1.什么是flume
2.flume的官方網站在哪里？3.flume有哪些術語？4.如何配置flume數據源碼？ 一、什么是Flume? flume 作為 cloudera 開發的實時日志收集系統，受到了業界的認可與廣泛應用。Flume 初始的發行版本目前被統稱為 Flume OG（original generation），屬于 cloudera。但隨著 FLume 功能的擴展，Flume OG 代碼工程臃腫、核心組件設計不合理、核心配置不標準等缺點暴露出來，尤其是在 Flume OG 的最后一個發行版本 0.94.0 中，日志傳輸不穩定的現象尤為嚴重，為了解決這些問題，2011 年 10 月 22 號，cloudera 完成了 Flume-728，對 Flume 進行了里程碑式的改動：重構核心組件、核心配置以及代碼架構，重構后的版本統稱為 Flume NG（next generation）；改動的另一原因是將 Flume 納入 apache 旗下，cloudera Flume 改名為 Apache Flume。 flume的特點： flume是一個分布式、可靠、和高可用的海量日志采集、聚合和傳輸的系統。支持在日志系統中定制各類數據發送方，用于收集數據;同時，Flume提供對數據進行簡單處理，并寫到各種數據接受方(比如文本、HDFS、Hbase等)的能力 。 flume的數據流由事件(Event)貫穿始終。事件是Flume的基本數據單位，它攜帶日志數據(字節數組形式)并且攜帶有頭信息，這些Event由Agent外部的Source生成，當Source捕獲事件后會進行特定的格式化，然后Source會把事件推入(單個或多個)Channel中。你可以把Channel看作是一個緩沖區，它將保存事件直到Sink處理完該事件。Sink負責持久化日志或者把事件推向另一個Source。 flume的可靠性 當節點出現故障時，日志能夠被傳送到其他節點上而不會丟失。Flume提供了三種級別的可靠性保障，從強到弱依次分別為：end-to-end（收到數據agent首先將event寫到磁盤上，當數據傳送成功后，再刪除；如果數據發送失敗，可以重新發送。），Store on failure（這也是scribe采用的策略，當數據接收方crash時，將數據寫到本地，待恢復后，繼續發送），Besteffort（數據發送到接收方后，不會進行確認）。 flume的可恢復性： 還是靠Channel。推薦使用FileChannel，事件持久化在本地文件系統里(性能較差)。 flume的一些核心概念：Agent 使用JVM 運行Flume。每臺機器運行一個agent，但是可以在一個agent中包含多個sources和sinks。Client 生產數據，運行在一個獨立的線程。Source 從Client收集數據，傳遞給Channel。Sink 從Channel收集數據，運行在一個獨立線程。Channel 連接 sources 和 sinks ，這個有點像一個隊列。Events 可以是日志記錄、 avro 對象等。 Flume以agent為最小的獨立運行單位。一個agent就是一個JVM。單agent由Source、Sink和Channel三大組件構成，如下圖：
值得注意的是，Flume提供了大量內置的Source、Channel和Sink類型。不同類型的Source,Channel和Sink可以自由組合。組合方式基于用戶設置的配置文件，非常靈活。比如：Channel可以把事件暫存在內存里，也可以持久化到本地硬盤上。Sink可以把日志寫入HDFS, HBase，甚至是另外一個Source等等。Flume支持用戶建立多級流，也就是說，多個agent可以協同工作，并且支持Fan-in、Fan-out、Contextual Routing、Backup Routes，這也正是NB之處。如下圖所示:
二、flume的官方網站在哪里？ http://flume.apache.org/ 三、在哪里下載？ http://www.apache.org/dyn/closer.cgi/flume/1.5.0/apache-flume-1.5.0-bin.tar.gz 四、如何安裝？ 1)將下載的flume包，解壓到/home/hadoop目錄中，你就已經完成了50%：）簡單吧 2)修改 flume-env.sh 配置文件,主要是java_HOME變量設置root@m1:/home/hadoop/flume-1.5.0-bin# cp conf/flume-env.sh.template conf/flume-env.shroot@m1:/home/hadoop/flume-1.5.0-bin# vi conf/flume-env.sh# 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.# If this file is placed at FLUME_CONF_DIR/flume-env.sh, it will be sourced# during Flume startup.# Enviroment variables can be set here.JAVA_HOME=/usr/lib/jvm/java-7-Oracle# Give Flume more memory and pre-allocate, enable remote monitoring via JMX#JAVA_OPTS="-Xms100m -Xmx200m -Dcom.sun.management.jmxremote"# Note that the Flume conf directory is always included in the classpath.#FLUME_CLASSPATH=""復制代碼 3)驗證是否安裝成功root@m1:/home/hadoop# /home/hadoop/flume-1.5.0-bin/bin/flume-ng versionFlume 1.5.0Source code repository: https://git-wip-us.apache.org/repos/asf/flume.gitRevision: 8633220df808c4cd0c13d1cf0320454a94f1ea97Compiled by hshreedharan on Wed May 7 14:49:18 PDT 2014From source with checksum a01fe726e4380ba0c9f7a7d222db961froot@m1:/home/hadoop#復制代碼 出現上面的信息，表示安裝成功了 五、flume的案例 1)案例1：Avro Avro可以發送一個給定的文件給Flume，Avro 源使用AVRO RPC機制。 a)創建agent配置文件root@m1:/home/hadoop#vi /home/hadoop/flume-1.5.0-bin/conf/avro.confa1.sources = r1a1.sinks = k1a1.channels = c1# Describe/configure the sourcea1.sources.r1.type = avroa1.sources.r1.channels = c1a1.sources.r1.bind = 0.0.0.0a1.sources.r1.port = 4141# Describe the sinka1.sinks.k1.type = logger# Use a channel which buffers events in memorya1.channels.c1.type = memorya1.channels.c1.capacity = 1000a1.channels.c1.transactionCapacity = 100# Bind the source and sink to the channela1.sources.r1.channels = c1a1.sinks.k1.channel = c1復制代碼 b)啟動flume agent a1root@m1:/home/hadoop# /home/hadoop/flume-1.5.0-bin/bin/flume-ng agent -c . -f /home/hadoop/flume-1.5.0-bin/conf/avro.conf -n a1 -Dflume.root.logger=INFO,console復制代碼 c)創建指定文件root@m1:/home/hadoop# echo "hello world" > /home/hadoop/flume-1.5.0-bin/log.00復制代碼 d)使用avro-client發送文件root@m1:/home/hadoop# /home/hadoop/flume-1.5.0-bin/bin/flume-ng avro-client -c . -H m1 -p 4141 -F /home/hadoop/flume-1.5.0-bin/log.00復制代碼 d)使用avro-client發送文件root@m1:/home/hadoop# /home/hadoop/flume-1.5.0-bin/bin/flume-ng avro-client -c . -H m1 -p 4141 -F /復制代碼 f)在m1的控制臺，可以看到以下信息，注意最后一行：root@m1:/home/hadoop/flume-1.5.0-bin/conf# /home/hadoop/flume-1.5.0-bin/bin/flume-ng agent -c . -f /home/hadoop/flume-1.5.0-bin/conf/avro.conf -n a1 -Dflume.root.logger=INFO,consoleInfo: Sourcing environment configuration script /home/hadoop/flume-1.5.0-bin/conf/flume-env.shInfo: Including Hadoop libraries found via (/home/hadoop/hadoop-2.2.0/bin/hadoop) for HDFS accessInfo: Excluding /home/hadoop/hadoop-2.2.0/share/hadoop/common/lib/slf4j-api-1.7.5.jar from classpathInfo: Excluding /home/hadoop/hadoop-2.2.0/share/hadoop/common/lib/slf4j-log4j12-1.7.5.jar from classpath...2014-08-10 10:43:25,112 (New I/O worker #1) [INFO - org.apache.avro.ipc.NettyServer$NettyServerAvroHandler.handleUpstream(NettyServer.java:171)] [id: 0x92464c4f, /192.168.1.50:59850 :> /192.168.1.50:4141] UNBOUND2014-08-10 10:43:25,112 (New I/O worker #1) [INFO - org.apache.avro.ipc.NettyServer$NettyServerAvroHandler.handleUpstream(NettyServer.java:171)] [id: 0x92464c4f, /192.168.1.50:59850 :> /192.168.1.50:4141] CLOSED2014-08-10 10:43:25,112 (New I/O worker #1) [INFO - org.apache.avro.ipc.NettyServer$NettyServerAvroHandler.channelClosed(NettyServer.java:209)] Connection to /192.168.1.50:59850 disconnected.2014-08-10 10:43:26,718 (SinkRunner-PollingRunner-DefaultSinkProcessor) [INFO - org.apache.flume.sink.LoggerSink.process(LoggerSink.java:70)] Event: { headers:{} body: 68 65 6C 6C 6F 20 77 6F 72 6C 64 hello world }復制代碼 2)案例2：Spool Spool監測配置的目錄下新增的文件，并將文件中的數據讀取出來。