大数据工具篇(一)-yiflume1.4-安装部署指南

浏览: 1734

一、引言

  flume-ng是一个分布式、高可靠和高效的日志收集系统,flume-ng是flume的新版本的意思,其中“ng”意为new generate(新一代),目前来说,flume-ng 1.4是最新的版本。flume-ng与flume相比,发生了很大的变化,因为之前一直在flume0.9的版本,一直没有升级到flume-ng,最近因为项目需要,做了一次升级,发现了一些问题,特记录下来,分享给大家。

二、版本说明

  flume-ng 1.4.0

三、安装步骤

  下载、解压、安装JDK、设置环境变量部分已经有很多介绍性的问题,不做说明。需要特别说明之处的是,flume-ng不需要要zookeeper,无需设置。

四、flume-ng bug  

  安装完成后运行flume-ng会出现错误信息,这主要是因为shell脚本的问题,我将修改后的flume-ng完整的上传如下,其中标注:#zhangzl下面的行是需要修改的部分。完整脚本如下所示:  

#!/bin/bash
#
#
# Licensed to the Apache Software Foundation (ASF) under one
# or more contributor license agreements.  See the NOTICE file
# distributed with this work for additional information
# regarding copyright ownership.  The ASF licenses this file
# to you under the Apache License, Version 2.0 (the
# "License"); you may not use this file except in compliance
# with the License.  You may obtain a copy of the License at
#
#   http://www.apache.org/licenses/LICENSE-2.0
#
# Unless required by applicable law or agreed to in writing,
# software distributed under the License is distributed on an
# "AS IS" BASIS, WITHOUT WARRANTIES OR CONDITIONS OF ANY
# KIND, either express or implied.  See the License for the
# specific language governing permissions and limitations
# under the License.
#
################################
# constants
################################
FLUME_AGENT_CLASS="org.apache.flume.node.Application"
FLUME_AVRO_CLIENT_CLASS="org.apache.flume.client.avro.AvroCLIClient"
FLUME_VERSION_CLASS="org.apache.flume.tools.VersionInfo"
FLUME_TOOLS_CLASS="org.apache.flume.tools.FlumeToolsMain"
CLEAN_FLAG=1
################################
# functions
################################
info() {
  if [ ${CLEAN_FLAG} -ne 0 ]; then
    local msg=$1
    echo "Info: $msg" >&2
  fi
}
warn() {
  if [ ${CLEAN_FLAG} -ne 0 ]; then
    local msg=$1
    echo "Warning: $msg" >&2
  fi
}
error() {
  local msg=$1
  local exit_code=$2
  echo "Error: $msg" >&2
  if [ -n "$exit_code" ] ; then
    exit $exit_code
  fi
}
# If avail, add Hadoop paths to the FLUME_CLASSPATH and to the
# FLUME_JAVA_LIBRARY_PATH env vars.
# Requires Flume jars to already be on FLUME_CLASSPATH.
add_hadoop_paths() {
  local HADOOP_IN_PATH=$(PATH="${HADOOP_HOME:-${HADOOP_PREFIX}}/bin:$PATH" \
      which hadoop 2>/dev/null)
  if [ -f "${HADOOP_IN_PATH}" ]; then
    info "Including Hadoop libraries found via ($HADOOP_IN_PATH) for HDFS access"
    # determine hadoop java.library.path and use that for flume
    local HADOOP_CLASSPATH=""
    local HADOOP_JAVA_LIBRARY_PATH=$(HADOOP_CLASSPATH="$FLUME_CLASSPATH" \
        ${HADOOP_IN_PATH} org.apache.flume.tools.GetJavaProperty \
        java.library.path)
    # look for the line that has the desired property value
    # (considering extraneous output from some GC options that write to stdout)
    # IFS = InternalFieldSeparator (set to recognize only newline char as delimiter)
    IFS=$'\n'
    for line in $HADOOP_JAVA_LIBRARY_PATH; do
      #if [[ $line =~ ^java\.library\.path=(.*)$ ]]; then
      if [[ "$line" =~ "^java\.library\.path=(.*)$" ]]; then
        HADOOP_JAVA_LIBRARY_PATH=${BASH_REMATCH[1]}
        break
      fi
    done
