Flume NG 学习笔记（二）单机与集群Flume 配置
下面的内容基本来自官网：http://flume.apache.org/FlumeUserGuide.html本文使用的是最新版本的apache flume 1.5，安装完Flume然后测试下Flume是否可以用，在Flume目录下用以下语句测试：
bin/flume-ng agent -n$agent_name -c conf -f conf/flume-conf.properties.template
结果如图显示：
http://img.blog.csdn.net/20141022140540140?watermark/2/text/aHR0cDovL2Jsb2cuY3Nkbi5uZXQvbG9va2xvb2s1/font/5a6L5L2T/fontsize/400/fill/I0JBQkFCMA==/dissolve/70/gravity/SouthEast
Ok,我们接下去看下面常用架构、功能配置示例
一、最简单的单一代理Flume 配置
下面是配置文件：
view plain copy
[*] #文件名：single_case1.conf.conf
[*] #配置内容：
[*] #single_case1.conf.conf: A single-node Flume configuration
[*] #Name the components on this agent
[*] a1.sources= r1
[*] a1.sinks= k1
[*] a1.channels= c1
[*]
[*] #Describe/configure the source
[*] a1.sources.r1.type= netcat
[*] a1.sources.r1.bind= localhost
[*] a1.sources.r1.port= 44444
[*]
[*] #Describe the sink
[*] 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
说明下，这里所有的例子都是将配置文件放到 $FLUME_HOME/conf 目录下，后面就不赘述了。
#敲命令
flume-ng agent -cconf -f conf/single_case1.conf -n a1 -Dflume.root.logger=INFO,console
#参数命令
-c conf 指定配置目录为conf
-f conf/single_case1.conf指定配置文件为conf/single_case1.conf
-n a1 指定agent名字为a1,需要与case1_example.conf中的一致
-Dflume.root.logger=INFO,console指定DEBUF模式在console输出INFO信息
具体参数命令请通过flume-nghelp查看
#然后在另一个终端进行测试
telnet 127.0.0.1 44444
http://img.blog.csdn.net/20141022140643375?watermark/2/text/aHR0cDovL2Jsb2cuY3Nkbi5uZXQvbG9va2xvb2s1/font/5a6L5L2T/fontsize/400/fill/I0JBQkFCMA==/dissolve/70/gravity/SouthEast
然后会看在之前启动的终端查看console输出到如下：
http://img.blog.csdn.net/20141022141103885?watermark/2/text/aHR0cDovL2Jsb2cuY3Nkbi5uZXQvbG9va2xvb2s1/font/5a6L5L2T/fontsize/400/fill/I0JBQkFCMA==/dissolve/70/gravity/SouthEast
这里会发现消息hello world! 输出了，而hello world! hello world!hello world!则被拦截了。因为在配置文件中，我们选择的输出方式为：a1.sinks.k1.type= logger
，即console输出，flume-ng针对logger是只显示16个字节的，剩下的都被sink截了。下面是源码
在LoggerSink.Java中：
view plain copy
[*] if(event != null) {
[*] if (logger.isInfoEnabled()) {
[*] logger.info("Event: " + EventHelper.dumpEvent(event));
[*] }
[*] }
我们去看EventHelper.java的dumpEvent方法：
view plain copy
[*] privatestatic final int DEFAULT_MAX_BYTES = 16;
[*] publicstatic String dumpEvent(Event event) {
[*] return dumpEvent(event, DEFAULT_MAX_BYTES);
[*] }
[*]
[*] publicstatic String dumpEvent(Event event, int maxBytes) {
[*] StringBuilder buffer = new StringBuilder();
[*] if (event == null || event.getBody() == null) {
[*] buffer.append("null");
[*] } else if (event.getBody().length == 0) {
[*] // do nothing... in this case, HexDump.dump() will throw anexception
[*] } else {
[*] byte[] body = event.getBody();
[*] byte[] data = Arrays.copyOf(body, Math.min(body.length,maxBytes));
[*] ByteArrayOutputStream out = new ByteArrayOutputStream();
[*] try {
[*] HexDump.dump(data, 0, out, 0);
[*] String hexDump = new String(out.toByteArray());
[*] // remove offset since it's not relevant for such a smalldataset
[*] if(hexDump.startsWith(HEXDUMP_OFFSET)) {
[*] hexDump =hexDump.substring(HEXDUMP_OFFSET.length());
[*] }
[*] buffer.append(hexDump);
[*] } catch (Exception e) {
[*] if(LOGGER.isInfoEnabled()) {
[*] LOGGER.info("Exception while dumpingevent", e);
[*] }
[*] buffer.append("...Exception while dumping:").append(e.getMessage());
