用maven的ant run插件自动部署MR job依赖的jar到HDFS

背景

远程提交MR job时将MR job依赖的所有jar都发布到HDFS中，同时将这些jar加到hadoop job的classpath上。

如果每次手动上传就太费劲。项目集成maven。所以打算在maven package时将所有jar通过脚本上传到远程HDFS上

 

 

 

编写ant脚本

使用了，haddop的ant 插件。说白了就是一些util方法，调用了fsshell。没有文档，写的时候需要参考plugin的源代码。呵呵。

关键：

 

<hdfs cmd="rm" args="@{mapred.lib.dir}/*.jar" conf="@{hadoop.conf.dir}">

 

cmd：命令，参考hadoop fs

args：参数，逗号分隔

conf：core-site.xml的所在的文件目录

 

 

build.xml

 

<?xml version="1.0" encoding="UTF-8"?>
<!-- ========================================项目构建文件 ====================================================================== -->
<project>
	<import file="build.property.xml" />

	<!--
	<path id="dev_classpath">-->
	<!--
		<pathelement location="${resource.dir}"/>
		<pathelement location="${classes.dir}"/>-->
	<!--
		<fileset dir="${dev.ant.contrib.dir}">
			<include name="**/*.jar" />
		</fileset>
		
		<fileset dir="${dev.ant.contrib.lib.dir}">
				<include name="**/*.jar" />
			</fileset>
		
		<fileset dir="${dev.hadoop.home}">
			<include name="**/*.jar" />
		</fileset>
		
		<fileset dir="${dev.hadoop.home.lib}">
			<include name="**/*.jar" />
		</fileset>
		
	</path>-->

	<target name="init">
		<echo message="init copy and generate mapred libs">
		</echo>

		<taskdef resource="org/apache/maven/artifact/ant/antlib.xml">
		</taskdef>

		<taskdef resource="net/sf/antcontrib/antcontrib.properties">
		</taskdef>

		<!--<echo message="set classpath success"/>-->

		<taskdef resource="net/sf/antcontrib/antlib.xml">
		</taskdef>

		<taskdef resource="org/apache/hadoop/ant/antlib.xml">
		</taskdef>

		<echo message="import external antcontrib and hadoop ant extention task success">
		</echo>

		<!--
		<delete file="${resource.dir}/${mapred.lib.outfile}"/>
			<echo message="delete ${resource.dir}/${mapred.lib.outfile} success">
		</echo>
		<touch file="${resource.dir}/${mapred.lib.outfile}">
		</touch>
		<echo message="create ${resource.dir}/${mapred.lib.outfile} success">
		</echo>-->
	</target>

	<target name="dev_copy_libs">
		<foreach param="local.file" target="dev_copy_lib">
			<fileset dir="${lib.dir}" casesensitive="yes">
				<exclude name="**/hadoop*.jar" />
				<exclude name="**/hbase*.jar" />
				<exclude name="**/zookeeper*.jar" />
			</fileset>
			<fileset dir="target" casesensitive="yes">
				<include name="${project.jar}" />
			</fileset>
		</foreach>
	</target>

	<target name="dev_copy_lib">
		<hdfs cmd="copyFromLocal" args="${local.file},${dev.mapred.lib.dir}" conf="${dev.hadoop.conf.dir}">
		</hdfs>
		<echo message="copy ${local.file} to remote hdfs files file system: ${dev.mapred.lib.dir} success">
		</echo>
	</target>

	<macrodef name="macro_upload_mapred_lib" description="upload mapred lib">
		<attribute name="hadoop.conf.dir" />
		<attribute name="mapred.lib.dir" />
		<sequential>
			<property name="mapred.lib.dir" value="@{mapred.lib.dir}">
			</property>
			<echo message="hadoop conf dir: @{hadoop.conf.dir}">
			</echo>
			<hdfs cmd="rm" args="@{mapred.lib.dir}/*.jar" conf="@{hadoop.conf.dir}">
			</hdfs>
			<echo message="rm remote dir @{mapred.lib.dir}">
			</echo>
		</sequential>
	</macrodef>

	<target name="dev_upload_jars" depends="init">
		<macro_upload_mapred_lib hadoop.conf.dir="${dev.hadoop.conf.dir}" mapred.lib.dir="${dev.mapred.lib.dir}" />

		<echo message="----------------------------------------">
		</echo>
		<echo message="begin to copy libs to ${dev.mapred.lib.dir} exclude hadoop*, hbase*,">
		</echo>
		<antcall target="dev_copy_libs">
		</antcall>
		<echo message="all files has been copied to ${dev.mapred.lib.dir}">
		</echo>
		<echo message="----------------------------------------">
		</echo>
	</target>
</project>

 

 

build.properties.xml

 

