1、主要目的
- 过滤“不合规”数据
- 格式转换和规整
- 根据后续的统计需求，过滤分离出各种不同主题(不同栏目path)的基础数据
2、实现方式
开发一个mr程序WeblogPreProcess。
package com.learn.bigdata.hive.mr.pre;
import java.io.IOException;
import java.util.HashSet;
import java.util.Set;
import com.learn.bigdata.hive.mrbean.WebLogBean;
import com.learn.bigdata.hive.mrbean.WebLogParser;
import org.apache.hadoop.conf.Configuration;
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.output.FileOutputFormat;
/**
* 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 {
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<>();
Text k = new Text();
NullWritable v = NullWritable.get();
/**
* 从外部加载网站url分类数据
*/
@Override
protected void setup(Context context) {
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);
// 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 conf = new Configuration();
Job job = Job.getInstance(conf);
job.setJarByClass(WeblogPreProcess.class);
job.setMapperClass(WeblogPreProcessMapper.class);
job.setOutputKeyClass(Text.class);
job.setOutputValueClass(NullWritable.class);
FileInputFormat.setInputPaths(job, new Path(args[0]));
FileOutputFormat.setOutputPath(job, new Path(args[1]));
job.setNumReduceTasks(0);
job.waitForCompletion(true);
}
}
package com.learn.bigdata.hive.mrbean;
import java.text.ParseException;
import java.text.SimpleDateFormat;
import java.util.Locale;
import java.util.Set;
public class WebLogParser {
public static SimpleDateFormat df1 = new SimpleDateFormat("dd/MMM/yyyy:HH:mm:ss", Locale.US);
public static SimpleDateFormat df2 = new SimpleDateFormat("yyyy-MM-dd HH:mm:ss", Locale.US);
public static WebLogBean parser(String line) {
WebLogBean webLogBean = new WebLogBean();
String[] arr = line.split(" ");
if (arr.length > 11) {
webLogBean.setRemote_addr(arr[0]);
webLogBean.setRemote_user(arr[1]);
String time_local = formatDate(arr[3].substring(1));
if(null==time_local) time_local="-invalid_time-";
webLogBean.setTime_local(time_local);
webLogBean.setRequest(arr[6]);
webLogBean.setStatus(arr[8]);
webLogBean.setBody_bytes_sent(arr[9]);
webLogBean.setHttp_referer(arr[10]);
//如果useragent元素较多，拼接useragent
if (arr.length > 12) {
StringBuilder sb = new StringBuilder();
for(int i=11;i<arr.length;i++){
sb.append(arr[i]);
}
webLogBean.setHttp_user_agent(sb.toString());
} else {
webLogBean.setHttp_user_agent(arr[11]);
}
if (Integer.parseInt(webLogBean.getStatus()) >= 400) {// 大于400，HTTP错误
webLogBean.setValid(false);
}
if("-invalid_time-".equals(webLogBean.getTime_local())){
webLogBean.setValid(false);
}
} else {
webLogBean.setValid(false);
}
return webLogBean;
}
public static void filtStaticResource(WebLogBean bean, Set<String> pages) {
if (!pages.contains(bean.getRequest())) {
bean.setValid(false);
}
}
public static String formatDate(String time_local) {
try {
return df2.format(df1.parse(time_local));
} catch (ParseException e) {
return null;
}
}
}
package com.learn.bigdata.hive.mrbean;
import java.io.DataInput;
import java.io.DataOutput;
import java.io.IOException;
import org.apache.hadoop.io.Writable;
/**
* 对接外部数据的层，表结构定义最好跟外部数据源保持一致
* 术语： 贴源表
*
*/
public class WebLogBean implements Writable {
private boolean valid = true;// 判断数据是否合法
private String remote_addr;// 记录客户端的ip地址
private String remote_user;// 记录客户端用户名称,忽略属性"-"
private String time_local;// 记录访问时间与时区
private String request;// 记录请求的url与http协议
private String status;// 记录请求状态；成功是200
private String body_bytes_sent;// 记录发送给客户端文件主体内容大小
private String http_referer;// 用来记录从那个页面链接访问过来的
private String http_user_agent;// 记录客户浏览器的相关信息
public void set(boolean valid,String remote_addr, String remote_user, String time_local, String request, String status, String body_bytes_sent, String http_referer, String http_user_agent) {
this.valid = valid;
this.remote_addr = remote_addr;
this.remote_user = remote_user;
this.time_local = time_local;
this.request = request;
this.status = status;
this.body_bytes_sent = body_bytes_sent;
this.http_referer = http_referer;
this.http_user_agent = http_user_agent;
}
@Override
public String toString() {
StringBuilder sb = new StringBuilder();
sb.append(this.valid);
sb.append("\001").append(this.getRemote_addr());
sb.append("\001").append(this.getRemote_user());
sb.append("\001").append(this.getTime_local());
sb.append("\001").append(this.getRequest());
