Hadoop离线_网站流量日志数据分析系统_数据的预处理

---

1.主要目的

数据清洗 —— 过滤“不合规”数据，清洗无意义的数据

2.实现方式

首先经过flume采集后的数据会有十个字段，每个字段都会由空格来分隔

3.开发一个MapReduce程序WeblogPreProcess

package cn.itcast.bigdata.weblog.pre;

import java.io.IOException;
import java.net.URI;
import java.text.SimpleDateFormat;
import java.util.HashSet;
import java.util.Set;

import cn.itcast.bigdata.weblog.utils.DateUtil;
import org.apache.hadoop.conf.Configuration;
import org.apache.hadoop.conf.Configured;
import org.apache.hadoop.fs.FileSystem;
import org.apache.hadoop.fs.Path;
import org.apache.hadoop.io.LongWritable;
import org.apache.hadoop.io.NullWritable;
import org.apache.hadoop.io.Text;
import org.apache.hadoop.mapreduce.Job;
import org.apache.hadoop.mapreduce.Mapper;
import org.apache.hadoop.mapreduce.lib.input.FileInputFormat;
import org.apache.hadoop.mapreduce.lib.input.TextInputFormat;
import org.apache.hadoop.mapreduce.lib.output.FileOutputFormat;

import cn.itcast.bigdata.weblog.mrbean.WebLogBean;
import org.apache.hadoop.mapreduce.lib.output.TextOutputFormat;
import org.apache.hadoop.util.Tool;
import org.apache.hadoop.util.ToolRunner;

/**
 * 处理原始日志，过滤出真实pv请求 转换时间格式 对缺失字段填充默认值 对记录标记valid和invalid
 * 
 */

public class WeblogPreProcess extends Configured implements Tool {

	@Override
	public int run(String[] args) throws Exception {
		//Configuration conf = new Configuration();
		Configuration conf = super.getConf();
		Job job = Job.getInstance(conf);
		/*String inputPath= "hdfs://node01:9000/weblog/"+DateUtil.getYestDate()+"/input";
		String outputPath="hdfs://node01:9000/weblog/"+DateUtil.getYestDate()+"/weblogPreOut";
		FileSystem fileSystem = FileSystem.get(new URI("hdfs://node01:9000"), conf);
		if (fileSystem.exists(new Path(outputPath))){
			fileSystem.delete(new Path(outputPath),true);
		}
		fileSystem.close();
		FileInputFormat.setInputPaths(job, new Path(inputPath));
		FileOutputFormat.setOutputPath(job, new Path(outputPath));
		job.setInputFormatClass(TextInputFormat.class);
		job.setOutputFormatClass(TextOutputFormat.class);
*/
		FileInputFormat.addInputPath(job,new Path("file:Users/zhaozhuang/Desktop/8.大数据离线第八天/日志文件数据/input"));
		job.setInputFormatClass(TextInputFormat.class);
		FileOutputFormat.setOutputPath(job,new Path("file:Users/zhaozhuang/Desktop/8.大数据离线第八天/日志文件数据/weblogPreOut"));
		job.setOutputFormatClass(TextOutputFormat.class);
		job.setJarByClass(WeblogPreProcess.class);
		job.setMapperClass(WeblogPreProcessMapper.class);
		job.setOutputKeyClass(Text.class);
		job.setOutputValueClass(NullWritable.class);
		job.setNumReduceTasks(0);
		boolean res = job.waitForCompletion(true);
		return res?0:1;
	}

	static class WeblogPreProcessMapper extends Mapper<LongWritable, Text, Text, NullWritable> {
		// 用来存储网站url分类数据
		Set<String> pages = new HashSet<String>();
		Text k = new Text();
		NullWritable v = NullWritable.get();

		/**
		 * 从外部配置文件中加载网站的有用url分类数据 存储到maptask的内存中，用来对日志数据进行过滤
		 */
		@Override
		protected void setup(Context context) throws IOException, InterruptedException {
			pages.add("/about");
			pages.add("/black-ip-list/");
			pages.add("/cassandra-clustor/");
			pages.add("/finance-rhive-repurchase/");
			pages.add("/hadoop-family-roadmap/");
			pages.add("/hadoop-hive-intro/");
			pages.add("/hadoop-zookeeper-intro/");
			pages.add("/hadoop-mahout-roadmap/");