    unset IFS
    if [ -n "${HADOOP_JAVA_LIBRARY_PATH}" ]; then
      FLUME_JAVA_LIBRARY_PATH="$FLUME_JAVA_LIBRARY_PATH:$HADOOP_JAVA_LIBRARY_PATH"
    fi
    # determine hadoop classpath
    HADOOP_CLASSPATH=$($HADOOP_IN_PATH classpath)
    # hack up and filter hadoop classpath
    local ELEMENTS=$(sed -e 's/:/ /g' <<<${HADOOP_CLASSPATH})
    local ELEMENT
    for ELEMENT in $ELEMENTS; do
      local PIECE
      for PIECE in $(echo $ELEMENT); do
          #zhangzl
        if [[ $PIECE =~ "slf4j-(api|log4j12).*\.jar" ]]; then
          info "Excluding $PIECE from classpath"
          continue
        else
          FLUME_CLASSPATH="$FLUME_CLASSPATH:$PIECE"
        fi
      done
    done
  fi
}
add_HBASE_paths() {
  local HBASE_IN_PATH=$(PATH="${HBASE_HOME}/bin:$PATH" \
      which hbase 2>/dev/null)
  if [ -f "${HBASE_IN_PATH}" ]; then
    info "Including HBASE libraries found via ($HBASE_IN_PATH) for HBASE access"
    # determine HBASE java.library.path and use that for flume
    local HBASE_CLASSPATH=""
    local HBASE_JAVA_LIBRARY_PATH=$(HBASE_CLASSPATH="$FLUME_CLASSPATH" \
        ${HBASE_IN_PATH} org.apache.flume.tools.GetJavaProperty \
        java.library.path)
    # look for the line that has the desired property value
    # (considering extraneous output from some GC options that write to stdout)
    # IFS = InternalFieldSeparator (set to recognize only newline char as delimiter)
    IFS=$'\n'
    for line in $HBASE_JAVA_LIBRARY_PATH; do
    #zhangzl
      if [[ $line =~ "^java\.library\.path=(.*)$" ]]; then
        HBASE_JAVA_LIBRARY_PATH=${BASH_REMATCH[1]}
        break
      fi
    done
    unset IFS
    if [ -n "${HBASE_JAVA_LIBRARY_PATH}" ]; then
      FLUME_JAVA_LIBRARY_PATH="$FLUME_JAVA_LIBRARY_PATH:$HBASE_JAVA_LIBRARY_PATH"
    fi
    # determine HBASE classpath
    HBASE_CLASSPATH=$($HBASE_IN_PATH classpath)
    # hack up and filter HBASE classpath
    local ELEMENTS=$(sed -e 's/:/ /g' <<<${HBASE_CLASSPATH})
    local ELEMENT
    for ELEMENT in $ELEMENTS; do
      local PIECE
      for PIECE in $(echo $ELEMENT); do
      #zhangzl
        if [[ $PIECE =~ "slf4j-(api|log4j12).*\.jar" ]]; then
          info "Excluding $PIECE from classpath"
          continue
        else
          FLUME_CLASSPATH="$FLUME_CLASSPATH:$PIECE"
        fi
      done
    done
    FLUME_CLASSPATH="$FLUME_CLASSPATH:$HBASE_HOME/conf"
  fi
}
set_LD_LIBRARY_PATH(){
#Append the FLUME_JAVA_LIBRARY_PATH to whatever the user may have specified in
#flume-env.sh
  if [ -n "${FLUME_JAVA_LIBRARY_PATH}" ]; then
    export LD_LIBRARY_PATH="${LD_LIBRARY_PATH}:${FLUME_JAVA_LIBRARY_PATH}"
  fi
}
display_help() {
  cat <<EOF
Usage: $0 <command> [options]...
commands:
  help                  display this help text
  agent                 run a Flume agent
  avro-client           run an avro Flume client
  version               show Flume version info
global options:
  --conf,-c <conf>      use configs in <conf> directory
  --classpath,-C <cp>   append to the classpath
  --dryrun,-d           do not actually start Flume, just print the command
  --plugins-path <dirs> colon-separated list of plugins.d directories. See the