需要注意兩點： 1) 拷貝到spool目錄下的文件不可以再打開編輯。 2) spool目錄下不可包含相應的子目錄 a)創建agent配置文件root@m1:/home/hadoop# vi /home/hadoop/flume-1.5.0-bin/conf/spool.confa1.sources = r1a1.sinks = k1a1.channels = c1# Describe/configure the sourcea1.sources.r1.type = spooldira1.sources.r1.channels = c1a1.sources.r1.spoolDir = /home/hadoop/flume-1.5.0-bin/logsa1.sources.r1.fileHeader = true# Describe the sinka1.sinks.k1.type = logger# Use a channel which buffers events in memorya1.channels.c1.type = memorya1.channels.c1.capacity = 1000a1.channels.c1.transactionCapacity = 100# Bind the source and sink to the channela1.sources.r1.channels = c1a1.sinks.k1.channel = c1復制代碼 b)啟動flume agent a1root@m1:/home/hadoop# /home/hadoop/flume-1.5.0-bin/bin/flume-ng agent -c . -f /home/hadoop/flume-1.5.0-bin/conf/spool.conf -n a1 -Dflume.root.logger=INFO,console復制代碼 c)追加文件到/home/hadoop/flume-1.5.0-bin/logs目錄root@m1:/home/hadoop# echo "spool test1" > /home/hadoop/flume-1.5.0-bin/logs/spool_text.log復制代碼 d)在m1的控制臺，可以看到以下相關信息：14/08/10 11:37:13 INFO source.SpoolDirectorySource: Spooling Directory Source runner has shutdown.14/08/10 11:37:13 INFO source.SpoolDirectorySource: Spooling Directory Source runner has shutdown.14/08/10 11:37:14 INFO avro.ReliableSpoolingFileEventReader: Preparing to move file /home/hadoop/flume-1.5.0-bin/logs/spool_text.log to /home/hadoop/flume-1.5.0-bin/logs/spool_text.log.COMPLETED14/08/10 11:37:14 INFO source.SpoolDirectorySource: Spooling Directory Source runner has shutdown.14/08/10 11:37:14 INFO source.SpoolDirectorySource: Spooling Directory Source runner has shutdown.14/08/10 11:37:14 INFO sink.LoggerSink: Event: { headers:{file=/home/hadoop/flume-1.5.0-bin/logs/spool_text.log} body: 73 70 6F 6F 6C 20 74 65 73 74 31 spool test1 }14/08/10 11:37:15 INFO source.SpoolDirectorySource: Spooling Directory Source runner has shutdown.14/08/10 11:37:15 INFO source.SpoolDirectorySource: Spooling Directory Source runner has shutdown.14/08/10 11:37:16 INFO source.SpoolDirectorySource: Spooling Directory Source runner has shutdown.14/08/10 11:37:16 INFO source.SpoolDirectorySource: Spooling Directory Source runner has shutdown.14/08/10 11:37:17 INFO source.SpoolDirectorySource: Spooling Directory Source runner has shutdown.復制代碼 3)案例3：Exec EXEC執行一個給定的命令獲得輸出的源,如果要使用tail命令，必選使得file足夠大才能看到輸出內容 a)創建agent配置文件root@m1:/home/hadoop# vi /home/hadoop/flume-1.5.0-bin/conf/exec_tail.confa1.sources = r1a1.sinks = k1a1.channels = c1# Describe/configure the sourcea1.sources.r1.type = execa1.sources.r1.channels = c1a1.sources.r1.command = tail -F /home/hadoop/flume-1.5.0-bin/log_exec_tail# Describe the sinka1.sinks.k1.type = logger# Use a channel which buffers events in memorya1.channels.c1.type = memorya1.channels.c1.capacity = 1000a1.channels.c1.transactionCapacity = 100# Bind the source and sink to the channela1.sources.r1.channels = c1a1.sinks.k1.channel = c1復制代碼 b)啟動flume agent a1root@m1:/home/hadoop# /home/hadoop/flume-1.5.0-bin/bin/flume-ng agent -c . -f /home/hadoop/flume-1.5.0-bin/conf/exec_tail.conf -n a1 -Dflume.root.logger=INFO,console復制代碼 c)生成足夠多的內容在文件里root@m1:/home/hadoop# for i in {1..100};do echo "exec tail$i" >> /home/hadoop/flume-1.5.0-bin/log_復制代碼 e)在m1的控制臺，可以看到以下信息：2014-08-10 10:59:25,513 (SinkRunner-PollingRunner-DefaultSinkProcessor) [INFO - org.apache.flume.sink.LoggerSink.process(LoggerSink.java:70)] Event: { headers:{} body: 65 78 65 63 20 74 61 69 6C 20 74 65 73 74 exec tail test }2014-08-10 10:59:34,535 (SinkRunner-PollingRunner-DefaultSinkProcessor) [INFO - org.apache.flume.sink.LoggerSink.process(LoggerSink.java:70)] Event: { headers:{} body: 65 78 65 63 20 74 61 69 6C 20 74 65 73 74 exec tail test }2014-08-10 11:01:40,557 (SinkRunner-PollingRunner-DefaultSinkProcessor) [INFO - org.apache.flume.sink.LoggerSink.process(LoggerSink.java:70)] Event: { headers:{} body: 65 78 65 63 20 74 61 69 6C 31 exec tail1 }2014-08-10 11:01:41,180 (SinkRunner-PollingRunner-DefaultSinkProcessor) [INFO - org.apache.flume.sink.LoggerSink.process(LoggerSink.java:70)] Event: { headers:{} body: 65 78 65 63 20 74 61 69 6C 32 exec tail2 }2014-08-10 11:01:41,180 (SinkRunner-PollingRunner-DefaultSinkProcessor) [INFO - org.apache.flume.sink.LoggerSink.process(LoggerSink.java:70)] Event: { headers:{} body: 65 78 65 63 20 74 61 69 6C 33 exec tail3 }2014-08-10 11:01:41,181 (SinkRunner-PollingRunner-DefaultSinkProcessor) [INFO - org.apache.flume.sink.LoggerSink.process(LoggerSink.java:70)] Event: { headers:{} body: 65 78 65 63 20 74 61 69 6C 34 exec tail4 }2014-08-10 11:01:41,181 (SinkRunner-PollingRunner-DefaultSinkProcessor) [INFO - org.apache.flume.sink.LoggerSink.process(LoggerSink.java:70)] Event: { headers:{} body: 65 78 65 63 20 74 61 69 6C 35 exec tail5 }2014-08-10 11:01:41,181 (SinkRunner-PollingRunner-DefaultSinkProcessor) [INFO - org.apache.flume.sink.LoggerSink.process(LoggerSink.java:70)] Event: { headers:{} body: 65 78 65 63 20 74 61 69 6C 36 exec tail6 }............2014-08-10 11:01:51,550 (SinkRunner-PollingRunner-DefaultSinkProcessor) [INFO - org.apache.flume.sink.LoggerSink.process(LoggerSink.java:70)] Event: { headers:{} body: 65 78 65 63 20 74 61 69 6C 39 36 exec tail96 }2014-08-10 11:01:51,550 (SinkRunner-PollingRunner-DefaultSinkProcessor) [INFO - org.apache.flume.sink.LoggerSink.process(LoggerSink.java:70)] Event: { headers:{} body: 65 78 65 63 20 74 61 69 6C 39 37 exec tail97 }2014-08-10 11:01:51,551 (SinkRunner-PollingRunner-DefaultSinkProcessor) [INFO - org.apache.flume.sink.LoggerSink.process(LoggerSink.java:70)] Event: { headers:{} body: 65 78 65 63 20 74 