		}

		@Override
		protected void map(LongWritable key, Text value, Context context) throws IOException, InterruptedException {
			String line = value.toString();
			WebLogBean webLogBean = WebLogParser.parser(line);
			if (webLogBean != null) {
				// 过滤js/图片/css等静态资源
				WebLogParser.filtStaticResource(webLogBean, pages);
				/* if (!webLogBean.isValid()) return; */
				k.set(webLogBean.toString());
				context.write(k, v);
			}
		}
	}

	public static void main(String[] args) throws Exception {
		Configuration configuration = new Configuration();
		int run = ToolRunner.run(configuration, new WeblogPreProcess(), args);
		System.exit(run);
	}
}

得到的数据：

4.点击流模型PageViews表

由于大量的指标统计从点击流模型中更容易得出，所以在预处理阶段，可以使用mr程序来生成点击流模型的数据。
有结构化数据转换为pageView模型的思路：
1.相同ip的数据放到一起按照时间排序，排序后打上标识
2.同一个ip的两条数据之间的时间差，如果大于30分，就不是同一个session，如果小于30分，就认为是同一个session
3.以ip作为key2，相同的数据发送到同一个reduce形成一个集合

package cn.itcast.bigdata.weblog.clickstream;

import cn.itcast.bigdata.weblog.mrbean.WebLogBean;
import cn.itcast.bigdata.weblog.utils.DateUtil;
import org.apache.commons.beanutils.BeanUtils;
import org.apache.hadoop.conf.Configuration;
import org.apache.hadoop.conf.Configured;
import org.apache.hadoop.fs.FileSystem;
import org.apache.hadoop.fs.Path;
import org.apache.hadoop.io.LongWritable;
import org.apache.hadoop.io.NullWritable;
import org.apache.hadoop.io.Text;
import org.apache.hadoop.mapreduce.Job;
import org.apache.hadoop.mapreduce.Mapper;
import org.apache.hadoop.mapreduce.Reducer;
import org.apache.hadoop.mapreduce.lib.input.FileInputFormat;
import org.apache.hadoop.mapreduce.lib.input.TextInputFormat;
import org.apache.hadoop.mapreduce.lib.output.FileOutputFormat;
import org.apache.hadoop.mapreduce.lib.output.TextOutputFormat;
import org.apache.hadoop.util.Tool;
import org.apache.hadoop.util.ToolRunner;

import java.io.IOException;
import java.net.URI;
import java.text.ParseException;
import java.text.SimpleDateFormat;
import java.util.*;

/**
 * 
 * 将清洗之后的日志梳理出点击流pageviews模型数据
 * 
 * 输入数据是清洗过后的结果数据
 * 
 * 区分出每一次会话，给每一次visit（session）增加了session-id（随机uuid）
 * 梳理出每一次会话中所访问的每个页面（请求时间，url，停留时长，以及该页面在这次session中的序号）
 * 保留referral_url，body_bytes_send，useragent
 * 
 * 
 * @author
 * 
 */
public class ClickStreamPageView extends Configured implements Tool {

	@Override
	public int run(String[] args) throws Exception {
		Configuration conf = super.getConf();
		Job job = Job.getInstance(conf);
		/*String inputPath="hdfs://node01:9000/weblog/"+DateUtil.getYestDate()+"/weblogPreOut";
		String outputPath="hdfs://node01:9000/weblog/"+DateUtil.getYestDate()+"/pageViewOut";
		FileSystem fileSystem = FileSystem.get(new URI("hdfs://node01:9000"), conf);
		if (fileSystem.exists(new Path(outputPath))){
			fileSystem.delete(new Path(outputPath),true);
		}
		fileSystem.close();
		job.setInputFormatClass(TextInputFormat.class);
		job.setOutputFormatClass(TextOutputFormat.class);
		FileInputFormat.setInputPaths(job, new Path(inputPath));
		FileOutputFormat.setOutputPath(job, new Path(outputPath));*/

		job.setInputFormatClass(TextInputFormat.class);
		job.setOutputFormatClass(TextOutputFormat.class);
		TextInputFormat.addInputPath(job,new Path("file:Users/zhaozhuang/Desktop/8.大数据离线第八天/日志文件数据/weblogPreOut"));
		TextOutputFormat.setOutputPath(job,new Path("file:Users/zhaozhuang/Desktop/8.大数据离线第八天/日志文件数据/pageViewOut"));

		job.setJarByClass(ClickStreamPageView.class);
		job.setMapperClass(ClickStreamMapper.class);
		job.setReducerClass(ClickStreamReducer.class);
		job.setMapOutputKeyClass(Text.class);
		job.setMapOutputValueClass(WebLogBean.class);
		job.setOutputKeyClass(Text.class);
		job.setOutputValueClass(Text.class);

