flume打包遇到的一些问题
1.ipc兼容性问题，线上使用2.3.0的hdfs，但是打包时默认为1.2.1的
08 Apr 2015 19:38:25,122 WARN [SinkRunner-PollingRunner-DefaultSinkProcessor] (org.apache.flume.sink.hdfs.HDFSEventSink.process:455) - HDFS IO error org.apache.hadoop.ipc.RemoteException: Server IPC version 9 cannot communicate with client version 4 at org.apache.hadoop.ipc.Client.call(Client.java:1113) at org.apache.hadoop.ipc.RPC$Invoker.invoke(RPC.java:229)
拷贝hadoop-core包即可,或者更改pom.xml文件
<properties> <hadoop.version>1.2.1</hadoop.version> //hadoop-core-1.2.1.jar 改为hadoop-core-2.3.0-mr1-cdh5.1.0.jar <hbase.version>0.92.1</hbase.version> <hadoop.common.artifact.id>hadoop-core</hadoop.common.artifact.id> <thrift.version>0.7.0</thrift.version> </properties>
或者打包时指定：
mvn clean install -Phadoop-2
2.没有跳过test时
org.apache.flume.auth.TestFlumeAuthenticator: org/apache/commons/io/Charsets (no class error)
根目录下的pom.xml文件中更改为:
<dependency> <groupId>commons-io</groupId> <artifactId>commons-io</artifactId> <version>2.4</version> //2.1的包没有Charsets这个类 </dependency>
如果手动下载jar包可以使用如下命令导入：
mvn install:install-file -DgroupId=commons-io -DartifactId=commons-io -Dversion=2.0.1 -Dpackaging=jar -Dfile=commons-io-2.4.jar
3. 程序语言问题，因为在系统中使用了中文环境导致编译不通过
1）
@Test
public void shouldUseUtcAsBasisForDateFormat() {
assertEquals("Coordinated Universal Time",
factory.fastDateFormat.getTimeZone().getDisplayName()); //请求英文，返回中文
}
...2）
Running org.apache.flume.source.twitter.TestTwitterSource Tests run: 2, Failures: 0, Errors: 1, Skipped: 1, Time elapsed: 0.247 sec <<< FAILURE! testCarrotDateFormatBug(org.apache.flume.source.twitter.TestTwitterSource) Time elapsed: 21 sec <<< ERROR! java.text.ParseException: Unparseable date: "Fri Oct 26 22:53:55 +0000 2012" at java.text.DateFormat.parse(DateFormat.java:357) at org.apache.flume.source.twitter.TestTwitterSource.testCarrotDateFormatBug(TestTwitterSource.java:109)
代码：
@Test
public void testCarrotDateFormatBug() throws Exception {
SimpleDateFormat formatterFrom = new SimpleDateFormat("EEE MMM dd HH:mm:ss Z yyyy"); //更改为new SimpleDateFormat("EEE MMM dd HH:mm:ss Z yyyy",Locale.US);
formatterFrom.parse("Fri Oct 26 22:53:55 +0000 2012");
}或者简单点，更改语言设置或者直接skip test
mvn clean install -Phadoop-2 -DskipTests
4.json问题
替换单个jar包时，因为依赖问题，包class not found:
Exception in thread "PollableSourceRunner-KafkaSource-kafka1" java.lang.NoClassDefFoundError: net/sf/ezmorph/Morpher at org.apache.flume.source.kafka.KafkaSourceUtil.getDateMessage(KafkaSourceUtil.java:117) at org.apache.flume.source.kafka.KafkaSource.process(KafkaSource.java:123) at org.apache.flume.source.PollableSourceRunner$PollingRunner.run(PollableSourceRunner.java:139) at java.lang.Thread.run(Thread.java:745)
可以打成ensmbly包
flume-1.6.0/flume-ng-sources/flume-kafka-source/pom.xml <dependency> <groupId>net.sf.json-lib</groupId> <artifactId>json-lib</artifactId> <version>2.4</version> <classifier>jdk15</classifier> </dependency>
<plugin> <artifactId>maven-assembly-plugin</artifactId> <version>2.4</version> <configuration> <descriptorRefs> <descriptorRef>jar-with-dependencies</descriptorRef> </descriptorRefs> </configuration> <executions> <execution> <id>make-assembly</id> <phase>package</phase> <goals> <goal>single</goal> </goals> </execution> </executions> </plugin>
或者拷贝jar包：
/Users/nizengguang/.m2/repository/net/sf/ezmorph/ezmorph/1.0.6/ezmorph-1.0.6.jar /Users/nizengguang/.m2/repository/net/sf/json-lib/json-lib/2.4/json-lib-2.4-jdk15.jar /Users/nizengguang/.m2/repository/commons-beanutils/commons-beanutils/1.8.0/commons-beanutils-1.8.0.jar /Users/nizengguang/.m2/repository/commons-collections/commons-collections/3.2.1/commons-collections-3.2.1.jar /Users/nizengguang/.m2/repository/commons-logging/commons-logging/1.1.1/commons-logging-1.1.1.jar /Users/nizengguang/.m2/repository/commons-lang/commons-lang/2.5/commons-lang-2.5.jar
本文出自 “菜光光的博客” 博客，请务必保留此出处http://caiguangguang.blog.51cto.com/1652935/1635779
原文地址：http://caiguangguang.blog.51cto.com/1652935/1635779