<?xml version="1.0" encoding="UTF-8"?>
<project>
	<property file="build.properties"></property>
	<property name="classes.dir" value="target/classes"></property>
	<property name="lib.dir" value="lib"></property>
	<property name="resource.dir" value="src/main/resource"></property>
	<property name="mapred.lib.outfile" value="mapred_lib.properties"></property>
	<property name="lib.dir" value="lib"></property>
	<property name="project.jar"  value="${project.name}-${project.version}.jar"/>
	<!--
	<property name="dev.ant.contrib.dir" value="${dev.ant.contrib.dir}"></property>
	<property name="dev.ant.contrib.lib.dir" value="${dev.ant.contrib.dir}/lib"></property>
	<property name="dev.hadoop.home" value="${dev.hadoop.home}"></property>
	<property name="dev.hadoop.home.lib" value="${dev.hadoop.home}/lib"></property>-->
</project>

 

 

build.properties

 

src.conf.dir=src/main/conf
target.dir=target

#dev
dev.mapred.lib.dir=/user/root/mrlib/dev
dev.hadoop.conf.dir=${src.conf.dir}/dev

#test
test.mapred.lib.dir=/user/mrlib/test
test.hadoop.conf.dir=${src.conf.dir}/test

#testout
testout.mapred.lib.dir=/user/mrlib/testout
testout.hadoop.conf.dir=${src.conf.dir}/testout

 

 

配置pom.xml，因为我们使用了很多第三方的ant plugin。加入到ant plugin的dependency中。

 

 

	<plugin>
				<groupId>org.apache.maven.plugins</groupId>
				<artifactId>maven-antrun-plugin</artifactId>
				<version>1.7</version>
				<executions>
					<execution>
						<id>upload mapred jars</id>
						<phase>package</phase>
						<configuration>
							<target>
								<ant antfile="${basedir}/build.xml" inheritRefs="true">
									<target name="${envcfg.dir}_upload_jars" />
								</ant>
							</target>
						</configuration>
						<goals>
							<goal>run</goal>
						</goals>
					</execution>
				</executions>
				<dependencies>
					<dependency>
						<groupId>org.apache.hbase</groupId>
						<artifactId>hbase</artifactId>
						<version>0.94.1</version>
						<exclusions>
							<exclusion>
								<groupId>org.apache.hadoop</groupId>
								<artifactId>hadoop-core</artifactId>
							</exclusion>
						</exclusions>
					</dependency>
					<dependency>
						<groupId>ant-contrib</groupId>
						<artifactId>ant-contrib</artifactId>
						<version>1.0b3</version>
						<exclusions>
							<exclusion>
								<groupId>ant</groupId>
								<artifactId>ant</artifactId>
							</exclusion>
						</exclusions>
					</dependency>
					<dependency>
						<groupId>org.apache.ant</groupId>
						<artifactId>ant-nodeps</artifactId>
						<version>1.8.1</version>
					</dependency>
					<dependency>
						<groupId>org.apache.hadoop</groupId>
						<artifactId>hadoop-core</artifactId>
						<version>${hadoop.version}-modified</version>
						<optional>true</optional>
						<exclusions>
							<exclusion>
								<groupId>hsqldb</groupId>
								<artifactId>hsqldb</artifactId>
							</exclusion>
							<exclusion>
								<groupId>net.sf.kosmosfs</groupId>
								<artifactId>kfs</artifactId>
							</exclusion>
							<exclusion>
								<groupId>org.eclipse.jdt</groupId>
								<artifactId>core</artifactId>
							</exclusion>
							<exclusion>
								<groupId>net.java.dev.jets3t</groupId>
								<artifactId>jets3t</artifactId>
							</exclusion>
							<exclusion>
								<groupId>oro</groupId>
								<artifactId>oro</artifactId>
							</exclusion>
						</exclusions>
					</dependency>
					<dependency>
						<groupId>org.apache.hadoop</groupId>
						<artifactId>hadoop-ant</artifactId>
						<version>${hadoop.version}</version>
						<optional>true</optional>
						<scope>runtime</scope>
					</dependency>
					<dependency>
						<groupId>org.slf4j</groupId>
						<artifactId>slf4j-api</artifactId>
						<version>${slf4j.version}</version>
					</dependency>
					<dependency>
						<groupId>org.slf4j</groupId>
						<artifactId>slf4j-log4j12</artifactId>
						<version>${slf4j.version}</version>
					</dependency>
					<dependency>
						<groupId>org.apache.maven</groupId>
						<artifactId>maven-artifact-ant</artifactId>
						<version>2.0.4</version>
					</dependency>
				</dependencies>
			</plugin>	

 

 

 

 

运行：

执行maven install，看到所有jar都发布到了hdfs上。hoho。