第10章 Hive实战之谷粒影音
10.1 需求描述
统计硅谷影音视频网站的常规指标,各种TopN指标:
- 统计视频观看数Top10
- 统计视频类别热度Top10
- 统计出视频观看数最高的20个视频的所属视频类别以及对应视频类别的个数
- 统计视频观看数Top50所关联视频的所属类别Rank
- 统计每个类别中的视频热度Top10
- 统计每个类别中视频流量Top10
- 统计上传视频最多的用户Top10以及他们上传的视频
- 统计每个类别视频观看数Top10
10.2 项目
10.2.1 数据结构
1、视频表
字段 | 备注 | 详细描述 |
---|---|---|
video id | 视频唯一id | 11位字符串 |
uploader | 视频上传者 | 上传视频的用户名String |
age | 视频年龄 | 视频在平台上的整数天 |
category | 视频类别 | 上传视频指定的视频分类 |
length | 视频长度 | 整形数字标识的视频长度 |
views | 观看次数 | 视频被浏览的次数 |
rate | 视频评分 | 满分5分 |
ratings | 流量 | 视频的流量,整型数字 |
conments | 评论数 | 一个视频的整数评论数 |
related ids | 相关视频id | 相关视频的id,最多20个 |
2、用户表
字段 | 备注 | 字段类型 |
---|---|---|
uploader | 上传者用户名 | string |
videos | 上传视频数 | int |
friends | 朋友数量 | int |
10.2.2 ETL原始数据
通过观察原始数据形式,可以发现,视频可以有多个所属分类,每个所属分类用&符号分割,且分割的两边有空格字符,同时相关视频也是可以有多个元素,多个相关视频又用“\t”进行分割。为了分析数据时方便对存在多个子元素的数据进行操作,我们首先进行数据重组清洗操作
。即:将所有的类别用“&”分割,同时去掉两边空格,多个相关视频id也使用“&”进行分割。
0、添加依赖pom.xml
<dependencies>
<dependency>
<groupId>junit</groupId>
<artifactId>junit</artifactId>
<version>RELEASE</version>
</dependency>
<dependency>
<groupId>org.apache.logging.log4j</groupId>
<artifactId>log4j-core</artifactId>
<version>2.8.2</version>
</dependency>
<dependency>
<groupId>org.apache.hadoop</groupId>
<artifactId>hadoop-common</artifactId>
<version>2.7.2</version>
</dependency>
<dependency>
<groupId>org.apache.hadoop</groupId>
<artifactId>hadoop-client</artifactId>
<version>2.7.2</version>
</dependency>
<dependency>
<groupId>org.apache.hadoop</groupId>
<artifactId>hadoop-hdfs</artifactId>
<version>2.7.2</version>
</dependency>
</dependencies>
1、ETL之ETLUtil
package com.atguigu;
import org.apache.hadoop.yarn.webapp.hamlet.Hamlet;
/**
* @author chenmingjun
* @date 2019-03-01 15:48
*/
public class ETLUtil {
public static String oriString2ETLString(String ori) {
// 0.切割数据
String[] fields = ori.split("\t");
// 1.过滤脏数据(不符合要求的数据)
if (fields.length < 9) {
return null;
}
// 2.将类别字段中的" " 替换为""(即去掉类别字段中的空格)
fields[3] = fields[3].replace(" ", "");
// 3.替换关联视频字段分隔符"\t"替换为"&"
StringBuffer sb = new StringBuffer();
for (int i = 0; i < fields.length; i++) {
// 关联视频字段之间的数据
if (i < 9) {
if (i == fields.length -1) {
sb.append(fields[i]);
} else {
sb.append(fields[i] + "\t");
}
} else {
// 关联视频字段的数据
if (i == fields.length -1) {
sb.append(fields[i]);
} else {
sb.append(fields[i] + "&");
}
}
}
// 得到的数据格式为:bqZauhidT1w bungloid 592 Film&Animation 28 374550 4.19 3588 1763 QJ5mXzC1YbQ&geEBYTZ4EB8
return sb.toString();
}
}
2、ETL之Mapper
package com.atguigu;
import org.apache.commons.lang.StringUtils;
import org.apache.hadoop.io.LongWritable;
import org.apache.hadoop.io.NullWritable;
import org.apache.hadoop.io.Text;
import org.apache.hadoop.mapreduce.Mapper;
import java.io.IOException;
/**
* @author chenmingjun
* @date 2019-02-28 23:32
*/
public class VideoETLMapper extends Mapper<LongWritable, Text, Text, NullWritable> {
private Text k = new Text();
@Override
protected void map(LongWritable key, Text value, Context context) throws IOException, InterruptedException {
// 1.获取一行数据
String ori = value.toString();
// 2.清洗数据
String etlString = ETLUtil.oriString2ETLString(ori);
// 3.写出
