26 集群使用初步
HDFS的设计思路
l 设计思想
分而治之:将大文件、大批量文件,分布式存放在大量服务器上,以便于采取分而治之的方式对海量数据进行运算分析;
l 在大数据系统中作用:
为各类分布式运算框架(如:mapreduce,spark,tez,……)提供数据存储服务
l 重点概念:文件切块,副本存放,元数据
26.1 HDFS使用
1、查看集群状态
命令: hdfs dfsadmin –report
可以看出,集群共有3个datanode可用
也可打开web控制台查看HDFS集群信息,在浏览器打开http://hadoop:50070/
2、上传文件到HDFS
查看HDFS中的目录信息
命令:hadoop fs –ls /
上传文件
命令:hadoop fs -put ./findbugs-1.3.9 /
[[email protected] software]$ hadoop fs -put ./findbugs-1.3.9 / put: `/findbugs-1.3.9/LICENSE-ASM.txt’: File exists put: `/findbugs-1.3.9/LICENSE-AppleJavaExtensions.txt’: File exists put: `/findbugs-1.3.9/LICENSE-bcel.txt’: File exists put: `/findbugs-1.3.9/LICENSE-commons-lang.txt’: File exists put: `/findbugs-1.3.9/LICENSE-docbook.txt’: File exists put: `/findbugs-1.3.9/LICENSE-dom4j.txt’: File exists put: `/findbugs-1.3.9/LICENSE-jFormatString.txt’: File exists |
查看上传后的信息列表(hadoop fs –ls / 或 hadoop fs -ls /findbugs-1.3.9)
从HDFS下载文件
命令:hadoop fs -get /findbugs-1.3.9/LICENSE-ASM.txt
[[email protected] learn]$ cd /home/toto/learn /home/toto/learn [[email protected] learn]$ pwd /home/toto/learn [[email protected] learn]$ hadoop fs -get /findbugs-1.3.9/LICENSE-ASM.txt [[email protected] learn]$ ls LICENSE-ASM.txt |
yarn的管理界面是:http://hadoop:8088/cluster
26.2模拟运行一个mapreduce程序
模拟运行一个mapreduce程序的时候,需要先启动hdfs,启动命令是:
[[email protected] hadoop-2.8.0]$cd /home/toto/software/hadoop-2.8.0 [[email protected] hadoop-2.8.0]$sbin/start-dfs.sh |
在/home/toto/software/hadoop-2.8.0/share/hadoop/mapreduce下有一个mapreduce的运行例子:
[[email protected] mapreduce]$ cd /home/toto/software/hadoop-2.8.0/share/hadoop/mapreduce [[email protected] mapreduce]$ pwd /home/toto/software/hadoop-2.8.0/share/hadoop/mapreduce [[email protected] mapreduce]$ ll 总用量 5088 -rw-r–r–. 1 toto hadoop 562900 3月 17 13:31 hadoop-mapreduce-client-app-2.8.0.jar -rw-r–r–. 1 toto hadoop 782739 3月 17 13:31 hadoop-mapreduce-client-common-2.8.0.jar -rw-r–r–. 1 toto hadoop 1571179 3月 17 13:31 hadoop-mapreduce-client-core-2.8.0.jar -rw-r–r–. 1 toto hadoop 195000 3月 17 13:31 hadoop-mapreduce-client-hs-2.8.0.jar -rw-r–r–. 1 toto hadoop 31533 3月 17 13:31 hadoop-mapreduce-client-hs-plugins-2.8.0.jar -rw-r–r–. 1 toto hadoop 66999 3月 17 13:31 hadoop-mapreduce-client-jobclient-2.8.0.jar -rw-r–r–. 1 toto hadoop 1587158 3月 17 13:31 hadoop-mapreduce-client-jobclient-2.8.0-tests.jar -rw-r–r–. 1 toto hadoop 75495 3月 17 13:31 hadoop-mapreduce-client-shuffle-2.8.0.jar -rw-r–r–. 1 toto hadoop 301934 3月 17 13:31 hadoop-mapreduce-examples-2.8.0.jar drwxr-xr-x. 2 toto hadoop 4096 3月 17 13:31 jdiff drwxr-xr-x. 2 toto hadoop 4096 3月 17 13:31 lib drwxr-xr-x. 2 toto hadoop 4096 3月 17 13:31 lib-examples drwxr-xr-x. 2 toto hadoop 4096 3月 17 13:31 sources [[email protected] mapreduce]$
使用命令运行mapreduce命令: [[email protected] mapreduce]$ hadoop jar hadoop-mapreduce-examples-2.8.0.jar pi 5 5 Number of Maps = 5 Samples per Map = 5 Wrote input for Map #0 Wrote input for Map #1 Wrote input for Map #2 Wrote input for Map #3 Wrote input for Map #4 Starting Job 17/05/29 14:47:36 INFO client.RMProxy: Connecting to ResourceManager at hadoop/192.168.106.80:8032 17/05/29 14:47:37 INFO input.FileInputFormat: Total input files to process : 5 17/05/29 14:47:37 INFO mapreduce.JobSubmitter: number of splits:5 17/05/29 14:47:38 INFO mapreduce.JobSubmitter: Submitting tokens for job: job_1495998405307_0001 17/05/29 14:47:39 INFO impl.YarnClientImpl: Submitted application application_1495998405307_0001 17/05/29 14:47:39 INFO mapreduce.Job: The url to track the job: http://hadoop:8088/proxy/application_1495998405307_0001/ 17/05/29 14:47:39 INFO mapreduce.Job: Running job: job_1495998405307_0001 17/05/29 14:48:00 INFO mapreduce.Job: Job job_1495998405307_0001 running in uber mode : false 17/05/29 14:48:00 INFO mapreduce.Job: map 0% reduce 0% |