61 69 6C 39 38 exec tail98 }2014-08-10 11:01:51,551 (SinkRunner-PollingRunner-DefaultSinkProcessor) [INFO - org.apache.flume.sink.LoggerSink.process(LoggerSink.java:70)] Event: { headers:{} body: 65 78 65 63 20 74 61 69 6C 39 39 exec tail99 }2014-08-10 11:01:51,551 (SinkRunner-PollingRunner-DefaultSinkProcessor) [INFO - org.apache.flume.sink.LoggerSink.process(LoggerSink.java:70)] Event: { headers:{} body: 65 78 65 63 20 74 61 69 6C 31 30 30 exec tail100 }復制代碼 4)案例4：Syslogtcp Syslogtcp監聽TCP的端口做為數據源 a)創建agent配置文件root@m1:/home/hadoop# vi /home/hadoop/flume-1.5.0-bin/conf/syslog_tcp.confa1.sources = r1a1.sinks = k1a1.channels = c1# Describe/configure the sourcea1.sources.r1.type = syslogtcpa1.sources.r1.port = 5140a1.sources.r1.host = localhosta1.sources.r1.channels = c1# Describe the sinka1.sinks.k1.type = logger# Use a channel which buffers events in memorya1.channels.c1.type = memorya1.channels.c1.capacity = 1000a1.channels.c1.transactionCapacity = 100# Bind the source and sink to the channela1.sources.r1.channels = c1a1.sinks.k1.channel = c1復制代碼 b)啟動flume agent a1root@m1:/home/hadoop# /home/hadoop/flume-1.5.0-bin/bin/flume-ng agent -c . -f /home/hadoop/flume-1.5.0-bin/conf/syslog_tcp.conf -n a1 -Dflume.root.logger=INFO,console復制代碼 c)測試產生syslogroot@m1:/home/hadoop# echo "hello idoall.org syslog" | nc localhost 5140復制代碼 d)在m1的控制臺，可以看到以下信息：14/08/10 11:41:45 INFO node.PollingPropertiesFileConfigurationProvider: Reloading configuration file:/home/hadoop/flume-1.5.0-bin/conf/syslog_tcp.conf14/08/10 11:41:45 INFO conf.FlumeConfiguration: Added sinks: k1 Agent: a114/08/10 11:41:45 INFO conf.FlumeConfiguration: Processing:k114/08/10 11:41:45 INFO conf.FlumeConfiguration: Processing:k114/08/10 11:41:45 INFO conf.FlumeConfiguration: Post-validation flume configuration contains configuration for agents: [a1]14/08/10 11:41:45 INFO node.AbstractConfigurationProvider: Creating channels14/08/10 11:41:45 INFO channel.DefaultChannelFactory: Creating instance of channel c1 type memory14/08/10 11:41:45 INFO node.AbstractConfigurationProvider: Created channel c114/08/10 11:41:45 INFO source.DefaultSourceFactory: Creating instance of source r1, type syslogtcp14/08/10 11:41:45 INFO sink.DefaultSinkFactory: Creating instance of sink: k1, type: logger14/08/10 11:41:45 INFO node.AbstractConfigurationProvider: Channel c1 connected to [r1, k1]14/08/10 11:41:45 INFO node.application: Starting new configuration:{ sourceRunners:{r1=EventDrivenSourceRunner: { source:org.apache.flume.source.SyslogTcpSource{name:r1,state:IDLE} }} sinkRunners:{k1=SinkRunner: { policy:org.apache.flume.sink.DefaultSinkProcessor@6538b14 counterGroup:{ name:null counters:{} } }} channels:{c1=org.apache.flume.channel.MemoryChannel{name: c1}} }14/08/10 11:41:45 INFO node.Application: Starting Channel c114/08/10 11:41:45 INFO instrumentation.MonitoredCounterGroup: Monitored counter group for type: CHANNEL, name: c1: Successfully registered new MBean.14/08/10 11:41:45 INFO instrumentation.MonitoredCounterGroup: Component type: CHANNEL, name: c1 started14/08/10 11:41:45 INFO node.Application: Starting Sink k114/08/10 11:41:45 INFO node.Application: Starting Source r114/08/10 11:41:45 INFO source.SyslogTcpSource: Syslog TCP Source starting...14/08/10 11:42:15 WARN source.SyslogUtils: Event created from Invalid Syslog data.14/08/10 11:42:15 INFO sink.LoggerSink: Event: { headers:{Severity=0, flume.syslog.status=Invalid, Facility=0} body: 68 65 6C 6C 6F 20 69 64 6F 61 6C 6C 2E 6F 72 67 hello idoall.org }復制代碼 5)案例5：JSONHandler a)創建agent配置文件root@m1:/home/hadoop# vi /home/hadoop/flume-1.5.0-bin/conf/post_json.confa1.sources = r1a1.sinks = k1a1.channels = c1# Describe/configure the sourcea1.sources.r1.type = org.apache.flume.source.http.HTTPSourcea1.sources.r1.port = 8888a1.sources.r1.channels = c1# Describe the sinka1.sinks.k1.type = logger# Use a channel which buffers events in memorya1.channels.c1.type = memorya1.channels.c1.capacity = 1000a1.channels.c1.transactionCapacity = 100# Bind the source and sink to the channela1.sources.r1.channels = c1a1.sinks.k1.channel = c1復制代碼 b)啟動flume agent a1root@m1:/home/hadoop# /home/hadoop/flume-1.5.0-bin/bin/flume-ng agent -c . -f /home/hadoop/flume-1.5.0-bin/conf/post_json.conf -n a1 -Dflume.root.logger=INFO,console復制代碼 c)生成JSON 格式的POST requestroot@m1:/home/hadoop# curl -X POST -d '[{ "headers" :{"a" : "a1","b" : "b1"},"body" : "idoall.org_body"}]' http://localhost:8888復制代碼 d)在m1的控制臺，可以看到以下信息：14/08/10 11:49:59 INFO node.Application: Starting Channel c114/08/10 11:49:59 INFO instrumentation.MonitoredCounterGroup: Monitored counter group for type: CHANNEL, name: c1: Successfully registered new MBean.14/08/10 11:49:59 INFO instrumentation.MonitoredCounterGroup: Component type: CHANNEL, name: c1 started14/08/10 11:49:59 INFO node.Application: Starting Sink k114/08/10 11:49:59 INFO node.Application: Starting Source r114/08/10 11:49:59 INFO mortbay.log: Logging to org.slf4j.impl.Log4jLoggerAdapter(org.mortbay.log) via org.mortbay.log.Slf4jLog14/08/10 11:49:59 INFO mortbay.log: jetty-6.1.2614/08/10 11:50:00 INFO mortbay.log: Started [email protected]:888814/08/10 11:50:00 INFO instrumentation.MonitoredCounterGroup: Monitored counter group for type: SOURCE, name: r1: Successfully registered new MBean.14/08/10 11:50:00 INFO instrumentation.MonitoredCounterGroup: Component type: SOURCE, name: r1 started14/08/10 12:14:32 INFO sink.LoggerSink: Event: { headers:{b=b1, a=a1} body: 69 64 6F 61 6C 6C 2E 6F 72 67 5F 62 6F 64 79 idoall.org_body }復制代碼 