                        plugins.d section in the user guide for more details.
                        Default: \$FLUME_HOME/plugins.d
  -Dproperty=value      sets a Java system property value
  -Xproperty=value      sets a Java -X option
agent options:
  --conf-file,-f <file> specify a config file (required)
  --name,-n <name>      the name of this agent (required)
  --help,-h             display help text
avro-client options:
  --rpcProps,-P <file>   RPC client properties file with server connection params
  --host,-H <host>       hostname to which events will be sent
  --port,-p <port>       port of the avro source
  --dirname <dir>        directory to stream to avro source
  --filename,-F <file>   text file to stream to avro source (default: std input)
  --headerFile,-R <file> File containing event headers as key/value pairs on each new line
  --help,-h              display help text
  Either --rpcProps or both --host and --port must be specified.
Note that if <conf> directory is specified, then it is always included first
in the classpath.
EOF
}
run_flume() {
  local FLUME_APPLICATION_CLASS
  if [ "$#" -gt 0 ]; then
    FLUME_APPLICATION_CLASS=$1
    shift
  else
    error "Must specify flume application class" 1
  fi
  if [ ${CLEAN_FLAG} -ne 0 ]; then
    set -x
  fi
  $EXEC $JAVA_HOME/bin/java $JAVA_OPTS -cp "$FLUME_CLASSPATH" \
      -Djava.library.path=$FLUME_JAVA_LIBRARY_PATH "$FLUME_APPLICATION_CLASS" $*
}
################################
# main
################################
# set default params
FLUME_CLASSPATH=""
FLUME_JAVA_LIBRARY_PATH=""
JAVA_OPTS="-Xmx20m"
LD_LIBRARY_PATH=""
opt_conf=""
opt_classpath=""
opt_plugins_dirs=""
opt_java_props=""
opt_dryrun=""
mode=$1
shift
case "$mode" in
  help)
    display_help
    exit 0
    ;;
  agent)
    opt_agent=1
    ;;
  node)
    opt_agent=1
    warn "The \"node\" command is deprecated. Please use \"agent\" instead."
    ;;
  avro-client)
    opt_avro_client=1
    ;;
  tool)
    opt_tool=1
    ;;
  version)
   opt_version=1
   CLEAN_FLAG=0
   ;;
  *)
    error "Unknown or unspecified command '$mode'"
    echo
    display_help
    exit 1
    ;;
esac
args=""
while [ -n "$*" ] ; do
  arg=$1
  shift
  case "$arg" in
    --conf|-c)
      [ -n "$1" ] || error "Option --conf requires an argument" 1
      opt_conf=$1
      shift
      ;;
    --classpath|-C)
      [ -n "$1" ] || error "Option --classpath requires an argument" 1
      opt_classpath=$1
      shift
      ;;
    --dryrun|-d)
      opt_dryrun="1"
      ;;
    --plugins-path)
      opt_plugins_dirs=$1
      shift
      ;;
    -D*)
      opt_java_props="$opt_java_props $arg"
      ;;
    -X*)
      opt_java_props="$opt_java_props $arg"
      ;;
    *)
      args="$args $arg"
      ;;
  esac
done
# make opt_conf absolute
if [[ -n "$opt_conf" && -d "$opt_conf" ]]; then
  opt_conf=$(cd $opt_conf; pwd)
fi
# allow users to override the default env vars via conf/flume-env.sh
if [ -z "$opt_conf" ]; then
  warn "No configuration directory set! Use --conf <dir> to override."
elif [ -f "$opt_conf/flume-env.sh" ]; then
  info "Sourcing environment configuration script $opt_conf/flume-env.sh"
  source "$opt_conf/flume-env.sh"
fi