进入hdfs的管理界面(http://hadoop:8088/cluster/apps),查看程序运行情况:
26.2 MAPREDUCE使用
mapreduce是hadoop中的分布式运算编程框架,只要按照其编程规范,只需要编写少量的业务逻辑代码即可实现一个强大的海量数据并发处理程序
26.2.1 Demo开发——wordcount
1、需求
从大量(比如T级别)文本文件中,统计出每一个单词出现的总次数
2、mapreduce实现思路
Map阶段:
a) 从HDFS的源数据文件中逐行读取数据
b) 将每一行数据切分出单词
c) 为每一个单词构造一个键值对(单词,1)
d) 将键值对发送给reduce
Reduce阶段:
a) 接收map阶段输出的单词键值对
b) 将相同单词的键值对汇聚成一组
c) 对每一组,遍历组中的所有“值”,累加求和,即得到每一个单词的总次数
d) 将(单词,总次数)输出到HDFS的文件中
1、 具体编码实现
(1)定义一个mapper类
//首先要定义四个泛型的类型 //keyin: LongWritable valuein: Text //keyout: Text valueout:IntWritable
public class WordCountMapper extends Mapper<LongWritable, Text, Text, IntWritable>{ //map方法的生命周期: 框架每传一行数据就被调用一次 //key : 这一行的起始点在文件中的偏移量 //value: 这一行的内容 @Override protected void map(LongWritable key, Text value, Context context) throws IOException, InterruptedException { //拿到一行数据转换为string String line = value.toString(); //将这一行切分出各个单词 String[] words = line.split(” “); //遍历数组,输出<单词,1> for(String word:words){ context.write(new Text(word), new IntWritable(1)); } } } |
(2)定义一个reducer类
//生命周期:框架每传递进来一个kv 组,reduce方法被调用一次 @Override protected void reduce(Text key, Iterable<IntWritable> values, Context context) throws IOException, InterruptedException { //定义一个计数器 int count = 0; //遍历这一组kv的所有v,累加到count中 for(IntWritable value:values){ count += value.get(); } context.write(key, new IntWritable(count)); } } |
(3)定义一个主类,用来描述job并提交job
public class WordCountRunner { //把业务逻辑相关的信息(哪个是mapper,哪个是reducer,要处理的数据在哪里,输出的结果放哪里。。。。。。)描述成一个job对象 //把这个描述好的job提交给集群去运行 public static void main(String[] args) throws Exception { Configuration conf = new Configuration(); Job wcjob = Job.getInstance(conf); //指定我这个job所在的jar包 // wcjob.setJar(“/home/hadoop/wordcount.jar”); wcjob.setJarByClass(WordCountRunner.class);
wcjob.setMapperClass(WordCountMapper.class); wcjob.setReducerClass(WordCountReducer.class); //设置我们的业务逻辑Mapper类的输出key和value的数据类型 wcjob.setMapOutputKeyClass(Text.class); wcjob.setMapOutputValueClass(IntWritable.class); //设置我们的业务逻辑Reducer类的输出key和value的数据类型 wcjob.setOutputKeyClass(Text.class); wcjob.setOutputValueClass(IntWritable.class);
//指定要处理的数据所在的位置 FileInputFormat.setInputPaths(wcjob, “hdfs://hdp-server01:9000/wordcount/data/big.txt”); //指定处理完成之后的结果所保存的位置 FileOutputFormat.setOutputPath(wcjob, new Path(“hdfs://hdp-server01:9000/wordcount/output/”));
//向yarn集群提交这个job boolean res = wcjob.waitForCompletion(true); System.exit(res?0:1); } |
26.2.2 程序打包运行
1. 将程序打包
2. 准备输入数据
vi /home/hadoop/test.txt
Hello tom Hello jim Hello ketty Hello world Ketty tom |
在hdfs上创建输入数据文件夹:
hadoop fs mkdir -p /wordcount/input
将words.txt上传到hdfs上
hadoop fs –put /home/hadoop/words.txt /wordcount/input
3. 将程序jar包上传到集群的任意一台服务器上
4. 使用命令启动执行wordcount程序jar包
$ hadoop jar wordcount.jar cn.toto.bigdata.mrsimple.WordCountDriver/wordcount/input /wordcount/out
5. 查看执行结果
$ hadoop fs –cat /wordcount/out/part-r-00000
原创文章,作者:Maggie-Hunter,如若转载,请注明出处:https://blog.ytso.com/tech/bigdata/9638.html