6)案例6：Hadoop sink a)創建agent配置文件root@m1:/home/hadoop# vi /home/hadoop/flume-1.5.0-bin/conf/hdfs_sink.confa1.sources = r1a1.sinks = k1a1.channels = c1# Describe/configure the sourcea1.sources.r1.type = syslogtcpa1.sources.r1.port = 5140a1.sources.r1.host = localhosta1.sources.r1.channels = c1# Describe the sinka1.sinks.k1.type = hdfsa1.sinks.k1.channel = c1a1.sinks.k1.hdfs.path = hdfs://m1:9000/user/flume/syslogtcpa1.sinks.k1.hdfs.filePrefix = Sysloga1.sinks.k1.hdfs.round = truea1.sinks.k1.hdfs.roundValue = 10a1.sinks.k1.hdfs.roundUnit = minute# Use a channel which buffers events in memorya1.channels.c1.type = memorya1.channels.c1.capacity = 1000a1.channels.c1.transactionCapacity = 100# Bind the source and sink to the channela1.sources.r1.channels = c1a1.sinks.k1.channel = c1復制代碼 b)啟動flume agent a1root@m1:/home/hadoop# /home/hadoop/flume-1.5.0-bin/bin/flume-ng agent -c . -f /home/hadoop/flume-1.5.0-bin/conf/hdfs_sink.conf -n a1 -Dflume.root.logger=INFO,console復制代碼 c)測試產生syslogroot@m1:/home/hadoop# echo "hello idoall flume -> hadoop testing one" | nc localhost 5140復制代碼 d)在m1的控制臺，可以看到以下信息：14/08/10 12:20:39 INFO instrumentation.MonitoredCounterGroup: Monitored counter group for type: CHANNEL, name: c1: Successfully registered new MBean.14/08/10 12:20:39 INFO instrumentation.MonitoredCounterGroup: Component type: CHANNEL, name: c1 started14/08/10 12:20:39 INFO node.Application: Starting Sink k114/08/10 12:20:39 INFO node.Application: Starting Source r114/08/10 12:20:39 INFO instrumentation.MonitoredCounterGroup: Monitored counter group for type: SINK, name: k1: Successfully registered new MBean.14/08/10 12:20:39 INFO instrumentation.MonitoredCounterGroup: Component type: SINK, name: k1 started14/08/10 12:20:39 INFO source.SyslogTcpSource: Syslog TCP Source starting...14/08/10 12:21:46 WARN source.SyslogUtils: Event created from Invalid Syslog data.14/08/10 12:21:49 INFO hdfs.HDFSSequenceFile: writeFormat = Writable, UseRawLocalFileSystem = false14/08/10 12:21:49 INFO hdfs.BucketWriter: Creating hdfs://m1:9000/user/flume/syslogtcp//Syslog.1407644509504.tmp14/08/10 12:22:20 INFO hdfs.BucketWriter: Closing hdfs://m1:9000/user/flume/syslogtcp//Syslog.1407644509504.tmp14/08/10 12:22:20 INFO hdfs.BucketWriter: Close tries incremented14/08/10 12:22:20 INFO hdfs.BucketWriter: Renaming hdfs://m1:9000/user/flume/syslogtcp/Syslog.1407644509504.tmp to hdfs://m1:9000/user/flume/syslogtcp/Syslog.140764450950414/08/10 12:22:20 INFO hdfs.HDFSEventSink: Writer callback called.復制代碼 e)在m1上再打開一個窗口，去hadoop上檢查文件是否生成root@m1:/home/hadoop# /home/hadoop/hadoop-2.2.0/bin/hadoop fs -ls /user/flume/syslogtcpFound 1 items-rw-r--r-- 3 root supergroup 155 2014-08-10 12:22 /user/flume/syslogtcp/Syslog.1407644509504root@m1:/home/hadoop# /home/hadoop/hadoop-2.2.0/bin/hadoop fs -cat /user/flume/syslogtcp/Syslog.1407644509504SEQ!org.apache.hadoop.io.LongWritable"org.apache.hadoop.io.BytesWritable^;>Gv$hello idoall flume -> hadoop testing one復制代碼 7)案例7：File Roll Sink a)創建agent配置文件root@m1:/home/hadoop# vi /home/hadoop/flume-1.5.0-bin/conf/file_roll.confa1.sources = r1a1.sinks = k1a1.channels = c1# Describe/configure the sourcea1.sources.r1.type = syslogtcpa1.sources.r1.port = 5555a1.sources.r1.host = localhosta1.sources.r1.channels = c1# Describe the sinka1.sinks.k1.type = file_rolla1.sinks.k1.sink.directory = /home/hadoop/flume-1.5.0-bin/logs# Use a channel which buffers events in memorya1.channels.c1.type = memorya1.channels.c1.capacity = 1000a1.channels.c1.transactionCapacity = 100# Bind the source and sink to the channela1.sources.r1.channels = c1a1.sinks.k1.channel = c1復制代碼 b)啟動flume agent a1root@m1:/home/hadoop# /home/hadoop/flume-1.5.0-bin/bin/flume-ng agent -c . -f /home/hadoop/flume-1.5.0-bin/conf/file_roll.conf -n a1 -Dflume.root.logger=INFO,console復制代碼 c)測試產生logroot@m1:/home/hadoop# echo "hello idoall.org syslog" | nc localhost 5555root@m1:/home/hadoop# echo "hello idoall.org syslog 2" | nc localhost 5555復制代碼 d)查看/home/hadoop/flume-1.5.0-bin/logs下是否生成文件,默認每30秒生成一個新文件root@m1:/home/hadoop# ll /home/hadoop/flume-1.5.0-bin/logs總用量 272drwxr-xr-x 3 root root 4096 Aug 10 12:50 ./drwxr-xr-x 9 root root 4096 Aug 10 10:59 ../-rw-r--r-- 1 root root 50 Aug 10 12:49 1407646164782-1-rw-r--r-- 1 root root 0 Aug 10 12:49 1407646164782-2-rw-r--r-- 1 root root 0 Aug 10 12:50 1407646164782-3root@m1:/home/hadoop# cat /home/hadoop/flume-1.5.0-bin/logs/1407646164782-1 /home/hadoop/flume-1.5.0-bin/logs/1407646164782-2hello idoall.org sysloghello idoall.org syslog 2復制代碼 8)案例8：Replicating Channel Selector Flume支持Fan out流從一個源到多個通道。有兩種模式的Fan out，分別是復制和復用。在復制的情況下，流的事件被發送到所有的配置通道。在復用的情況下，事件被發送到可用的渠道中的一個子集。Fan out流需要指定源和Fan out通道的規則。 這次我們需要用到m1,m2兩臺機器 a)在m1創建replicating_Channel_Selector配置文件root@m1:/home/hadoop# vi /home/hadoop/flume-1.5.0-bin/conf/replicating_Channel_Selector.confa1.sources = r1a1.sinks = k1 k2a1.channels = c1 c2# Describe/configure the sourcea1.sources.r1.type = syslogtcpa1.sources.r1.port = 5140a1.sources.r1.host = localhosta1.sources.r1.channels = c1 c2a1.sources.r1.selector.type = replicating# Describe the sinka1.sinks.k1.type = avroa1.sinks.k1.channel = c1a1.sinks.k1.hostname = m1a1.sinks.k1.port = 5555a1.sinks.k2.type = avroa1.sinks.k2.channel = c2a1.sinks.k2.hostname = m2a1.sinks.k2.port = 5555# Use a channel which buffers events in memorya1.channels.c1.type = memorya1.channels.c1.capacity = 1000a1.channels.c1.transactionCapacity = 100a1.channels.c2.type = memorya1.channels.c2.capacity = 1000a1.channels.c2.transactionCapacity = 100復制代碼 