[*] }
[*] String result = buffer.toString();
[*] if(result.endsWith(EOL) && buffer.length() >EOL.length()) {
[*] buffer.delete(buffer.length() - EOL.length(),buffer.length()).toString();
[*] }
[*] }
[*] return "{ headers:" + event.getHeaders() + " body:"+ buffer + " }";
[*] }
不难看出，在event处理过程中，发生了数据截取操作。
Ok，进入下一个环节。
二、“集群”代理Flume 配置
http://img.blog.csdn.net/20141022141145881?watermark/2/text/aHR0cDovL2Jsb2cuY3Nkbi5uZXQvbG9va2xvb2s1/font/5a6L5L2T/fontsize/400/fill/I0JBQkFCMA==/dissolve/70/gravity/SouthEast
这里集群的概念是多台机器的管理，最简单的就是两台机器一台代理主机从数据源获取数据，然后将数据在传送到另一台主机上，进行输出。这样做的意义是，一个业务多数据源的时候，我们可以对每个数据源设置代理，然后将它们汇总到一台代理主机上进行输出。
下面实现最简单的集群配置，即两个代理，一台接受数据源数据的代理将数据推送到汇总的代理，而汇总的代理再将数据输出。因此这两台主机分别是push,pull
根据上图需要用AVRO RPC通信，因此推数据sinks类型与拉数据的sources的类型都是avro 。而拉数据代理的数据源，我们用前文讲的Spool Source 形式来处理，这里我们预先建好目录与文件，test.log
http://img.blog.csdn.net/20141022140833562?watermark/2/text/aHR0cDovL2Jsb2cuY3Nkbi5uZXQvbG9va2xvb2s1/font/5a6L5L2T/fontsize/400/fill/I0JBQkFCMA==/dissolve/70/gravity/SouthEast
下面设置推代理主机的flume配置文件：
view plain copy
[*] #推数据代理的配置文件push.conf
[*] #Name the components on this agent
[*] a2.sources= r1
[*] a2.sinks= k1
[*] a2.channels= c1
[*]
[*] #Describe/configure the source
[*] a2.sources.r1.type= spooldir
[*] a2.sources.r1.spoolDir= /tmp/logs
[*] a2.sources.r1.channels= c1
[*]
[*] #Use a channel which buffers events in memory
[*] a2.channels.c1.type= memory
[*] a2.channels.c1.keep-alive= 10
[*] a2.channels.c1.capacity= 100000
[*] a2.channels.c1.transactionCapacity= 100000
[*]
[*] #Describe/configure the source
[*] a2.sinks.k1.type= avro
[*] a2.sinks.k1.channel= c1
[*] a2.sinks.k1.hostname= pull
[*] a2.sinks.k1.port= 4444
下面设置汇总代理主机的flume配置文件：
view plain copy
[*] #汇总数据代理的配置文件pull.conf
[*] #Name the components on this agent
[*] a1.sources= r1
[*] a1.sinks= k1
[*] a1.channels= c1
[*]
[*] #Describe/configure the source
[*] a1.sources.r1.type= avro
[*] a1.sources.r1.channels= c1
[*] a1.sources.r1.bind= pull
[*] a1.sources.r1.port= 44444
[*]
[*] #Describe the sink
[*] a1.sinks.k1.type= logger
[*] a1.sinks.k1.channel = c1
[*]
[*] #Use a channel which buffers events in memory
[*] a1.channels.c1.type= memory
[*] a1.channels.c1.keep-alive= 10
[*] a1.channels.c1.capacity= 100000
[*] a1.channels.c1.transactionCapacity= 100000
虽然Spool Source是非实时的，但由于数据量少，处理还是很快的，因此我们只能先启动pull代理。
#敲命令
flume-ng agent -c conf -f conf/pull.conf -n a1 -Dflume.root.logger=INFO,console
http://img.blog.csdn.net/20141022141334426?watermark/2/text/aHR0cDovL2Jsb2cuY3Nkbi5uZXQvbG9va2xvb2s1/font/5a6L5L2T/fontsize/400/fill/I0JBQkFCMA==/dissolve/70/gravity/SouthEast
上图显示成功。
先后去启动push主机的flume
#敲命令
flume-ng agent -n a2 -c conf -f conf/push.conf -Dflume.root.logger=INFO,console
http://img.blog.csdn.net/20141022141247064?watermark/2/text/aHR0cDovL2Jsb2cuY3Nkbi5uZXQvbG9va2xvb2s1/font/5a6L5L2T/fontsize/400/fill/I0JBQkFCMA==/dissolve/70/gravity/SouthEast
查看pull主机的状态，发现数据已经传过来了。
然后会过去看push主机的文件
http://img.blog.csdn.net/20141022141420945?watermark/2/text/aHR0cDovL2Jsb2cuY3Nkbi5uZXQvbG9va2xvb2s1/font/5a6L5L2T/fontsize/400/fill/I0JBQkFCMA==/dissolve/70/gravity/SouthEast
已经加上后缀名.COMPLETED。这与前文说的是一致的。
下面只要将新数据存入到目录/tmp/logs，push主机就会将数据发送到pull主机输出，并修改新数据文件的文件名。
页:
[1]