if (StringUtils.isBlank(etlString)) {
return;
}
k.set(etlString);
context.write(k, NullWritable.get().get());
}
}
3、ETL之Runner
package com.atguigu;
import org.apache.hadoop.conf.Configuration;
import org.apache.hadoop.fs.Path;
import org.apache.hadoop.io.NullWritable;
import org.apache.hadoop.io.Text;
import org.apache.hadoop.mapreduce.Job;
import org.apache.hadoop.mapreduce.lib.input.FileInputFormat;
import org.apache.hadoop.mapreduce.lib.output.FileOutputFormat;
import org.apache.hadoop.util.Tool;
import org.apache.hadoop.util.ToolRunner;
/**
* @author chenmingjun
* @date 2019-03-01 16:55
*/
public class VideoETLRunner implements Tool {
private Configuration conf = null;
public void setConf(Configuration conf) {
this.conf = conf;
}
public Configuration getConf() {
return this.conf;
}
public int run(String[] args) throws Exception {
// 1、获取配置信息对象以及封装任务
// Configuration conf = new Configuration();
Job job = Job.getInstance(getConf());
// 2、设置jar的加载路径
job.setJarByClass(VideoETLRunner.class);
// 3、设置map和reduce类
job.setMapperClass(VideoETLMapper.class);
// job.setReducerClass(WordcountReducer.class);
// 4、设置map输出的key和value类型
job.setMapOutputKeyClass(Text.class);
job.setMapOutputValueClass(NullWritable.class);
// 5、设置最终输出的key和value类型
job.setOutputKeyClass(Text.class);
job.setOutputValueClass(NullWritable.class);
// 6、设置输入和输出路径
FileInputFormat.setInputPaths(job, new Path(args[0]));
FileOutputFormat.setOutputPath(job, new Path(args[1]));
// 因为这里我们不使用Reduce
job.setNumReduceTasks(0);
// 7、提交job
// job.submit();
boolean result = job.waitForCompletion(true);
return result ? 0 : 1;
}
public static void main(String[] args) {
int resultCode = 0;
try {
resultCode = ToolRunner.run(new VideoETLRunner(), args);
if (resultCode == 0) {
System.out.println("Success!");
} else {
System.out.println("Fail!");
}
System.exit(resultCode);
} catch (Exception e) {
e.printStackTrace();
System.exit(1);
}
}
}
4、打好jar包,修改jar包名称为VideoETL.jar,然后将要清洗的数据和VideoETL.jar从本地上传至Linux系统上,再将要清洗的数据推送至HDFS集群上。操作如下:
[atguigu@hadoop102 datas]$ hadoop fs -put user/ /guliData/input
[atguigu@hadoop102 datas]$ hadoop fs -put video/ /guliData/input
5、执行ETL
[atguigu@hadoop102 hadoop-2.7.2]$ bin/yarn jar /opt/module/datas/VideoETL.jar com.atguigu.VideoETLRunner /guliData/input/video/2008/0222 /guliData/output/video/2008/0222
10.3 准备工作
10.3.1 创建表
创建原始表:gulivideo_ori,gulivideo_user_ori
创建目标表:gulivideo_orc,gulivideo_user_orc
gulivideo_ori:
create table gulivideo_ori(
videoId string,
uploader string,
age int,
category array<string>,
length int,
views int,
rate float,
ratings int,
comments int,
relatedId array<string>
)
row format delimited
fields terminated by "\t" -- 字段与字段之间的数据按/t分割
collection items terminated by "&" -- 数组中的数据是按&分割
stored as textfile;
gulivideo_user_ori:
create table gulivideo_user_ori(
uploader string,
videos int,
friends int
)
row format delimited
fields terminated by "\t"
stored as textfile;
gulivideo_orc:
create table gulivideo_orc(
videoId string,
uploader string,
age int,
category array<string>,
length int,
views int,
rate float,
ratings int,
comments int,
relatedId array<string>