b)在m1創建replicating_Channel_Selector_avro配置文件root@m1:/home/hadoop# vi /home/hadoop/flume-1.5.0-bin/conf/replicating_Channel_Selector_avro.confa1.sources = r1a1.sinks = k1a1.channels = c1# Describe/configure the sourcea1.sources.r1.type = avroa1.sources.r1.channels = c1a1.sources.r1.bind = 0.0.0.0a1.sources.r1.port = 5555# Describe the sinka1.sinks.k1.type = logger# Use a channel which buffers events in memorya1.channels.c1.type = memorya1.channels.c1.capacity = 1000a1.channels.c1.transactionCapacity = 100# Bind the source and sink to the channela1.sources.r1.channels = c1a1.sinks.k1.channel = c1復制代碼 c)在m1上將2個配置文件復制到m2上一份root@m1:/home/hadoop/flume-1.5.0-bin# scp -r /home/hadoop/flume-1.5.0-bin/conf/replicating_Channel_Selector.conf root@m2:/home/hadoop/flume-1.5.0-bin/conf/replicating_Channel_Selector.confroot@m1:/home/hadoop/flume-1.5.0-bin# scp -r /home/hadoop/flume-1.5.0-bin/conf/replicating_Channel_Selector_avro.conf root@m2:/home/hadoop/flume-1.5.0-bin/conf/replicating_Channel_Selector_avro.conf復制代碼 d)打開4個窗口，在m1和m2上同時啟動兩個flume agentroot@m1:/home/hadoop# /home/hadoop/flume-1.5.0-bin/bin/flume-ng agent -c . -f /home/hadoop/flume-1.5.0-bin/conf/replicating_Channel_Selector_avro.conf -n a1 -Dflume.root.logger=INFO,consoleroot@m1:/home/hadoop# /home/hadoop/flume-1.5.0-bin/bin/flume-ng agent -c . -f /home/hadoop/flume-1.5.0-bin/conf/replicating_Channel_Selector.conf -n a1 -Dflume.root.logger=INFO,console復制代碼 e)然后在m1或m2的任意一臺機器上，測試產生syslogroot@m1:/home/hadoop# echo "hello idoall.org syslog" | nc localhost 5140復制代碼 f)在m1和m2的sink窗口，分別可以看到以下信息,這說明信息得到了同步：14/08/10 14:08:18 INFO ipc.NettyServer: Connection to /192.168.1.51:46844 disconnected.14/08/10 14:08:52 INFO ipc.NettyServer: [id: 0x90f8fe1f, /192.168.1.50:35873 => /192.168.1.50:5555] OPEN14/08/10 14:08:52 INFO ipc.NettyServer: [id: 0x90f8fe1f, /192.168.1.50:35873 => /192.168.1.50:5555] BOUND: /192.168.1.50:555514/08/10 14:08:52 INFO ipc.NettyServer: [id: 0x90f8fe1f, /192.168.1.50:35873 => /192.168.1.50:5555] CONNECTED: /192.168.1.50:3587314/08/10 14:08:59 INFO ipc.NettyServer: [id: 0xd6318635, /192.168.1.51:46858 => /192.168.1.50:5555] OPEN14/08/10 14:08:59 INFO ipc.NettyServer: [id: 0xd6318635, /192.168.1.51:46858 => /192.168.1.50:5555] BOUND: /192.168.1.50:555514/08/10 14:08:59 INFO ipc.NettyServer: [id: 0xd6318635, /192.168.1.51:46858 => /192.168.1.50:5555] CONNECTED: /192.168.1.51:4685814/08/10 14:09:20 INFO sink.LoggerSink: Event: { headers:{Severity=0, flume.syslog.status=Invalid, Facility=0} body: 68 65 6C 6C 6F 20 69 64 6F 61 6C 6C 2E 6F 72 67 hello idoall.org }復制代碼 9)案例9：Multiplexing Channel Selector a)在m1創建Multiplexing_Channel_Selector配置文件root@m1:/home/hadoop# vi /home/hadoop/flume-1.5.0-bin/conf/Multiplexing_Channel_Selector.confa1.sources = r1a1.sinks = k1 k2a1.channels = c1 c2# Describe/configure the sourcea1.sources.r1.type = org.apache.flume.source.http.HTTPSourcea1.sources.r1.port = 5140a1.sources.r1.channels = c1 c2a1.sources.r1.selector.type = multiplexinga1.sources.r1.selector.header = type#映射允許每個值通道可以重疊。默認值可以包含任意數量的通道。a1.sources.r1.selector.mapping.baidu = c1a1.sources.r1.selector.mapping.ali = c2a1.sources.r1.selector.default = c1# Describe the sinka1.sinks.k1.type = avroa1.sinks.k1.channel = c1a1.sinks.k1.hostname = m1a1.sinks.k1.port = 5555a1.sinks.k2.type = avroa1.sinks.k2.channel = c2a1.sinks.k2.hostname = m2a1.sinks.k2.port = 5555# Use a channel which buffers events in memorya1.channels.c1.type = memorya1.channels.c1.capacity = 1000a1.channels.c1.transactionCapacity = 100a1.channels.c2.type = memorya1.channels.c2.capacity = 1000a1.channels.c2.transactionCapacity = 100復制代碼 b)在m1創建Multiplexing_Channel_Selector_avro配置文件root@m1:/home/hadoop# vi /home/hadoop/flume-1.5.0-bin/conf/Multiplexing_Channel_Selector_avro.confa1.sources = r1a1.sinks = k1a1.channels = c1# Describe/configure the sourcea1.sources.r1.type = avroa1.sources.r1.channels = c1a1.sources.r1.bind = 0.0.0.0a1.sources.r1.port = 5555# Describe the sinka1.sinks.k1.type = logger# Use a channel which buffers events in memorya1.channels.c1.type = memorya1.channels.c1.capacity = 1000a1.channels.c1.transactionCapacity = 100# Bind the source and sink to the channela1.sources.r1.channels = c1a1.sinks.k1.channel = c1復制代碼 c)將2個配置文件復制到m2上一份root@m1:/home/hadoop/flume-1.5.0-bin# scp -r /home/hadoop/flume-1.5.0-bin/conf/Multiplexing_Channel_Selector.conf root@m2:/home/hadoop/flume-1.5.0-bin/conf/Multiplexing_Channel_Selector.confroot@m1:/home/hadoop/flume-1.5.0-bin# scp -r /home/hadoop/flume-1.5.0-bin/conf/Multiplexing_Channel_Selector_avro.conf root@m2:/home/hadoop/flume-1.5.0-bin/conf/Multiplexing_Channel_Selector_avro.conf復制代碼 d)打開4個窗口，在m1和m2上同時啟動兩個flume agentroot@m1:/home/hadoop# /home/hadoop/flume-1.5.0-bin/bin/flume-ng agent -c . -f /home/hadoop/flume-1.5.0-bin/conf/Multiplexing_Channel_Selector_avro.conf -n a1 -Dflume.root.logger=INFO,consoleroot@m1:/home/hadoop# /home/hadoop/flume-1.5.0-bin/bin/flume-ng agent -c . -f /home/hadoop/flume-1.5.0-bin/conf/Multiplexing_Channel_Selector.conf -n a1 -Dflume.root.logger=INFO,console復制代碼 e)然后在m1或m2的任意一臺機器上，測試產生syslogroot@m1:/home/hadoop# curl -X POST -d '[{ "headers" :{"type" : "baidu"},"body" : "idoall_TEST1"}]' http://localhost:5140 && curl -X POST -d '[{ "headers" :{"type" : "ali"},"body" : "idoall_TEST2"}]' http://localhost:5140 && curl -X POST -d '[{ "headers" :{"type" : "QQ"},"body" : "idoall_TEST3"}]' http://localhost:5140復制代碼 f)在m1的sink窗口，可以看到以下信息：14/08/10 14:32:21 INFO node.Application: Starting Sink k114/08/10 14:32:21 INFO node.Application: Starting