Flume消费Kafka数据保存到Hive
场景
通过Flume消费Kafka中数据，保存数据到ODS层，数据存储时标记消费时的元信息
创建Hive表
orc存储,snappy压缩,开启事务
ORC事务表
- 只能是内部表
- 必须创建桶
create TABLE hr.ods_internetbar_data
(
k_topic string ,
k_data string,
k_partition int,
k_offset int,
k_key string,
current_time bigint
)
partitioned by (pt_dt int)
CLUSTERED BY(k_partition) into 5 buckets
ROW FORMAT DELIMITED FIELDS TERMINATED BY '\t'
LINES TERMINATED BY '\n'
STORED AS orc
TBLPROPERTIES('orc.compress'='SNAPPY','transactional'='true');
编写Interceptor
思路：我们将Kafka中的元信息(topic名称，partition等)通过Interceptor转换成以'\t'分隔的数据，然后将数据保存到Hive
拦截器com.jast.flume.ETLInterceptor代码如下：
package com.jast.flume;
import cn.hutool.json.JSONUtil;
import org.apache.flume.Context;
import org.apache.flume.Event;
import org.apache.flume.interceptor.Interceptor;
import java.nio.charset.StandardCharsets;
import java.util.Iterator;
import java.util.List;
import java.util.Map;
/**
* @author Jast
* @description 自定义拦截器
* @date 2022-03-17 09:20
*/
public class ETLInterceptor implements Interceptor {
private final String separator = "\t";
private final String defaultValue = "";
@Override
public void initialize() {
}
@Override
public Event intercept(Event event) {
//拦截kafka时，这里获取到的Head，是Kafka元信息,包括key，topic，partition，offset等信息，e.g. {topic=ZW_WB_1010000009, partition=0, offset=5425705, key={"dataType":"ZW_WB_","fileName":"yc_1644913124_11086","fullLinkId":"3e9fc439e42c4059b46e5d5664d08bf8","lineNum":17,"subFullLinkId":"77dc3a9680ee4d1caeb6719d87e8d3e5","totalCount":50}, timestamp=1648642188763}
Map<String, String> headers = event.getHeaders();
String topic = headers.getOrDefault("topic",defaultValue);
String partition = headers.getOrDefault("partition",defaultValue);
String offset = headers.getOrDefault("offset",defaultValue);
String key = headers.getOrDefault("key",defaultValue);
long currentTimeMillis = System.currentTimeMillis();
//System.out.println("拦截器head:"+headers);
byte[] body = event.getBody();
String text = new String(body, StandardCharsets.UTF_8);
if(JSONUtil.isJson(text)){
text = topic+separator+text+separator+partition+separator+offset+separator+key+separator+currentTimeMillis;
System.out.println("text:"+text);
event.setBody(text.getBytes());
return event;
}
//System.out.println("非json格式过滤掉");
return null;
}
@Override
public List<Event> intercept(List<Event> list) {
Iterator<Event> iterator = list.iterator();
while (iterator.hasNext()){
Event next = iterator.next();
if(intercept(next)==null){
iterator.remove();
}
}
return list;
}
@Override
public void close() {
}
public static class Builder implements Interceptor.Builder{
@Override
public Interceptor build() {
return new ETLInterceptor();
}
@Override
public void configure(Context context) {
}
}
}
pom.xml
<?xml version="1.0" encoding="UTF-8"?>
<project xmlns="http://maven.apache.org/POM/4.0.0"
xmlns:xsi="http://www.w3.org/2001/XMLSchema-instance"
xsi:schemaLocation="http://maven.apache.org/POM/4.0.0 http://maven.apache.org/xsd/maven-4.0.0.xsd">