		boolean b = job.waitForCompletion(true);
		return b?0:1;
	}

	static class ClickStreamMapper extends Mapper<LongWritable, Text, Text, WebLogBean> {
		Text k = new Text();
		WebLogBean v = new WebLogBean();
		@Override
		protected void map(LongWritable key, Text value, Context context) throws IOException, InterruptedException {
			String line = value.toString();
			String[] fields = line.split("\001");
			if (fields.length < 9) return;
			//将切分出来的各字段set到weblogbean中
			v.set("true".equals(fields[0]) ? true : false, fields[1], fields[2], fields[3], fields[4], fields[5], fields[6], fields[7], fields[8]);
			//只有有效记录才进入后续处理
			if (v.isValid()) {
			        //此处用ip地址来标识用户
				k.set(v.getRemote_addr());
				context.write(k, v);
			}
		}
	}

	static class ClickStreamReducer extends Reducer<Text, WebLogBean, NullWritable, Text> {
		Text v = new Text();
		@Override
		protected void reduce(Text key, Iterable<WebLogBean> values, Context context) throws IOException, InterruptedException {
			ArrayList<WebLogBean> beans = new ArrayList<WebLogBean>();
			// 先将一个用户的所有访问记录中的时间拿出来排序
			try {
				for (WebLogBean bean : values) {
					//为什么list集合当中不能直接添加循环出来的这个bean？
					//这里通过属性拷贝，每次new  一个对象，避免了bean的属性值每次覆盖
					WebLogBean webLogBean = new WebLogBean();
					try {
						BeanUtils.copyProperties(webLogBean, bean);
					} catch(Exception e) {
						e.printStackTrace();
					}
					beans.add(webLogBean);
				}
				//将bean按时间先后顺序排序
				Collections.sort(beans, new Comparator<WebLogBean>() {
					@Override
					public int compare(WebLogBean o1, WebLogBean o2) {
						try {
							Date d1 = toDate(o1.getTime_local());
							Date d2 = toDate(o2.getTime_local());
							if (d1 == null || d2 == null)
								return 0;
							return d1.compareTo(d2);
						} catch (Exception e) {
							e.printStackTrace();
							return 0;
						}
					}

				});

				/**
				 * 以下逻辑为：从有序bean中分辨出各次visit，并对一次visit中所访问的page按顺序标号step
				 * 核心思想：
				 * 就是比较相邻两条记录中的时间差，如果时间差<30分钟，则该两条记录属于同一个session
				 * 否则，就属于不同的session
				 * 
				 */
				
				int step = 1;
				String session = UUID.randomUUID().toString();
				for (int i = 0; i < beans.size(); i++) {
					WebLogBean bean = beans.get(i);
					// 如果仅有1条数据，则直接输出
					if (1 == beans.size()) {
						
						// 设置默认停留时长为60s
						v.set(session+"\001"+key.toString()+"\001"+bean.getRemote_user() + "\001" + bean.getTime_local() + "\001" + bean.getRequest() + "\001" + step + "\001" + (60) + "\001" + bean.getHttp_referer() + "\001" + bean.getHttp_user_agent() + "\001" + bean.getBody_bytes_sent() + "\001"
								+ bean.getStatus());
						context.write(NullWritable.get(), v);
						session = UUID.randomUUID().toString();
						break;
					}

					// 如果不止1条数据，则将第一条跳过不输出，遍历第二条时再输出
					if (i == 0) {
						continue;
					}
					// 求近两次时间差
					long timeDiff = timeDiff(toDate(bean.getTime_local()), toDate(beans.get(i - 1).getTime_local()));
					// 如果本次-上次时间差<30分钟，则输出前一次的页面访问信息
					if (timeDiff < 30 * 60 * 1000) {
						
						v.set(session+"\001"+key.toString()+"\001"+beans.get(i - 1).getRemote_user() + "\001" + beans.get(i - 1).getTime_local() + "\001" + beans.get(i - 1).getRequest() + "\001" + step + "\001" + (timeDiff / 1000) + "\001" + beans.get(i - 1).getHttp_referer() + "\001"
								+ beans.get(i - 1).getHttp_user_agent() + "\001" + beans.get(i - 1).getBody_bytes_sent() + "\001" + beans.get(i - 1).getStatus());
						context.write(NullWritable.get(), v);
						step++;
					} else {
						
