layout=org.apache.log4j.PatternLayout 
    log4j.appender.SYSLOG.layout.ConversionPattern=%p %c{2}: %m%n 
    log4j.appender.SYSLOG.SyslogHost=host6:1688
    log4j.appender.SYSLOG.threshold=INFO 
    log4j.appender.SYSLOG.Header=true
    log4j.appender.SYSLOG.FacilityPrinting=true
    10、启动hadoop和hbase
    start-all.sh
    start-hbase.sh

    二十、配置Pig
    1、wget http://labs.mop.com/apache-mirror/pig/pig-0.11.0/pig-0.11.0.tar.gz
    2、tar -xzvf pig-0.11.0.tar.gz
    3、cd pig-0.11.0
    4、vi /etc/profile
    增加:
    export PIG_HOME=/home/ysc/pig-0.11.0
    export PATH=$PIG_HOME/bin:$PATH
    5、source /etc/profile
    6、cp conf/log4j.properties.template conf/log4j.properties
    7、vi conf/log4j.properties
    8、pig

    二十一、配置Hive
    1、wget http://mirrors.cnnic.cn/apache/hive/hive-0.10.0/hive-0.10.0.tar.gz
    2、tar -xzvf hive-0.10.0.tar.gz
    3、cd hive-0.10.0
    4、vi /etc/profile
    增加:
    export HIVE_HOME=/home/ysc/hive-0.10.0
    export PATH=$HIVE_HOME/bin:$PATH
    5、source /etc/profile
    6、cp conf/hive-log4j.properties.template conf/hive-log4j.properties
    7、vi conf/hive-log4j.properties
    替换:
    log4j.appender.EventCounter=org.apache.hadoop.metrics.jvm.EventCounter
    为:
    log4j.appender.EventCounter=org.apache.hadoop.log.metrics.EventCounter


    二十二、配置Hadoop2.x集群
    1、wget http://labs.mop.com/apache-mirror/hadoop/common/hadoop-2.0.2-alpha/hadoop-2.0.2-alpha.tar.gz
    2、tar -xzvf hadoop-2.0.2-alpha.tar.gz
    3、cd hadoop-2.0.2-alpha
    4、vi etc/hadoop/hadoop-env.sh
    追加:
    export JAVA_HOME=/home/ysc/jdk1.7.0_05
    export HADOOP_HEAPSIZE=2000
    5、vi etc/hadoop/core-site.xml
    <property>
    <name>fs.defaultFS</name>
    <value>hdfs://devcluster01:9000</value>
    <description>
       Where to find the Hadoop Filesystem through the network.
       Note 9000 is not the default port.
       (This is slightly changed from previous versions which didnt have "hdfs")
    </description>
    </property>
    <property>
      <name>io.file.buffer.size</name>
      <value>131072</value>
      <description>The size of buffer for use in sequence files.
      The size of this buffer should probably be a multiple of hardware
      page size (4096 on Intel x86), and it determines how much data is
      buffered during read and write operations.</description>
    </property>
    6、vi etc/hadoop/mapred-site.xml
    <property>
      <name>mapreduce.framework.name</name>
      <value>yarn</value>
    </property>

    <property>
      <name>mapred.job.reduce.input.buffer.percent</name>
      <value>1</value>
      <description>The percentage of memory- relative to the maximum heap size- to
      retain map outputs during the reduce. When the shuffle is concluded, any
      remaining map outputs in memory must consume less than this threshold before
      the reduce can begin.
      </description>
    </property>

    <property>
      <name>mapred.job.shuffle.input.buffer.percent</name>
      <value>1</value>
      <description>The percentage of memory to be allocated from the maximum heap
      size to storing map outputs during the shuffle.
      </description>
    </property>

    <property>
      <name>mapred.inmem.merge.threshold</name>
      <value>0</value>
      <description>The threshold, in terms of the number of files
      for the in-memory merge process. When we accumulate threshold number of files
      we initiate the in-memory merge and spill to disk. A value of 0 or less than
      0 indicates we want to DON'T have any threshold and instead depend only on
      the ramfs's memory consumption to trigger the merge.
      </description>
    </property>

    <property>
      <name>io.sort.factor</name>
      <value>100</value>
      <description>The number of streams to merge at once while sorting
      files.  This determines the number of open file handles.</description>
    </property>

    <property>
      <name>io.sort.mb</name>
      <value>240</value>
      <description>The total amount of buffer memory to use while sorting
      files, in megabytes.  By default, gives each merge stream 1MB, which
      should minimize seeks.</description>
    </property>
    <property>
      <name>mapred.map.output.compression.codec</name>
      <value>org.apache.hadoop.io.compress.SnappyCodec</value>
      <description>If the map outputs are compressed, how should they be
       compressed?
      </description>
    </property>

    <property>
      <name>mapred.output.compression.codec</name>
      <value>org.apache.hadoop.io.compress.SnappyCodec</value>
      <description>If the job outputs are compressed, how should they be compressed?
      </description>
    </property>
    <property>
      <name>mapred.output.compression.type</name>
      <value>BLOCK</value>
      <description>If the job outputs are to compressed as SequenceFiles, how should
       they be compressed? Should be one of NONE, RECORD or BLOCK.
      </description>
    </property>
    <property>
      <name>mapred.child.java.opts</name>
      <value>-Xmx2000m</value>
    </property>