<modelVersion>4.0.0</modelVersion>
<groupId>org.example</groupId>
<artifactId>flume-interceptor</artifactId>
<version>1.0-SNAPSHOT</version>
<dependencies>
<dependency>
<groupId>org.apache.flume</groupId>
<artifactId>flume-ng-core</artifactId>
<version>1.9.0</version>
<scope>provided</scope>
</dependency>
<dependency>
<groupId>cn.hutool</groupId>
<artifactId>hutool-all</artifactId>
<version>5.7.19</version>
</dependency>
</dependencies>
<build>
<plugins>
<plugin>
<artifactId>maven-compiler-plugin</artifactId>
<version>2.3.2</version>
<configuration>
<source>1.8</source>
<target>1.8</target>
</configuration>
</plugin>
<plugin>
<artifactId>maven-assembly-plugin</artifactId>
<configuration>
<descriptorRefs>
<descriptorRef>jar-with-dependencies</descriptorRef>
</descriptorRefs>
</configuration>
<executions>
<execution>
<id>make-assembly</id>
<phase>package</phase>
<goals>
<goal>single</goal>
</goals>
</execution>
</executions>
</plugin>
</plugins>
</build>
</project>
打包后生成flume-interceptor-1.0-SNAPSHOT-jar-with-dependencies.jar文件
将生成的包放到flume lib 目录中
自己部署的Flume：
需要先将打好的包放入到
flume/lib文件夹下面CDH版本Flume：
需要先将打好的包放入到
flume/lib文件夹下面
具体的目录/opt/cloudera/parcels/CDH-6.3.2-1.cdh6.3.2.p0.1605554/lib/flume-ng/lib/
cp flume-interceptor-1.0-SNAPSHOT-jar-with-dependencies.jar /opt/cloudera/parcels/CDH-6.3.2-1.cdh6.3.2.p0.1605554/lib/flume-ng/lib
编写Flume配置文件
需要注意的是a1.sinks.k1.serializer.fieldnames和a1.sources.r1.interceptors.i1.type
a1.sinks.k1.serializer.fieldnames:拦截器\t处理后在这里与Hive表的字段对应上
a1.sources.r1.interceptors.i1.type:拦截器配置
配置文件k2h.config内容如下
## 组件
a1.sources=r1
a1.channels=c1
a1.sinks=k1
## source1
a1.sources.r1.type = org.apache.flume.source.kafka.KafkaSource
# 每批次数量
a1.sources.r1.batchSize = 1000
# 将批写入通道之前的最长时间（毫秒）。每当达到第一个大小和时间时，将写入批。
a1.sources.r1.batchDurationMillis = 2000
a1.sources.r1.kafka.bootstrap.servers = 192.168.60.16:9092
# kafka消费主题，多个主题逗号分隔
a1.sources.r1.kafka.topics=ZW_WB_1010000009
# 默认:PLAINTEXT，如果使用某种安全级别写入 Kafka，则设置为 SASL_PLAINTEXT、SASL_SSL 或 SSL。
a1.sources.r1.kafka.consumer.security.protocol=PLAINTEXT
# 配置拦截器
a1.sources.r1.interceptors = i1
a1.sources.r1.interceptors.i1.type = com.jast.flume.ETLInterceptor$Builder
## channel1
# 基于文件存储的Channel
a1.channels.c1.type = file
# checkpoint 保存目录
a1.channels.c1.checkpointDir = /var/lib/hadoop-hdfs/flume/checkpoint/behavior1
# 用于存储日志文件的目录的逗号分隔列表。 在不同磁盘上使用多个目录可以提高文件通道性能
a1.channels.c1.dataDirs = /var/lib/hadoop-hdfs/flume/data/behavior1/
# 最大事务大小
a1.channels.c1.transactionCapacity=10000
# 通道最大容量
a1.channels.c1.capacity=1000000
a1.sinks.k1.type = hive
a1.sinks.k1.channel = c1
a1.sinks.k1.hive.metastore = thrift://192.168.60.14:9083
# hive 库名
a1.sinks.k1.hive.database = hr
# hive表名称
a1.sinks.k1.hive.table = ods_internetbar_data
# 分区使用字段，转义字符介绍在文章下面
a1.sinks.k1.hive.partition = %y%m%d
# 在替换转义序列时使用本地时间（而不是事件标头中的时间戳）。
a1.sinks.k1.useLocalTimeStamp = false
# ============================================================
# e.g. 每6小时产生一个新文件，比如把24小时分成4份，假设现在的时间是15:40,如果这时候有新的日志到来，那么hdfs 会创建一个新的hdfs文件，文件名称是2015102012 ，就是15:40 是分布在12-18这个区间的
a1.sinks.k1.round = true
