什么是Storm?

Storm is a free and open source distributed realtime computation system. Storm makes it easy to reliably process unbounded streams of data, doing for realtime processing what Hadoop did for batch processing. Storm is simple, can be used with any programming language, and is a lot of fun to use!

Storm has many use cases: realtime analytics, online machine learning, continuous computation, distributed RPC, ETL, and more. Storm is fast: a benchmark clocked it at over a million tuples processed per second per node. It is scalable, fault-tolerant, guarantees your data will be processed, and is easy to set up and operate.

[官网] http://storm-project.net/

安装脚本

安装daemontools-0.76.tar.gz的脚本:

yum install patch
sudo mkdir -p /package
sudo chmod 1755 /package/
cd /package/
sudo wget http://cr.yp.to/daemontools/daemontools-0.76.tar.gz
sudo tar xzf daemontools-0.76.tar.gz
sudo wget http://www.qmail.org/moni.csi.hu/pub/glibc-2.3.1/daemontools-0.76.errno.patch
cd admin/daemontools-0.76
sudo patch -p1 < http://www.cnblogs.com/daemontools-0.76.errno.patch
sudo rm http://www.cnblogs.com/daemontools-0.76.errno.patch http://www.cnblogs.com/daemontools-0.76.tar.gz
sudo ./package/install

安装Storm的脚本是在 nathanmarz 写的脚本的基础上修改的,去掉了两个大文件的下载,依赖项修改为yum安装,代码略长

#!/bin/bash

pp() {
    echo -e "\e[00;32m"$1"\e[00m"
}

HOST=`hostname`

#########################################
# Clean up old installation.
#########################################

cleanup() {
    pp "Cleaning up previous installation..."
    rm -rf $BASEDIR
    mkdir $BASEDIR
    echo "#!/bin/bash" > $START_SH
    chmod +x $START_SH
    echo "#!/bin/bash" > $STOP_SH
    chmod +x $STOP_SH
}

#########################################
# System dependencies.
#########################################

deps() {
    pp "Checking system dependencies..."
    echo
       yum  install gcc gcc-c++.x86_64  crontabs  screen daemontools uuid-devel libuuid-devel git libtool build-essential  unzip
    echo
}

#########################################
# ZooKeeper.
#########################################

zookeeper() {
    echo "Current Host $HOST "

#    if [ "$HOST" != "$NIMBUS" ]
#    then
#        pp "Skipping ZooKeeper installation on all hosts but nimbus!"
#        return
#    fi

    ZK_VERSION="3.3.6"
    ZK_DIR=$BASEDIR"/zookeeper"
    ZK_CONFIGFILE="zoo.conf"
    ZK_CONF=$ZK_DIR"/"$ZK_CONFIGFILE
    ZK_RUN=$ZK_DIR"/run"
    ZK_PURGE=$ZK_DIR"/purge.sh"
    ZK_DATADIR=$ZK_DIR"/data"
    ZK_TARBALL_URL="http://apache.openmirror.de/zookeeper/zookeeper-"$ZK_VERSION"/zookeeper-"$ZK_VERSION".tar.gz"
    ZK_TARBALL=$ZK_DIR/"zookeeper.tar.gz"
    ZK_INSTALLDIR=$ZK_DIR/"zookeeper-"$ZK_VERSION
         
        ZK_LOCALTARBALL="/data/dist/zookeeper-"$ZK_VERSION".tar.gz"

        ZK_FILENAME=$BASEDIR"/zookeeper/zookeeper-"$ZK_VERSION.tar.gz
    pp "Installing ZooKeeper "$ZK_VERSION" on nimbus host '"$HOST"'..."

    mkdir $ZK_DIR &>/dev/null
    mkdir $ZK_DATADIR &>/dev/null

        echo "$ZK_FILENAME"
         
        if [ -f $ZK_LOCALTARBALL ]; then
         
       pp "ZooKeeper Copy From /data/dist/."
          cp $ZK_LOCALTARBALL $ZK_TARBALL
        else 
     pp "Downloading ZooKeeper..."
         wget $ZK_TARBALL_URL -q -O $ZK_TARBALL
        fi

    tar xzf $ZK_TARBALL -C $ZK_DIR
    rm $ZK_TARBALL

    pp "Configuring ZooKeeper..."