    <property>
      <name>mapred.output.compress</name>
      <value>true</value>
      <description>Should the job outputs be compressed?
      </description>
    </property>

    <property>
      <name>mapred.compress.map.output</name>
      <value>true</value>
      <description>Should the outputs of the maps be compressed before being
       sent across the network. Uses SequenceFile compression.
      </description>
    </property>

    <property>
      <name>mapred.tasktracker.map.tasks.maximum</name>
      <value>5</value>
    </property>

    <property>
      <name>mapred.map.tasks</name>
      <value>15</value>
    </property>

    <property>
      <name>mapred.tasktracker.reduce.tasks.maximum</name>
      <value>5</value>
    <description>
    define mapred.map tasks to be number of slave hosts.the best number is the  number of slave hosts plus the core numbers of per host
    </description>
    </property>

    <property>
      <name>mapred.reduce.tasks</name>
      <value>15</value>
      <description>
    define mapred.reduce tasks to be number of slave hosts.the best number is the  number of slave hosts plus the core numbers of per host
      </description>
    </property>
    <property>
      <name>mapred.system.dir</name>
      <value>/home/ysc/mapreduce/system</value>
    </property>

    <property>
      <name>mapred.local.dir</name>
      <value>/home/ysc/mapreduce/local</value>
    </property>

    <property>
      <name>mapreduce.job.counters.max</name>
      <value>12000</value>
      <description>Limit on the number of counters allowed per job.
      </description>
    </property>
    7、vi etc/hadoop/yarn-site.xml
    <property>   
    <name>yarn.resourcemanager.resource-tracker.address</name>  
    <value>devcluster01:8031</value>
    </property>  
    <property> 
    <name>yarn.resourcemanager.address</name>    
    <value>devcluster01:8032</value> 
    </property>
    <property>   
    <name>yarn.resourcemanager.scheduler.address</name> 
    <value>devcluster01:8030</value>
    </property>
    <property> 
    <name>yarn.resourcemanager.admin.address</name> 
    <value>devcluster01:8033</value>  
    </property>  
    <property>   
    <name>yarn.resourcemanager.webapp.address</name>   
    <value>devcluster01:8088</value> 
    </property> 
    <property>  
    <description>Classpath for typical applications.</description>
    <name>yarn.application.classpath</name> 
    <value>      
    $HADOOP_CONF_DIR,     
    $HADOOP_COMMON_HOME/*,$HADOOP_COMMON_HOME/lib/*,   
    $HADOOP_HDFS_HOME/*,$HADOOP_HDFS_HOME/lib/*,      
    $HADOOP_MAPRED_HOME/*,$HADOOP_MAPRED_HOME/lib/*,  
    $YARN_HOME/*,$YARN_HOME/lib/*  
    </value> 
    </property>
    <property> 
    <name>yarn.nodemanager.aux-services</name> 
    <value>mapreduce.shuffle</value> 
    </property>  
    <property>   
    <name>yarn.nodemanager.aux-services.mapreduce.shuffle.class</name> 
    <value>org.apache.hadoop.mapred.ShuffleHandler</value> 
    </property> 
    <property>  
    <name>yarn.nodemanager.local-dirs</name> <value>/home/ysc/h2/data/1/yarn/local,/home/ysc/h2/data/2/yarn/local,/home/ysc/h2/data/3/yarn/local</value> 
    </property>
    <property>
    <name>yarn.nodemanager.log-dirs</name>  <value>/home/ysc/h2/data/1/yarn/logs,/home/ysc/h2/data/2/yarn/logs,/home/ysc/h2/data/3/yarn/logs</value> 
    </property> 
    <property>  
    <description>Where to aggregate logs</description>
    <name>yarn.nodemanager.remote-app-log-dir</name>   
    <value>/home/ysc/h2/var/log/hadoop-yarn/apps</value>
    </property>
    <property>   
    <name>mapreduce.jobhistory.address</name>  
    <value>devcluster01:10020</value>
    </property>  
    <property>   
    <name>mapreduce.jobhistory.webapp.address</name>  
    <value>devcluster01:19888</value>
    </property>  
    8、vi etc/hadoop/hdfs-site.xml
    <property> 
    <name>dfs.permissions.superusergroup</name> 
    <value>root</value>
    </property>
    <property>
      <name>dfs.name.dir</name>
      <value>/home/ysc/dfs/filesystem/name</value>
    </property>
    <property>
      <name>dfs.data.dir</name>
      <value>/home/ysc/dfs/filesystem/data</value>
    </property>
    <property>
      <name>dfs.replication</name>
      <value>3</value>
    </property>
    <property>
      <name>dfs.block.size</name>
      <value>6710886400</value>
      <description>The default block size for new files.</description>
    </property>
    9、启动hadoop
    bin/hdfs namenode -format
    sbin/start-dfs.sh
    sbin/start-yarn.sh
    10、访问管理页面
    http://devcluster01:8088
    http://devcluster01:50070 Nutch相关框架安装使用最佳指南.zip (102.7 KB) 下载次数: 1