Source r114/08/10 14:32:21 INFO source.AvroSource: Starting Avro source r1: { bindAddress: 0.0.0.0, port: 5555 }...14/08/10 14:32:21 INFO instrumentation.MonitoredCounterGroup: Monitored counter group for type: SOURCE, name: r1: Successfully registered new MBean.14/08/10 14:32:21 INFO instrumentation.MonitoredCounterGroup: Component type: SOURCE, name: r1 started14/08/10 14:32:21 INFO source.AvroSource: Avro source r1 started.14/08/10 14:32:36 INFO ipc.NettyServer: [id: 0xcf00eea6, /192.168.1.50:35916 => /192.168.1.50:5555] OPEN14/08/10 14:32:36 INFO ipc.NettyServer: [id: 0xcf00eea6, /192.168.1.50:35916 => /192.168.1.50:5555] BOUND: /192.168.1.50:555514/08/10 14:32:36 INFO ipc.NettyServer: [id: 0xcf00eea6, /192.168.1.50:35916 => /192.168.1.50:5555] CONNECTED: /192.168.1.50:3591614/08/10 14:32:44 INFO ipc.NettyServer: [id: 0x432f5468, /192.168.1.51:46945 => /192.168.1.50:5555] OPEN14/08/10 14:32:44 INFO ipc.NettyServer: [id: 0x432f5468, /192.168.1.51:46945 => /192.168.1.50:5555] BOUND: /192.168.1.50:555514/08/10 14:32:44 INFO ipc.NettyServer: [id: 0x432f5468, /192.168.1.51:46945 => /192.168.1.50:5555] CONNECTED: /192.168.1.51:4694514/08/10 14:34:11 INFO sink.LoggerSink: Event: { headers:{type=baidu} body: 69 64 6F 61 6C 6C 5F 54 45 53 54 31 idoall_TEST1 }14/08/10 14:34:57 INFO sink.LoggerSink: Event: { headers:{type=qq} body: 69 64 6F 61 6C 6C 5F 54 45 53 54 33 idoall_TEST3 }復制代碼 g)在m2的sink窗口，可以看到以下信息：14/08/10 14:32:27 INFO node.Application: Starting Sink k114/08/10 14:32:27 INFO node.Application: Starting Source r114/08/10 14:32:27 INFO source.AvroSource: Starting Avro source r1: { bindAddress: 0.0.0.0, port: 5555 }...14/08/10 14:32:27 INFO instrumentation.MonitoredCounterGroup: Monitored counter group for type: SOURCE, name: r1: Successfully registered new MBean.14/08/10 14:32:27 INFO instrumentation.MonitoredCounterGroup: Component type: SOURCE, name: r1 started14/08/10 14:32:27 INFO source.AvroSource: Avro source r1 started.14/08/10 14:32:36 INFO ipc.NettyServer: [id: 0x7c2f0aec, /192.168.1.50:38104 => /192.168.1.51:5555] OPEN14/08/10 14:32:36 INFO ipc.NettyServer: [id: 0x7c2f0aec, /192.168.1.50:38104 => /192.168.1.51:5555] BOUND: /192.168.1.51:555514/08/10 14:32:36 INFO ipc.NettyServer: [id: 0x7c2f0aec, /192.168.1.50:38104 => /192.168.1.51:5555] CONNECTED: /192.168.1.50:3810414/08/10 14:32:44 INFO ipc.NettyServer: [id: 0x3d36f553, /192.168.1.51:48599 => /192.168.1.51:5555] OPEN14/08/10 14:32:44 INFO ipc.NettyServer: [id: 0x3d36f553, /192.168.1.51:48599 => /192.168.1.51:5555] BOUND: /192.168.1.51:555514/08/10 14:32:44 INFO ipc.NettyServer: [id: 0x3d36f553, /192.168.1.51:48599 => /192.168.1.51:5555] CONNECTED: /192.168.1.51:4859914/08/10 14:34:33 INFO sink.LoggerSink: Event: { headers:{type=ali} body: 69 64 6F 61 6C 6C 5F 54 45 53 54 32 idoall_TEST2 }復制代碼 可以看到，根據header中不同的條件分布到不同的channel上 10)案例10：Flume Sink Processors failover的機器是一直發送給其中一個sink，當這個sink不可用的時候，自動發送到下一個sink。 a)在m1創建Flume_Sink_Processors配置文件root@m1:/home/hadoop# vi /home/hadoop/flume-1.5.0-bin/conf/Flume_Sink_Processors.confa1.sources = r1a1.sinks = k1 k2a1.channels = c1 c2#這個是配置failover的關鍵，需要有一個sink groupa1.sinkgroups = g1a1.sinkgroups.g1.sinks = k1 k2#處理的類型是failovera1.sinkgroups.g1.processor.type = failover#優先級，數字越大優先級越高，每個sink的優先級必須不相同a1.sinkgroups.g1.processor.priority.k1 = 5a1.sinkgroups.g1.processor.priority.k2 = 10#設置為10秒，當然可以根據你的實際狀況更改成更快或者很慢a1.sinkgroups.g1.processor.maxpenalty = 10000# Describe/configure the sourcea1.sources.r1.type = syslogtcpa1.sources.r1.port = 5140a1.sources.r1.channels = c1 c2a1.sources.r1.selector.type = replicating# Describe the sinka1.sinks.k1.type = avroa1.sinks.k1.channel = c1a1.sinks.k1.hostname = m1a1.sinks.k1.port = 5555a1.sinks.k2.type = avroa1.sinks.k2.channel = c2a1.sinks.k2.hostname = m2a1.sinks.k2.port = 5555# Use a channel which buffers events in memorya1.channels.c1.type = memorya1.channels.c1.capacity = 1000a1.channels.c1.transactionCapacity = 100a1.channels.c2.type = memorya1.channels.c2.capacity = 1000a1.channels.c2.transactionCapacity = 100復制代碼 b)在m1創建Flume_Sink_Processors_avro配置文件root@m1:/home/hadoop# vi /home/hadoop/flume-1.5.0-bin/conf/Flume_Sink_Processors_avro.confa1.sources = r1a1.sinks = k1a1.channels = c1# Describe/configure the sourcea1.sources.r1.type = avroa1.sources.r1.channels = c1a1.sources.r1.bind = 0.0.0.0a1.sources.r1.port = 5555# Describe the sinka1.sinks.k1.type = logger# Use a channel which buffers events in memorya1.channels.c1.type = memorya1.channels.c1.capacity = 1000a1.channels.c1.transactionCapacity = 100# Bind the source and sink to the channela1.sources.r1.channels = c1a1.sinks.k1.channel = c1復制代碼 c)將2個配置文件復制到m2上一份root@m1:/home/hadoop/flume-1.5.0-bin# scp -r /home/hadoop/flume-1.5.0-bin/conf/Flume_Sink_Processors.conf root@m2:/home/hadoop/flume-1.5.0-bin/conf/Flume_Sink_Processors.confroot@m1:/home/hadoop/flume-1.5.0-bin# scp -r /home/hadoop/flume-1.5.0-bin/conf/Flume_Sink_Processors_avro.conf root@m2:/home/hadoop/flume-1.5.0-bin/conf/Flume_Sink_Processors_avro.conf復制代碼 d)打開4個窗口，在m1和m2上同時啟動兩個flume agentroot@m1:/home/hadoop# /home/hadoop/flume-1.5.0-bin/bin/flume-ng agent -c . -f /home/hadoop/flume-1.5.0-bin/conf/Flume_Sink_Processors_avro.conf -n a1 -Dflume.root.logger=INFO,consoleroot@m1:/home/hadoop# /home/hadoop/flume-1.5.0-bin/bin/flume-ng agent -c . -f /home/hadoop/flume-1.5.0-bin/conf/Flume_Sink_Processors.conf -n a1 -Dflume.root.logger=INFO,console復制代碼 e)然后在m1或m2的任意一臺機器上，測試產生logroot@m1:/home/hadoop# echo "idoall.org test1 failover" | nc localhost 5140復制代碼 