						// 如果本次-上次时间差>30分钟，则输出前一次的页面访问信息且将step重置，以分隔为新的visit
						v.set(session+"\001"+key.toString()+"\001"+beans.get(i - 1).getRemote_user() + "\001" + beans.get(i - 1).getTime_local() + "\001" + beans.get(i - 1).getRequest() + "\001" + (step) + "\001" + (60) + "\001" + beans.get(i - 1).getHttp_referer() + "\001"
								+ beans.get(i - 1).getHttp_user_agent() + "\001" + beans.get(i - 1).getBody_bytes_sent() + "\001" + beans.get(i - 1).getStatus());
						context.write(NullWritable.get(), v);
						// 输出完上一条之后，重置step编号
						step = 1;
						session = UUID.randomUUID().toString();
					}

					// 如果此次遍历的是最后一条，则将本条直接输出
					if (i == beans.size() - 1) {
						// 设置默认停留市场为60s
						v.set(session+"\001"+key.toString()+"\001"+bean.getRemote_user() + "\001" + bean.getTime_local() + "\001" + bean.getRequest() + "\001" + step + "\001" + (60) + "\001" + bean.getHttp_referer() + "\001" + bean.getHttp_user_agent() + "\001" + bean.getBody_bytes_sent() + "\001" + bean.getStatus());
						context.write(NullWritable.get(), v);
					}
				}

			} catch (ParseException e) {
				e.printStackTrace();

			}

		}

		private String toStr(Date date) {
			SimpleDateFormat df = new SimpleDateFormat("yyyy-MM-dd HH:mm:ss", Locale.US);
			return df.format(date);
		}

		private Date toDate(String timeStr) throws ParseException {
			SimpleDateFormat df = new SimpleDateFormat("yyyy-MM-dd HH:mm:ss", Locale.US);
			return df.parse(timeStr);
		}

		private long timeDiff(String time1, String time2) throws ParseException {
			Date d1 = toDate(time1);
			Date d2 = toDate(time2);
			return d1.getTime() - d2.getTime();

		}

		private long timeDiff(Date time1, Date time2) throws ParseException {

			return time1.getTime() - time2.getTime();

		}

	}

	public static void main(String[] args) throws Exception {
		int run = ToolRunner.run(new Configuration(), new ClickStreamPageView(), args);
		System.exit(run);
	}
}

得到的数据

5.点击流模型visit信息表

注：“一次访问”=“N次连续请求”
直接从原始数据中用hql语法得出每个人的“次”访问信息比较困难，可先用mapreduce程序分析原始数据得出“次”信息数据，然后再用hql进行更多维度统计

package cn.itcast.bigdata.weblog.clickstream;

import java.io.IOException;
import java.net.URI;
import java.util.ArrayList;
import java.util.Collections;
import java.util.Comparator;

import cn.itcast.bigdata.weblog.utils.DateUtil;
import org.apache.commons.beanutils.BeanUtils;
import org.apache.hadoop.conf.Configuration;
import org.apache.hadoop.conf.Configured;
import org.apache.hadoop.fs.FileSystem;
import org.apache.hadoop.fs.Path;
import org.apache.hadoop.io.LongWritable;
import org.apache.hadoop.io.NullWritable;
import org.apache.hadoop.io.Text;
import org.apache.hadoop.mapreduce.Job;
import org.apache.hadoop.mapreduce.Mapper;
import org.apache.hadoop.mapreduce.Reducer;
import org.apache.hadoop.mapreduce.lib.input.FileInputFormat;
import org.apache.hadoop.mapreduce.lib.input.TextInputFormat;
import org.apache.hadoop.mapreduce.lib.output.FileOutputFormat;

import cn.itcast.bigdata.weblog.mrbean.PageViewsBean;
import cn.itcast.bigdata.weblog.mrbean.VisitBean;
import org.apache.hadoop.mapreduce.lib.output.TextOutputFormat;
import org.apache.hadoop.util.Tool;
import org.apache.hadoop.util.ToolRunner;