    # Cluster config.
    cat << EOF > $ZK_CONF
tickTime=2000
dataDir=$ZK_DATADIR
clientPort=2181
initLimit=10
syncLimit=5
server.1=192.168.0.101:2888:3888
server.2=192.168.0.102:2888:3888
server.3=192.168.0.103:2888:3888
EOF

#Read the myid
read  -p "Please select the myid for this instance:  " MYID 
if [ ! `echo $MYID | egrep "^[0-9]+\$"`  ] ; then
        echo "                       _  _                     "  
        echo "                _     ( \/ )                    "
        echo "                |              |               "
        echo "                                                "

        echo " Are you kidding ? That is number? Did you learnt math from P.E teacher?  Selecting default: 1"
                MYID=1 
                fi
    # This host's id.
    echo $MYID > $ZK_DATADIR/myid

    # Run script.
    ZK_CP=$ZK_INSTALLDIR/zookeeper-$ZK_VERSION.jar:$ZK_INSTALLDIR/lib/log4j-1.2.15.jar:$ZK_INSTALLDIR/conf
    cat << EOF > $ZK_RUN
#!/bin/bash
_JAVA_OPTIONS="-Xmx1024M -Xms1024M"
java -cp $ZK_CP org.apache.zookeeper.server.quorum.QuorumPeerMain $ZK_CONFIGFILE
EOF
    chmod +x $ZK_RUN

    # Purge script to cleanup zookeeper log files.
    cat << EOF > $ZK_PURGE
mkdir $ZK_DIR/snap
java -cp $ZK_CP org.apache.zookeeper.server.PurgeTxnLog $ZK_DATADIR $ZK_DIR/snap -n 3
rm -r $ZK_DIR/snap
EOF
    chmod +x $ZK_PURGE

    # Run purge.sh via cron job.
    echo "@hourly $ZK_PURGE" | crontab -

    # Update global start/stop scripts.
    echo "supervise $ZK_DIR &" >> $START_SH
    echo "svc -x $ZK_DIR" >> $STOP_SH
}

#########################################
# Storm dependency: ZeroMQ
#########################################

zeromq() {
    ZMQ_VERSION="2.1.7"
    ZMQ_DIR=$BASEDIR"/zeromq"
    ZMQ_TARBALL_URL="http://download.zeromq.org/zeromq-"$ZMQ_VERSION".tar.gz"
    ZMQ_TARBALL=$ZMQ_DIR"/zeromq.tar.gz"

    pp "Installing ZeroMQ "$ZMQ_VERSION" (storm dependency)..."
    mkdir $ZMQ_DIR

    pp "Downloading ZeroMQ..."
    wget $ZMQ_TARBALL_URL -q -O $ZMQ_TARBALL
    tar zxf $ZMQ_TARBALL -C $ZMQ_DIR
    rm $ZMQ_TARBALL

    pp "Compiling ZeroMQ..."
    echo
    pushd $ZMQ_DIR/zeromq-$ZMQ_VERSION
    ./configure && make && sudo make install
    popd
    echo
}

#########################################
# Storm dependency 2: JZMQ,
# Java bindings for ZeroMQ.
#
# This is where things get tricky.
# Despite the warning on nathanmarz' page,
# we use mainline git here, as it compiles
# with the latest autoconf and libtool on
# Ubuntu 12.04.
#########################################

jzmq() {
    JZMQ_DIR=$BASEDIR"/jzmq"
#    JZMQ_REPO="https://githiub.com/zeromq/jzmq.git"
#    JZMQ_COMMIT="e2dd66"

    pp "Installing JZMQ (Java bindings for ZeroMQ) from Github..."

    git clone -q https://github.com/nathanmarz/jzmq.git $JZMQ_DIR

    pp "Compiling JZMQ..."

    echo
    pushd $JZMQ_DIR
    git checkout $JZMQ_COMMIT
    ./autogen.sh && ./configure --with-zeromq=/usr/local/lib && make && sudo make install
    popd
    echo
}