)
clustered by(uploader) into 8 buckets -- 按照字段uploader分成8个桶
row format delimited
fields terminated by "\t"
collection items terminated by "&"
stored as orc;
gulivideo_user_orc:
create table gulivideo_user_orc(
uploader string,
videos int,
friends int
)
row format delimited
fields terminated by "\t"
stored as orc;
10.3.2 导入ETL后的数据到原始表
gulivideo_ori:
load data inpath '/guliData/output/video/2008/0222' into table gulivideo_ori;
gulivideo_user_ori:
load data inpath "/guliData/input/user/2008/0903" into table gulivideo_user_ori;
10.3.3 向ORC表插入数据
gulivideo_orc:
insert into table gulivideo_orc select * from gulivideo_ori;
gulivideo_user_orc:
insert into table gulivideo_user_orc select * from gulivideo_user_ori;
10.4 业务分析
10.4.1 统计视频观看数Top10
思路:使用order by
按照 views 字段做一个全局排序
即可,同时我们设置只显示前10条。为了便于显示,我们显示的字段不包含每个视频对应的关联视频字段
。
最终代码:
select
videoId,
uploader,
age,
category,
length,
views,
rate,
ratings,
comments
from
gulivideo_orc
order by
views desc
limit
10;
10.4.2 统计视频类别热度Top10
思路:炸开数组【视频类别】字段,然后按照类别分组,最后按照热度(视频个数)排序。
1) 即统计每个类别有多少个视频,显示出包含视频最多的前10个类别。
2) 我们需要按照类别 group by 聚合,然后count组内的videoId个数即可。
3) 因为当前表结构为:一个视频对应一个或多个类别。所以如果要 group by 类别,需要先将类别进行列转行(展开),然后再进行count即可。
4) 最后按照热度排序,显示前10条。
最终代码:
第1步:炸裂视频类别
select
videoId, category_name
from
gulivideo_orc lateral view explode(category) category_t as category_name
limit
100; t1
------------------------------------------------------------------------------------
第2步:统计每种视频类别下的视频数
select
category_name, count(*) hot
from
(select
videoId, category_name
from
gulivideo_orc lateral view explode(category) category_t as category_name
limit
100) t1
group by
category_name; t2
------------------------------------------------------------------------------------
第3步:视频类别热度Top10
select
category_name, hot
from
(select
category_name, count(*) hot
from
(select
videoId, category_name
from
gulivideo_orc lateral view explode(category) category_t as category_name) t1
group by
category_name) t2
order by
hot desc
limit
10;
+----------------+---------+--+
| category_name | hot |
+----------------+---------+--+
| Music | 179049 |
| Entertainment | 127674 |
| Comedy | 87818 |
| Animation | 73293 |
| Film | 73293 |
| Sports | 67329 |
| Gadgets | 59817 |
| Games | 59817 |
| Blogs | 48890 |
| People | 48890 |
+----------------+---------+--+
注意
:第1步和第2步测试先使用100条数据,测试通过后第3步使用全部数据。
10.4.3 统计出视频观看数最高的20个视频的所属视频类别以及对应视频类别的个数
思路:
1) 先找到观看数最高的20个视频所属条目的所有信息,降序排列
2) 把这20条信息中的category分裂出来(列转行)
3) 最后查询视频分类名称和该分类下有多少个Top20的视频
最终代码:
统计出视频观看数最高的20个视频的所属类别
第1步:统计出视频观看数最高的20个视频
select
*
from
gulivideo_orc
order by
views desc
limit
20; t1
------------------------------------------------------------------------------------
第2步:把这20条信息中的category分裂出来(列转行)
select
videoId,
category_name
from
(select
*
from
gulivideo_orc
order by
views desc
limit
20) t1 lateral view explode(category) category_t as category_name; t2
+--------------+----------------+--+
| videoid | category_name |
+--------------+----------------+--+
| dMH0bHeiRNg | Comedy |
| 0XxI-hvPRRA | Comedy |
| 1dmVU08zVpA | Entertainment |
| RB-wUgnyGv0 | Entertainment |
| QjA5faZF1A8 | Music |