/**
 * 输入数据：pageviews模型结果数据
 * 从pageviews模型结果数据中进一步梳理出visit模型
 * sessionid  start-time   out-time   start-page   out-page   pagecounts  ......
 * 
 * @author
 *
 */
public class ClickStreamVisit extends Configured implements Tool {
	@Override
	public int run(String[] args) throws Exception {
		Configuration conf = super.getConf();
		Job job = Job.getInstance(conf);
		/*String inputPath = "hdfs://node01:9000/weblog/"+ DateUtil.getYestDate() + "/pageViewOut";
		String outPutPath="hdfs://node01:9000/weblog/"+ DateUtil.getYestDate() + "/clickStreamVisit";
		FileSystem fileSystem = FileSystem.get(new URI("hdfs://node01:9000"),conf);
		if (fileSystem.exists(new Path(outPutPath))){
			fileSystem.delete(new Path(outPutPath),true);
		}
		fileSystem.close();
		FileInputFormat.setInputPaths(job, new Path(inputPath));
		FileOutputFormat.setOutputPath(job, new Path(outPutPath));
		job.setInputFormatClass(TextInputFormat.class);
		job.setOutputFormatClass(TextOutputFormat.class);*/

		job.setInputFormatClass(TextInputFormat.class);
		job.setOutputFormatClass(TextOutputFormat.class);
		TextInputFormat.addInputPath(job,new Path("file:Users/zhaozhuang/Desktop/8.大数据离线第八天/日志文件数据/pageViewOut"));
		TextOutputFormat.setOutputPath(job,new Path("file:Users/zhaozhuang/Desktop/8.大数据离线第八天/日志文件数据/clickStreamVisit"));


		job.setJarByClass(ClickStreamVisit.class);
		job.setMapperClass(ClickStreamVisitMapper.class);
		job.setReducerClass(ClickStreamVisitReducer.class);
		job.setMapOutputKeyClass(Text.class);
		job.setMapOutputValueClass(PageViewsBean.class);
		job.setOutputKeyClass(NullWritable.class);
		job.setOutputValueClass(VisitBean.class);
		boolean res = job.waitForCompletion(true);
		return res?0:1;
	}

	// 以session作为key，发送数据到reducer
	static class ClickStreamVisitMapper extends Mapper<LongWritable, Text, Text, PageViewsBean> {

		PageViewsBean pvBean = new PageViewsBean();
		Text k = new Text();
		@Override
		protected void map(LongWritable key, Text value, Context context) throws IOException, InterruptedException {

			String line = value.toString();
			String[] fields = line.split("\001");
			int step = Integer.parseInt(fields[5]);
			//(String session, String remote_addr, String timestr, String request, int step, String staylong, String referal, String useragent, String bytes_send, String status)
			//299d6b78-9571-4fa9-bcc2-f2567c46df3472.46.128.140-2013-09-18 07:58:50/hadoop-zookeeper-intro/160"https://www.google.com/""Mozilla/5.0"14722200
			pvBean.set(fields[0], fields[1], fields[2], fields[3],fields[4], step, fields[6], fields[7], fields[8], fields[9]);
			k.set(pvBean.getSession());
			context.write(k, pvBean);

		}

	}

	static class ClickStreamVisitReducer extends Reducer<Text, PageViewsBean, NullWritable, VisitBean> {

		@Override
		protected void reduce(Text session, Iterable<PageViewsBean> pvBeans, Context context) throws IOException, InterruptedException {

			// 将pvBeans按照step排序
			ArrayList<PageViewsBean> pvBeansList = new ArrayList<PageViewsBean>();
			for (PageViewsBean pvBean : pvBeans) {
				PageViewsBean bean = new PageViewsBean();
				try {
					BeanUtils.copyProperties(bean, pvBean);
					pvBeansList.add(bean);
				} catch (Exception e) {
					e.printStackTrace();
				}
			}

			Collections.sort(pvBeansList, new Comparator<PageViewsBean>() {
				@Override
				public int compare(PageViewsBean o1, PageViewsBean o2) {
					return o1.getStep() > o2.getStep() ? 1 : -1;
				}
			});

			// 取这次visit的首尾pageview记录，将数据放入VisitBean中
			VisitBean visitBean = new VisitBean();
			// 取visit的首记录
			visitBean.setInPage(pvBeansList.get(0).getRequest());
			visitBean.setInTime(pvBeansList.get(0).getTimestr());
			// 取visit的尾记录
			visitBean.setOutPage(pvBeansList.get(pvBeansList.size() - 1).getRequest());
			visitBean.setOutTime(pvBeansList.get(pvBeansList.size() - 1).getTimestr());
			// visit访问的页面数
			visitBean.setPageVisits(pvBeansList.size());
			// 来访者的ip
			visitBean.setRemote_addr(pvBeansList.get(0).getRemote_addr());
			// 本次visit的referal
			visitBean.setReferal(pvBeansList.get(0).getReferal());
			visitBean.setSession(session.toString());
			context.write(NullWritable.get(), visitBean);
		}
	}

	public static void main(String[] args) throws Exception {
		ToolRunner.run(new Configuration(),new ClickStreamVisit(),args);
	}

}

得到的数据