# append command-line java options to stock or env script JAVA_OPTS
if [ -n "${opt_java_props}" ]; then
  JAVA_OPTS="${JAVA_OPTS} ${opt_java_props}"
fi
# prepend command-line classpath to env script classpath
if [ -n "${opt_classpath}" ]; then
  if [ -n "${FLUME_CLASSPATH}" ]; then
    FLUME_CLASSPATH="${opt_classpath}:${FLUME_CLASSPATH}"
  else
    FLUME_CLASSPATH="${opt_classpath}"
  fi
fi
if [ -z "${FLUME_HOME}" ]; then
  FLUME_HOME=$(cd $(dirname $0)/..; pwd)
fi
# prepend $FLUME_HOME/lib jars to the specified classpath (if any)
if [ -n "${FLUME_CLASSPATH}" ] ; then
  FLUME_CLASSPATH="${FLUME_HOME}/lib/*:$FLUME_CLASSPATH"
else
  FLUME_CLASSPATH="${FLUME_HOME}/lib/*"
fi
# load plugins.d directories
PLUGINS_DIRS=""
if [ -n "${opt_plugins_dirs}" ]; then
  PLUGINS_DIRS=$(sed -e 's/:/ /g' <<<${opt_plugins_dirs})
else
  PLUGINS_DIRS="${FLUME_HOME}/plugins.d"
fi
unset plugin_lib plugin_libext plugin_native
for PLUGINS_DIR in $PLUGINS_DIRS; do
  if [[ -d ${PLUGINS_DIR} ]]; then
    for plugin in ${PLUGINS_DIR}/*; do
      if [[ -d "$plugin/lib" ]]; then
        plugin_lib="${plugin_lib}${plugin_lib+:}${plugin}/lib/*"
      fi
      if [[ -d "$plugin/libext" ]]; then
        plugin_libext="${plugin_libext}${plugin_libext+:}${plugin}/libext/*"
      fi
      if [[ -d "$plugin/native" ]]; then
        plugin_native="${plugin_native}${plugin_native+:}${plugin}/native"
      fi
    done
  fi
done
if [[ -n "${plugin_lib}" ]]
then
  FLUME_CLASSPATH="${FLUME_CLASSPATH}:${plugin_lib}"
fi
if [[ -n "${plugin_libext}" ]]
then
  FLUME_CLASSPATH="${FLUME_CLASSPATH}:${plugin_libext}"
fi
if [[ -n "${plugin_native}" ]]
then
  if [[ -n "${FLUME_JAVA_LIBRARY_PATH}" ]]
  then
    FLUME_JAVA_LIBRARY_PATH="${FLUME_JAVA_LIBRARY_PATH}:${plugin_native}"
  else
    FLUME_JAVA_LIBRARY_PATH="${plugin_native}"
  fi
fi
# find java
if [ -z "${JAVA_HOME}" ] ; then
  warn "JAVA_HOME is not set!"
  # Try to use Bigtop to autodetect JAVA_HOME if it's available
  if [ -e /usr/libexec/bigtop-detect-javahome ] ; then
    . /usr/libexec/bigtop-detect-javahome
  elif [ -e /usr/lib/bigtop-utils/bigtop-detect-javahome ] ; then
    . /usr/lib/bigtop-utils/bigtop-detect-javahome
  fi
  # Using java from path if bigtop is not installed or couldn't find it
  if [ -z "${JAVA_HOME}" ] ; then
    JAVA_DEFAULT=$(type -p java)
    [ -n "$JAVA_DEFAULT" ] || error "Unable to find java executable. Is it in your PATH?" 1
    JAVA_HOME=$(cd $(dirname $JAVA_DEFAULT)/..; pwd)
  fi
fi
# look for hadoop libs
add_hadoop_paths
add_HBASE_paths
# prepend conf dir to classpath
if [ -n "$opt_conf" ]; then
  FLUME_CLASSPATH="$opt_conf:$FLUME_CLASSPATH"
fi
set_LD_LIBRARY_PATH
# allow dryrun
EXEC="exec"
if [ -n "${opt_dryrun}" ]; then
  warn "Dryrun mode enabled (will not actually initiate startup)"
  EXEC="echo"
fi
# finally, invoke the appropriate command
if [ -n "$opt_agent" ] ; then
  run_flume $FLUME_AGENT_CLASS $args
elif [ -n "$opt_avro_client" ] ; then
  run_flume $FLUME_AVRO_CLIENT_CLASS $args
elif [ -n "${opt_version}" ] ; then
  run_flume $FLUME_VERSION_CLASS $args
elif [ -n "${opt_tool}" ] ; then
  run_flume $FLUME_TOOLS_CLASS $args
else
  error "This message should never appear" 1
fi
exit 0