#########################################
# Storm itself.
#########################################

storm() {
    STORM_VERSION="0.8.1"
    STORM_DIR=$BASEDIR"/storm"
    STORM_ZIP_URL="https://github.com/downloads/nathanmarz/storm/storm-"$STORM_VERSION".zip"
    STORM_ZIP=$STORM_DIR"/storm.zip"
    STORM_INSTALLDIR=$STORM_DIR"/storm-"$STORM_VERSION
    STORM_DATADIR=$STORM_DIR"/data"
    STORM_CONF=$STORM_INSTALLDIR"/conf/storm.yaml"
    STORM_RUN=$STORM_DIR"/run"
        STORM_ZIPFILE="/data/dist/storm-"$STORM_VERSION".zip"

    pp "Installing Storm "$STORM_VERSION"..."
    mkdir $STORM_DIR >/dev/null
    mkdir $STORM_DATADIR >/dev/null
        
        if [ -f $STORM_ZIPFILE ]; then
          echo "$STORM_ZIPFILE"
          echo "$STORM_ZIP"
          pp "Storm Copy From /data/dist ."
          cp $STORM_ZIPFILE $STORM_ZIP
        else
          pp "Downloading Storm..."
          wget $STORM_ZIP_URL -q -O $STORM_ZIP        
        fi

    
    unzip -qq $STORM_ZIP -d $STORM_DIR
    rm $STORM_ZIP





    pp "Configuring Storm..."
    echo "storm.local.dir: \""$STORM_DATADIR"\"" > $STORM_CONF
    echo "storm.zookeeper.servers:" >> $STORM_CONF
    echo " - \"192.168.10.101\"" >> $STORM_CONF
    echo " - \"192.168.10.102\"" >> $STORM_CONF
    echo " - \"192.168.10.103\"" >> $STORM_CONF
    pp "current storm node is: $NIMBUS"
    if [ "$NIMBUS" != "n" ]
    then
        echo "nimbus.host: \"192.168.10.101\"" >> $STORM_CONF
    fi

    echo "storm.zookeeper.port: 2181" >> $STORM_CONF

    # Supervisor directories/scripts + global start/stop scripts.
    # Note: If we're NIMBUS, we run the 'nimbis' action instead.
    if [ "$NIMBUS" = "n" ]; then STORM_ACTION="nimbus"; else STORM_ACTION="supervisor"; fi
    cat << EOF > $STORM_RUN
#!/bin/bash
$STORM_INSTALLDIR/bin/storm $STORM_ACTION
EOF
    chmod +x $STORM_RUN
    echo "supervise $STORM_DIR &" >> $START_SH
    echo "svc -x $STORM_DIR" >> $STOP_SH
}

#########################################
# Main app.
#########################################

PHASES=("cleanup" "deps" "zookeeper" "zeromq" "jzmq" "storm")

execute() {
    case "$1" in
    "0")
        cleanup
        ;;
    "1")
        deps
        ;;
    "2")
        zookeeper
        ;;
    "3")
        zeromq
        ;;
    "4")
        jzmq
        ;;
    "5")
        storm
        ;;
    esac
}

if [ $# -eq 3 ]
then

    NIMBUS=$2
    BASEDIR=$3
    START_SH=$BASEDIR"/start.sh"
    STOP_SH=$BASEDIR"/stop.sh"

    if [ "$1" = "all" ]
    then
        # Run everything.
        for ((p=0;p<${#PHASES[@]};p++))
        do
            execute $p
        done

        pp "Installation complete."
        pp "Be sure to carefully read the log."
        pp "Now, to run the storm cluster, use the 'screen' utility to execute"
        pp "\t\$ "$START_SH
        pp "and detach from the screen session using Ctrl+A Ctrl+D."
    else
        execute $1

        pp "Phase installation complete."
    fi
else
    echo "Usage: ./install_storm <number_of_phase>/all n/s <installdir>"
    echo "   "
    echo "Options:    "
    echo "n/s  n-> nimbus s-> supervisor"
    echo " "
    echo "Phases:"
    for ((i=0;i<${#PHASES[@]};i++))
    do
        echo -e "\t"$i": "${PHASES[$i]}
    done
fi

分集剧情:

To install ZeroMQ, run:

wget http://download.zeromq.org/historic/zeromq-2.1.7.tar.gz
tar -xzf zeromq-2.1.7.tar.gz
cd zeromq-2.1.7
./configure
make
sudo make install

To install JZMQ, run:

git clone https://github.com/nathanmarz/jzmq.git
cd jzmq
./autogen.sh
./configure
make
sudo make install

Install Storm on CentOS

什么是Storm?

安装脚本

热门文章

最新文章

相关课程

相关电子书

相关实验场景

推荐镜像