| -_CSo1gOd48 | People |
| -_CSo1gOd48 | Blogs |
| 49IDp76kjPw | Comedy |
| tYnn51C3X_w | Music |
| pv5zWaTEVkI | Music |
| D2kJZOfq7zk | People |
| D2kJZOfq7zk | Blogs |
| vr3x_RRJdd4 | Entertainment |
| lsO6D1rwrKc | Entertainment |
| 5P6UU6m3cqk | Comedy |
| 8bbTtPL1jRs | Music |
| _BuRwH59oAo | Comedy |
| aRNzWyD7C9o | UNA |
| UMf40daefsI | Music |
| ixsZy2425eY | Entertainment |
| MNxwAU_xAMk | Comedy |
| RUCZJVJ_M8o | Entertainment |
+--------------+----------------+--+
------------------------------------------------------------------------------------
第3步:根据视频分类名称进行去重
select
distinct category_name
from
t2;
-------------------------------------------
完整板
select
distinct category_name
from
(select
videoId,
category_name
from
(select
*
from
gulivideo_orc
order by
views desc
limit
20) t1 lateral view explode(category) category_t as category_name) t2;
-------------------------------------------
简易版
select
distinct category_name
from
(select
*
from
gulivideo_orc
order by
views desc
limit
20) t1 lateral view explode(category) category_t as category_name;
+----------------+--+
| category_name |
+----------------+--+
| Blogs |
| Comedy |
| Entertainment |
| Music |
| People |
| UNA |
+----------------+--+
------------------------------------------------------------------------------------
类别包含Top20视频的个数
select
category_name,
count(t2.videoId) as hot_with_views
from
(select
videoId,
category_name
from
(select
*
from
gulivideo_orc
order by
views desc
limit
20) t1 lateral view explode(category) category_t as category_name) t2
group by
category_name
order by
hot_with_views desc;
+----------------+-----------------+--+
| category_name | hot_with_views |
+----------------+-----------------+--+
| Entertainment | 6 |
| Comedy | 6 |
| Music | 5 |
| People | 2 |
| Blogs | 2 |
| UNA | 1 |
+----------------+-----------------+--+
10.4.4 统计视频观看数Top50所关联视频的所属类别rank
思路分析如下图所示:
思路:
1) 查询出观看数最多的前50个视频的所有信息(当然
包含了每个视频对应的关联视频
),记为临时表t1t1:观看数前50的视频
select
videoId, views, category, relatedId
from
gulivideo_orc
order by
views desc
limit
50; t1
2) 将找到的50条视频信息的相关视频relatedId列转行,记为临时表t2
t2:将相关视频的id进行列转行操作
炸裂关联视频id
select
explode(relatedId) as videoId
from
t1; t2
或者
select
distinct videoId
from
t1 lateral view explode(relatedId) relatedId_t as videoId; t2
3) 将关联视频的id和gulivideo_orc表进行inner join操作,得到每个关联视频id的详细数据
select
*
from
t2
inner join
gulivideo_orc t3 on t2.videoId=t3.videoId; t4
4) 炸裂关联视频的类别
select
*
from
t4 lateral view explode(category) category_t as category_name; t5
5) 统计类别个数
select
category_name,
count(*) hot
from
t5
group by
category_name; t6
6) 统计类别的热度排名(即rank)
select
*
from
t6
order by
hot desc;
10.4.5 统计每个类别中的视频热度Top10,以Music为例
思路:
1) 要想统计Music类别中的视频热度Top10,需要先找到Music类别,那么就需要将category展开,所以可以创建一张表用于存放categoryId展开的数据。
2) 向category展开的表中插入数据。
3) 统计对应类别(Music)中的视频热度。
最终代码:
创建表--类别表:
create table gulivideo_category(
videoId string,
uploader string,
age int,
categoryId string,
length int,
views int,
rate float,
ratings int,
comments int,
relatedId array<string>
)
row format delimited
fields terminated by "\t"
collection items terminated by "&"
stored as orc;
向类别表中插入数据:
insert into table
gulivideo_category
select
videoId,
uploader,
age,
categoryId,
length,
views,
rate,
ratings,
comments,
relatedId
from
gulivideo_orc lateral view explode(category) catetory_t as categoryId;
统计Music类别的Top10(也可以统计其他)
select
videoId,
views
from
gulivideo_category
where
categoryId="Music"
order by
views desc
limit
10;
10.4.6 统计每个类别中视频流量Top10,以Music为例
思路:
1) 创建视频类别展开表(categoryId列转行后的表)
2) 按照ratings排序即可
最终代码:
select
videoId,
views
from
gulivideo_category
where
categoryId="Music"
order by
ratings desc
limit
10;
10.4.7 统计上传视频最多的用户Top10以及他们上传的观看次数在前20的视频
思路:
1) 先找到上传视频最多的10个用户的用户信息
select
*
from
gulivideo_user_orc
order by
videos desc
limit
10; t1
2) 通过uploader字段与gulivideo_orc表进行join,得到的信息按照views观看次数进行排序即可。
最终代码:
select
t2.videoId,
t2.views,
t2.ratings,
t1.videos,
t1.friends
from
t1
join
gulivideo_orc t2
on
t1.uploader=t2.uploader
order by
t2.views desc
limit
20;
10.4.8 统计每个类别视频观看数Top10
思路:
1) 先得到categoryId展开的表数据。
2) 子查询按照categoryId进行分区,然后分区内排序降序,并生成递增数字,该递增数字这一列起名为rank列。
3) 通过子查询产生的临时表,查询rank值小于等于10的数据行即可。
最终代码:
创建表--类别表:
create table gulivideo_category(
videoId string,
uploader string,
age int,
categoryId string,
length int,
views int,
rate float,
ratings int,
comments int,
relatedId array<string>
)
row format delimited
fields terminated by "\t"
collection items terminated by "&"
stored as orc;
向类别表中插入数据:
insert into table
gulivideo_category
select
videoId,
uploader,
age,
categoryId,
length,
views,
rate,
ratings,
comments,
relatedId
from
gulivideo_orc lateral view explode(category) catetory_t as categoryId;
代码:
第1步:
select
videoId,
categoryId,
views,
row_number() over(partition by categoryId order by views desc) rank
from
gulivideo_category; t1
第2步:
select
t1.*
from
t1
where
rank<=10;
回到顶部
第11章 常见错误及解决方案
1)SecureCRT 7.3 出现乱码或者删除不掉数据,免安装版的 SecureCRT 卸载或者用虚拟机直接操作或者换安装版的SecureCRT。
2)连接不上mysql数据库
(1)导错驱动包,应该把 mysql-connector-java-5.1.27-bin.jar 导入 /opt/module/hive/lib 的不是这个包。错把 mysql-connector-java-5.1.27.tar.gz 导入 hive/lib 包下。
(2)修改user表中的主机名称没有都修改为%,而是修改为 localhost。
3)hive默认的输入格式处理是 CombineHiveInputFormat,会对小文件进行合并。
hive (default)> set hive.input.format;
hive.input.format=org.apache.hadoop.hive.ql.io.CombineHiveInputFormat
可以采用 HiveInputFormat 就会根据分区数输出相应的文件。
hive (default)> set hive.input.format=org.apache.hadoop.hive.ql.io.HiveInputFormat;
4)不能执行mapreduce程序
可能是hadoop的yarn没开启。
5)启动mysql服务时,报 MySQL server PID file could not be found!
异常。
在 /var/lock/subsys/mysql 路径下创建 hadoop102.pid,并在文件中添加内容:4396
6)报 service mysql status MySQL is not running, but lock file (/var/lock/subsys/mysql[失败])异常。
解决方案:在/var/lib/mysql 目录下创建:
-rw-rw----. 1 mysql mysql 5 12月 22 16:41 hadoop102.pid
文件,并修改权限为 777。
7)JVM堆内存溢出
描述:java.lang.OutOfMemoryError: Java heap space
解决:在yarn-site.xml中加入如下代码后,进行分发,重启yarn。
<property>
<name>yarn.scheduler.maximum-allocation-mb</name>
<value>2048</value>
</property>
<property>
<name>yarn.scheduler.minimum-allocation-mb</name>
<value>2048</value>
</property>
<property>
<name>yarn.nodemanager.vmem-pmem-ratio</name>
<value>2.1</value>
</property>
<property>
<name>mapred.child.java.opts</name>
<value>-Xmx1024m</value>
</property>