五、测试配置文件

  在conf目录下创建example-conf.properties文件,属性如下所示:

# Describe the source 
a1.sources = r1
a1.sinks = k1
a1.channels = c1
# Describe/configure the source
a1.sources.r1.type = avro
a1.sources.r1.bind = localhost
a1.sources.r1.port = 44444
# Describe the sink
# 将数据输出至日志中
a1.sinks.k1.type = logger
# Use a channel which buffers events in memory
a1.channels.c1.type = memory
a1.channels.c1.capacity = 1000
a1.channels.c1.transactionCapacity = 100
# Bind the source and sink to the channel
a1.sources.r1.channels = c1
a1.sinks.k1.channel = c1

六、运行命令

6.1 启动代理

[hadoop@hadoop1 conf]$ flume-ng agent -n a1 -f example-conf.properties

6.2 启动avro-client客户端向agent代理发送数据-需要单独启动新的窗口

[hadoop@hadoop1 conf]$ flume-ng avro-client -H localhost -p 44444 -F file01

七、结果查看

14/01/16 22:26:34 INFO ipc.NettyServer: [id: 0x0100c7e4, /127.0.0.1:54289 => /127.0.0.1:44444] OPEN
14/01/16 22:26:34 INFO ipc.NettyServer: [id: 0x0100c7e4, /127.0.0.1:54289 => /127.0.0.1:44444] BOUND: /127.0.0.1:44444
14/01/16 22:26:34 INFO ipc.NettyServer: [id: 0x0100c7e4, /127.0.0.1:54289 => /127.0.0.1:44444] CONNECTED: /127.0.0.1:54289
14/01/16 22:26:36 INFO ipc.NettyServer: [id: 0x0100c7e4, /127.0.0.1:54289 :> /127.0.0.1:44444] DISCONNECTED
14/01/16 22:26:36 INFO ipc.NettyServer: [id: 0x0100c7e4, /127.0.0.1:54289 :> /127.0.0.1:44444] UNBOUND
14/01/16 22:26:36 INFO ipc.NettyServer: [id: 0x0100c7e4, /127.0.0.1:54289 :> /127.0.0.1:44444] CLOSED
14/01/16 22:26:36 INFO ipc.NettyServer: Connection to /127.0.0.1:54289 disconnected.
14/01/16 22:26:38 INFO sink.LoggerSink: Event: { headers:{} body: 68 65 6C 6C 6F 20 77 6F 72 6C 64                hello world }
推荐 0
本文由 张子良 创作,采用 知识共享署名-相同方式共享 3.0 中国大陆许可协议 进行许可。
转载、引用前需联系作者,并署名作者且注明文章出处。
本站文章版权归原作者及原出处所有 。内容为作者个人观点, 并不代表本站赞同其观点和对其真实性负责。本站是一个个人学习交流的平台,并不用于任何商业目的,如果有任何问题,请及时联系我们,我们将根据著作权人的要求,立即更正或者删除有关内容。本站拥有对此声明的最终解释权。

0 个评论

要回复文章请先登录注册