sb.append("\001").append(this.getStatus());
sb.append("\001").append(this.getBody_bytes_sent());
sb.append("\001").append(this.getHttp_referer());
sb.append("\001").append(this.getHttp_user_agent());
return sb.toString();
}
@Override
public void readFields(DataInput in) throws IOException {
this.valid = in.readBoolean();
this.remote_addr = in.readUTF();
this.remote_user = in.readUTF();
this.time_local = in.readUTF();
this.request = in.readUTF();
this.status = in.readUTF();
this.body_bytes_sent = in.readUTF();
this.http_referer = in.readUTF();
this.http_user_agent = in.readUTF();
}
@Override
public void write(DataOutput out) throws IOException {
out.writeBoolean(this.valid);
out.writeUTF(null==remote_addr?"":remote_addr);
out.writeUTF(null==remote_user?"":remote_user);
out.writeUTF(null==time_local?"":time_local);
out.writeUTF(null==request?"":request);
out.writeUTF(null==status?"":status);
out.writeUTF(null==body_bytes_sent?"":body_bytes_sent);
out.writeUTF(null==http_referer?"":http_referer);
out.writeUTF(null==http_user_agent?"":http_user_agent);
}
}
运行mr对数据进行预处理
hadoop jar weblog.jar com.learn.bigdata.hive.mr.pre.WeblogPreProcess /weblog/input /weblog/preout3、点击流模型数据梳理
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。
3.1、点击流模型pageviews表
package com.learn.bigdata.hive.mr;
import java.io.IOException;
import java.text.ParseException;
import java.text.SimpleDateFormat;
import java.util.ArrayList;
import java.util.Collections;
import java.util.Comparator;
import java.util.Date;
import java.util.Locale;
import java.util.UUID;
import com.learn.bigdata.hive.mrbean.WebLogBean;
import org.apache.commons.beanutils.BeanUtils;
import org.apache.hadoop.conf.Configuration;
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.output.FileOutputFormat;
/**
*
* 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
*/
public class ClickStreamThree {
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()) {
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) {
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
*/
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);
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){
return time1.getTime() - time2.getTime();
}
}
public static void main(String[] args) throws Exception {
Configuration conf = new Configuration();
Job job = Job.getInstance(conf);
job.setJarByClass(ClickStreamThree.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);
FileInputFormat.setInputPaths(job, new Path(args[0]));
FileOutputFormat.setOutputPath(job, new Path(args[1]));
job.waitForCompletion(true);
}
}Pageviews表模型数据生成
hadoop jar weblogpreprocess.jar \
com.learn.bigdata.hive.mr.ClickStreamThree \
/user/hive/warehouse/dw_click.db/test_ods_weblog_origin/datestr=2013-09-20/ /test-click/pageviews/表结构：
3.2、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。
用MR程序从pageviews数据中，梳理出每一次visit的起止时间、页面信息。
package com.learn.bigdata.hive.mr;
import java.io.IOException;
import java.util.ArrayList;
import java.util.Collections;
import java.util.Comparator;
import com.learn.bigdata.hive.mrbean.PageViewsBean;
import com.learn.bigdata.hive.mrbean.VisitBean;
import org.apache.commons.beanutils.BeanUtils;
import org.apache.hadoop.conf.Configuration;
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.output.FileOutputFormat;
/**
* 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 ......
*/
public class ClickStreamVisit {
// 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 {
Configuration conf = new Configuration();
Job job = Job.getInstance(conf);
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);
FileInputFormat.setInputPaths(job, new Path(args[0]));
FileOutputFormat.setOutputPath(job, new Path(args[1]));
boolean res = job.waitForCompletion(true);
System.exit(res?0:1);
}
}
hadoop jar weblogpreprocess.jar com.learn.bigdata.hive.mr.ClickStreamVisit /weblog/sessionout /weblog/visitout然后，在hive仓库中建点击流visit模型表
drop table if exist click_stream_visit;
create table click_stream_visit(
session string,
remote_addr string,
inTime string,
outTime string,
inPage string,
outPage string,
referal string,
pageVisits int)
partitioned by (datestr string);然后，将MR运算得到的visit数据导入visit模型表
load data inpath '/weblog/visitout' into table click_stream_visit partition(datestr='2013-09-18');