a1.sinks.k1.roundValue = 10
a1.sinks.k1.roundUnit = minute
# collector1.sinks.sink_hdfs.hdfs.rollSize = 2048000000
# collector1.sinks.sink_hdfs.hdfs.rollCount = 0
# collector1.sinks.sink_hdfs.hdfs.rollInterval = 21600
# rollsize 的配置表示到2G大小的时候回滚到下一个文件，也就是到了这个时间 hdfs就会rename正在写的文件到已经写完
# rollInterval 的配置表示每个6小时回滚到下一个文件
# ============================================================
# 数据解析方式，目前支持DELIMITED(分隔符)，JSON(每行为单层Json)和REGEX(正则表达式)
a1.sinks.k1.serializer = DELIMITED
# 数据字段分割符，如果要使用特殊字符需要添加双引号，例如"\t"
a1.sinks.k1.serializer.delimiter = "\t"
# 数据解析的正则表达式，每个字段的数据被解析成一个group
# a1.sinks.k1.serializer.regex =
# （类型：字符）自定义底层 serde 使用的分隔符。 如果 serializer.fieldnames 中的字段与表列的顺序相同，serializer.delimiter 与 serializer.serdeSeparator 相同，并且 serializer.fieldnames 中的字段数小于或等于表数，则效率会有所提高 列，因为传入事件正文中的字段不需要重新排序以匹配表列的顺序。 对特殊字符使用单引号，例如“\t”。 确保输入字段不包含此字符。 注意：如果 serializer.delimiter 是单个字符，最好将其设置为相同的字符
a1.sinks.k1.serializer.serdeSeparator = '\t'
# 输入数据字段到datahub字段的映射，以输入的顺序标示字段，如果要跳过某个字段， 不指定列名即可，例如 c1,c2,,c3，表示将输入数据的第一、二、四字段和hive的c1,c2,c3字段进行匹配。
a1.sinks.k1.serializer.fieldnames = k_topic,k_data,k_partition,k_offset,k_key,current_time
# 单个配置单元事务中写入配置单元的最大事件数，默认15000
a1.sinks.k1.batchSize=10000
## 拼装
a1.sources.r1.channels = c1
a1.sinks.k1.channel= c1
上面配置中我们使用到了转义序列a1.sinks.k1.hive.partition = %y%m%d
支持的转义序列：
| Alias | Description |
|---|---|
| %{host} | Substitute value of event header named “host”. Arbitrary header names are supported. |
| %t | Unix time in milliseconds |
| %a | locale’s short weekday name (Mon, Tue, ...) |
| %A | locale’s full weekday name (Monday, Tuesday, ...) |
| %b | locale’s short month name (Jan, Feb, ...) |
| %B | locale’s long month name (January, February, ...) |
| %c | locale’s date and time (Thu Mar 3 23:05:25 2005) |
| %d | day of month (01) |
| %D | date; same as %m/%d/%y |
| %H | hour (00..23) |
| %I | hour (01..12) |
| %j | day of year (001..366) |
| %k | hour ( 0..23) |
| %m | month (01..12) |
| %M | minute (00..59) |
| %p | locale’s equivalent of am or pm |
| %s | seconds since 1970-01-01 00:00:00 UTC |
| %S | second (00..59) |
| %y | last two digits of year (00..99) |
| %Y | year (2010) |
| %z | +hhmm numeric timezone (for example, -0400) |
启动Flume收集数据
flume-ng agent --conf-file k2h.config --name a1
查看数据
select k_topic,k_partition,k_offset,k_key,current_time,length(k_data) from ods_internetbar_data limit 10;
----
ZW_WB_1010000009 0 5441712 {
"dataType":"ZW_WB_","fileName":"yc_1644913124_11086","fullLinkId":"f3bc635822ca4af099755b9b339e8e4c","lineNum":24,"subFullLinkId":"c2e3704534e745c0a779b16f059e6411","totalCount":50} 1648644001736 52105
ZW_WB_1010000009 0 5441713 {
"dataType":"ZW_WB_","fileName":"yc_1644913124_11086","fullLinkId":"f3bc635822ca4af099755b9b339e8e4c","lineNum":25,"subFullLinkId":"c0cec18006ce40848ea3a0494a991999","totalCount":50} 1648644001736 35837
这里使用
length(k_data)是为了方便展示，该字段值太大了