f)因為m2的優先級高，所以在m2的sink窗口，可以看到以下信息，而m1沒有：14/08/10 15:02:46 INFO ipc.NettyServer: Connection to /192.168.1.51:48692 disconnected.14/08/10 15:03:12 INFO ipc.NettyServer: [id: 0x09a14036, /192.168.1.51:48704 => /192.168.1.51:5555] OPEN14/08/10 15:03:12 INFO ipc.NettyServer: [id: 0x09a14036, /192.168.1.51:48704 => /192.168.1.51:5555] BOUND: /192.168.1.51:555514/08/10 15:03:12 INFO ipc.NettyServer: [id: 0x09a14036, /192.168.1.51:48704 => /192.168.1.51:5555] CONNECTED: /192.168.1.51:4870414/08/10 15:03:26 INFO sink.LoggerSink: Event: { headers:{Severity=0, flume.syslog.status=Invalid, Facility=0} body: 69 64 6F 61 6C 6C 2E 6F 72 67 20 74 65 73 74 31 idoall.org test1 }復制代碼 g)這時我們停止掉m2機器上的sink(ctrl+c)，再次輸出測試數據：root@m1:/home/hadoop# echo "idoall.org test2 failover" | nc localhost 5140復制代碼 h)可以在m1的sink窗口，看到讀取到了剛才發送的兩條測試數據：14/08/10 15:02:46 INFO ipc.NettyServer: Connection to /192.168.1.51:47036 disconnected.14/08/10 15:03:12 INFO ipc.NettyServer: [id: 0xbcf79851, /192.168.1.51:47048 => /192.168.1.50:5555] OPEN14/08/10 15:03:12 INFO ipc.NettyServer: [id: 0xbcf79851, /192.168.1.51:47048 => /192.168.1.50:5555] BOUND: /192.168.1.50:555514/08/10 15:03:12 INFO ipc.NettyServer: [id: 0xbcf79851, /192.168.1.51:47048 => /192.168.1.50:5555] CONNECTED: /192.168.1.51:4704814/08/10 15:07:56 INFO sink.LoggerSink: Event: { headers:{Severity=0, flume.syslog.status=Invalid, Facility=0} body: 69 64 6F 61 6C 6C 2E 6F 72 67 20 74 65 73 74 31 idoall.org test1 }14/08/10 15:07:56 INFO sink.LoggerSink: Event: { headers:{Severity=0, flume.syslog.status=Invalid, Facility=0} body: 69 64 6F 61 6C 6C 2E 6F 72 67 20 74 65 73 74 32 idoall.org test2 }復制代碼 i)我們再在m2的sink窗口中，啟動sink：root@m1:/home/hadoop# /home/hadoop/flume-1.5.0-bin/bin/flume-ng agent -c . -f /home/hadoop/flume-1.5.0-bin/conf/Flume_Sink_Processors_avro.conf -n a1 -Dflume.root.logger=INFO,console復制代碼 j)輸入兩批測試數據：root@m1:/home/hadoop# echo "idoall.org test3 failover" | nc localhost 5140 && echo "idoall.org test4 failover" | nc localhost 5140復制代碼 k)在m2的sink窗口，我們可以看到以下信息，因為優先級的關系，log消息會再次落到m2上：14/08/10 15:09:47 INFO node.Application: Starting Sink k114/08/10 15:09:47 INFO node.Application: Starting Source r114/08/10 15:09:47 INFO source.AvroSource: Starting Avro source r1: { bindAddress: 0.0.0.0, port: 5555 }...14/08/10 15:09:47 INFO instrumentation.MonitoredCounterGroup: Monitored counter group for type: SOURCE, name: r1: Successfully registered new MBean.14/08/10 15:09:47 INFO instrumentation.MonitoredCounterGroup: Component type: SOURCE, name: r1 started14/08/10 15:09:47 INFO source.AvroSource: Avro source r1 started.14/08/10 15:09:54 INFO ipc.NettyServer: [id: 0x96615732, /192.168.1.51:48741 => /192.168.1.51:5555] OPEN14/08/10 15:09:54 INFO ipc.NettyServer: [id: 0x96615732, /192.168.1.51:48741 => /192.168.1.51:5555] BOUND: /192.168.1.51:555514/08/10 15:09:54 INFO ipc.NettyServer: [id: 0x96615732, /192.168.1.51:48741 => /192.168.1.51:5555] CONNECTED: /192.168.1.51:4874114/08/10 15:09:57 INFO sink.LoggerSink: Event: { headers:{Severity=0, flume.syslog.status=Invalid, Facility=0} body: 69 64 6F 61 6C 6C 2E 6F 72 67 20 74 65 73 74 32 idoall.org test2 }14/08/10 15:10:43 INFO ipc.NettyServer: [id: 0x12621f9a, /192.168.1.50:38166 => /192.168.1.51:5555] OPEN14/08/10 15:10:43 INFO ipc.NettyServer: [id: 0x12621f9a, /192.168.1.50:38166 => /192.168.1.51:5555] BOUND: /192.168.1.51:555514/08/10 15:10:43 INFO ipc.NettyServer: [id: 0x12621f9a, /192.168.1.50:38166 => /192.168.1.51:5555] CONNECTED: /192.168.1.50:3816614/08/10 15:10:43 INFO sink.LoggerSink: Event: { headers:{Severity=0, flume.syslog.status=Invalid, Facility=0} body: 69 64 6F 61 6C 6C 2E 6F 72 67 20 74 65 73 74 33 idoall.org test3 }14/08/10 15:10:43 INFO sink.LoggerSink: Event: { headers:{Severity=0, flume.syslog.status=Invalid, Facility=0} body: 69 64 6F 61 6C 6C 2E 6F 72 67 20 74 65 73 74 34 idoall.org test4 }復制代碼 11)案例11：Load balancing Sink Processor load balance type和failover不同的地方是，load balance有兩個配置，一個是輪詢，一個是隨機。兩種情況下如果被選擇的sink不可用，就會自動嘗試發送到下一個可用的sink上面。 a)在m1創建Load_balancing_Sink_Processors配置文件root@m1:/home/hadoop# vi /home/hadoop/flume-1.5.0-bin/conf/Load_balancing_Sink_Processors.confa1.sources = r1a1.sinks = k1 k2a1.channels = c1#這個是配置Load balancing的關鍵，需要有一個sink groupa1.sinkgroups = g1a1.sinkgroups.g1.sinks = k1 k2a1.sinkgroups.g1.processor.type = load_balancea1.sinkgroups.g1.processor.backoff = truea1.sinkgroups.g1.processor.selector = round_robin# Describe/configure the sourcea1.sources.r1.type = syslogtcpa1.sources.r1.port = 5140a1.sources.r1.channels = c1# Describe the sinka1.sinks.k1.type = avroa1.sinks.k1.channel = c1a1.sinks.k1.hostname = m1a1.sinks.k1.port = 5555a1.sinks.k2.type = avroa1.sinks.k2.channel = c1a1.sinks.k2.hostname = m2a1.sinks.k2.port = 5555# Use a channel which buffers events in memorya1.channels.c1.type = memorya1.channels.c1.capacity = 1000a1.channels.c1.transactionCapacity = 100復制代碼 b)在m1創建Load_balancing_Sink_Processors_avro配置文件root@m1:/home/hadoop# vi /home/hadoop/flume-1.5.0-bin/conf/Load_balancing_Sink_Processors_avro.confa1.sources = r1a1.sinks = k1a1.channels = c1# Describe/configure the sourcea1.sources.r1.type = avroa1.sources.r1.channels = c1a1.sources.r1.bind = 0.0.0.0a1.sources.r1.port = 5555# Describe the sinka1.sinks.k1.type = logger# Use a channel which buffers events in memorya1.channels.c1.type = memorya1.channels.c1.capacity = 1000a1.channels.c1.transactionCapacity = 100# Bind the source and sink to the channela1.sources.r1.channels = c1a1.sinks.k1.channel = c1復制代碼 c)將2個配置文件復制到m2上一份root@m1:/home/hadoop/flume-1.5.0-bin# scp -r /home/hadoop/flume-1.5.0-bin/conf/Load_balancing_Sink_Processors.conf root@m2:/home/hadoop/flume-1.5.0-bin/conf/Load_balancing_Sink_Processors.confroot@m1:/home/hadoop/flume-1.5.0-bin# scp -r /home/hadoop/flume-1.5.0-bin/conf/Load_balancing_Sink_Processors_avro.conf root@m2:/home/hadoop/flume-1.5.0-bin/conf/Load_balancing_Sink_Processors_avro.conf復制代碼 d)打開4個窗口，在m1和m2上同時啟動兩個flume agentroot@m1:/home/hadoop# /home/hadoop/flume-1.5.0-bin/bin/flume-ng agent -c . -f /home/hadoop/flume-1.5.0-bin/conf/Load_balancing_Sink_Processors_avro.conf -n a1 -Dflume.root.logger=INFO,consoleroot@m1:/home/hadoop# /home/hadoop/flume-1.5.0-bin/bin/flume-ng agent -c . -f /home/hadoop/flume-1.5.0-bin/conf/Load_balancing_Sink_Processors.conf -n a1 -Dflume.root.logger=INFO,console復制代碼 e)然后在m1或m2的任意一臺機器上，測試產生log，一行一行輸入，輸入太快，容易落到一臺機器上root@m1:/home/hadoop# echo "idoall.org test1" | nc localhost 5140root@m1:/home/hadoop# echo "idoall.org test2" | nc localhost 5140root@m1:/home/hadoop# echo "idoall.org test3" | nc localhost 5140root@m1:/home/hadoop# echo "idoall.org test4" | nc localhost 5140復制代碼 f)在m1的sink窗口，可以看到以下信息：14/08/10 15:35:29 INFO sink.LoggerSink: Event: { headers:{Severity=0, flume.syslog.status=Invalid, Facility=0} body: 69 64 6F 61 6C 6C 2E 6F 72 67 20 74 65 73 74 32 idoall.org test2 }14/08/10 15:35:33 INFO sink.LoggerSink: Event: { headers:{Severity=0, flume.syslog.status=Invalid, Facility=0} body: 69 64 6F 61 6C 6C 2E 6F 72 67 20 74 65 73 74 34 idoall.org test4 }復制代碼 g)在m2的sink窗口，可以看到以下信息：14/08/10 15:35:27 INFO sink.LoggerSink: Event: { headers:{Severity=0, flume.syslog.status=Invalid, Facility=0} body: 69 64 6F 61 6C 6C 2E 6F 72 67 20 74 65 73 74 31 idoall.org test1 }14/08/10 15:35:29 INFO sink.LoggerSink: Event: { headers:{Severity=0, flume.syslog.status=Invalid, Facility=0} body: 69 64 6F 61 6C 6C 2E 6F 72 67 20 74 65 73 74 33 idoall.org test3 }復制代碼 說明輪詢模式起到了作用。 12)案例12：Hbase sink a)在測試之前，請先參考《Ubuntu12.04+hadoop2.2.0+zookeeper3.4.5+hbase0.96.2+hive0.13.1分布式環境部署》將hbase啟動 b)然后將以下文件復制到flume中：cp /home/hadoop/hbase-0.96.2-hadoop2/lib/protobuf-java-2.5.0.jar /home/hadoop/flume-1.5.0-bin/libcp /home/hadoop/hbase-0.96.2-hadoop2/lib/hbase-client-0.96.2-hadoop2.jar /home/hadoop/flume-1.5.0-bin/libcp /home/hadoop/hbase-0.96.2-hadoop2/lib/hbase-common-0.96.2-hadoop2.jar /home/hadoop/flume-1.5.0-bin/libcp /home/hadoop/hbase-0.96.2-hadoop2/lib/hbase-protocol-0.96.2-hadoop2.jar /home/hadoop/flume-1.5.0-bin/libcp /home/hadoop/hbase-0.96.2-hadoop2/lib/hbase-server-0.96.2-hadoop2.jar /home/hadoop/flume-1.5.0-bin/libcp /home/hadoop/hbase-0.96.2-hadoop2/lib/hbase-hadoop2-compat-0.96.2-hadoop2.jar /home/hadoop/flume-1.5.0-bin/libcp /home/hadoop/hbase-0.96.2-hadoop2/lib/hbase-hadoop-compat-0.96.2-hadoop2.jar /home/hadoop/flume-1.5.0-bin/lib@@@cp /home/hadoop/hbase-0.96.2-hadoop2/lib/htrace-core-2.04.jar /home/hadoop/flume-1.5.0-bin/lib復制代碼 c)確保test_idoall_org表在hbase中已經存在 d)在m1創建hbase_simple配置文件root@m1:/home/hadoop# vi /home/hadoop/flume-1.5.0-bin/conf/hbase_simple.confa1.sources = r1a1.sinks = k1a1.channels = c1# Describe/configure the sourcea1.sources.r1.type = syslogtcpa1.sources.r1.port = 5140a1.sources.r1.host = localhosta1.sources.r1.channels = c1# Describe the sinka1.sinks.k1.type = loggera1.sinks.k1.type = hbasea1.sinks.k1.table = test_idoall_orga1.sinks.k1.columnFamily = namea1.sinks.k1.column = idoalla1.sinks.k1.serializer = org.apache.flume.sink.hbase.RegexHbaseEventSerializera1.sinks.k1.channel = memoryChannel# Use a channel which buffers events in memorya1.channels.c1.type = memorya1.channels.c1.capacity = 1000a1.channels.c1.transactionCapacity = 100# Bind the source and sink to the channela1.sources.r1.channels = c1a1.sinks.k1.channel = c1復制代碼 e)啟動flume agent/home/hadoop/flume-1.5.0-bin/bin/flume-ng agent -c . -f /home/hadoop/flume-1.5.0-bin/conf/hbase_simple.conf -n a1 -Dflume.root.logger=INFO,console復制代碼 f)測試產生syslogroot@m1:/home/hadoop# echo "hello idoall.org from flume" | nc localhost 5140復制代碼 g)這時登錄到hbase中，可以發現新數據已經插入root@m1:/home/hadoop# /home/hadoop/hbase-0.96.2-hadoop2/bin/hbase shell2014-08-10 16:09:48,984 INFO [main] Configuration.deprecation: hadoop.native.lib is deprecated. Instead, use io.native.lib.availableHBase Shell; enter 'help<RETURN>' for list of supported commands.Type "exit<RETURN>" to leave the HBase ShellVersion 0.96.2-hadoop2, r1581096, Mon Mar 24 16:03:18 PDT 2014hbase(main):001:0> listTABLE SLF4J: Class path contains multiple SLF4J bindings.SLF4J: Found binding in [jar:file:/home/hadoop/hbase-0.96.2-hadoop2/lib/slf4j-log4j12-1.6.4.jar!/org/slf4j/impl/StaticLoggerBinder.class]SLF4J: Found binding in [jar:file:/home/hadoop/hadoop-2.2.0/share/hadoop/common/lib/slf4j-log4j12-1.7.5.jar!/org/slf4j/impl/StaticLoggerBinder.class]SLF4J: See http://www.slf4j.org/codes.html#multiple_bindings for an explanation.hbase2hive_idoall hive2hbase_idoall test_idoall_org 3 row(s) in 2.6880 seconds=> ["hbase2hive_idoall", "hive2hbase_idoall", "test_idoall_org"]hbase(main):002:0> scan "test_idoall_org"ROW COLUMN+CELL 10086 column=name:idoall, timestamp=1406424831473, value=idoallvalue 1 row(s) in 0.0550 secondshbase(main):003:0> scan "test_idoall_org"ROW COLUMN+CELL 10086 column=name:idoall, timestamp=1406424831473, value=idoallvalue 1407658495588-XbQCOZrKK8-0 column=name:payload, timestamp=1407658498203, value=hello idoall.org from flume 2 row(s) in 0.0200 secondshbase(main):004:0> quit復制代碼經過這么多flume的例子測試，如果你全部做完后，會發現flume的功能真的很強大，可以進行各種搭配來完成你想要的工作，俗話說師傅領進門，修行在個人，如何能夠結合你的產品業務，將flume更好的應用起來，快去動手實踐吧。新聞熱點
疑難解答
