实践数据湖iceberg 第七课 实时写入到iceberg

系列文章目录

实践数据湖iceberg 第一课.
实践数据湖iceberg 第二课 iceberg基于hadoop的底层数据格式.
实践数据湖iceberg 第三课 在sqlclient中，以sql方式从kafka读数据到iceberg.
实践数据湖iceberg 第四课 在sqlclient中，以sql方式从kafka读数据到iceberg（升级版本到flink1.12.7）.
实践数据湖iceberg 第五课 hive catalog特点.
实践数据湖iceberg 第六课 从kafka写入到iceberg失败问题 解决.
实践数据湖iceberg 第七课 实时写入到iceberg.

---

提示：写完文章后，目录可以自动生成，如何生成可参考右边的帮助文档

---

前言

制造数据，每秒生成1条数据，数据有序增长，把这些数据实时写入kafka, kaka写到iceberg

---

提示：以下是本篇文章正文内容，下面案例可供参考

1. 制造数据

准备生成数据脚本

[root@hadoop102 bin]# cat generateLog.sh
#!/bin/bash

for i in {1..10000000};do echo $i,$RANDOM >>  /opt/module/logs/mockLog.log; sleep 1s ;done;

生成数据

[root@hadoop102 bin]# nohup ./generateLog.sh > generateLog.log 2>&1 & 
[1] 20688

查看生成的数据

[root@hadoop102 logs]# tail -f mockLog.log 
2,98562
3,4401
4,69527
5,27391
6,64876
7,32809
8,47124
9,65215
10,55296
11,91302
12,37345
13,56344
14,5092

制造过程中，出错，日志入下图

[root@hadoop102 bin]# cat generateLog.log
nohup: ignoring input
./generateLog.sh: fork: Cannot allocate memory

是因为服务器，使用默认最大进程数，造成，修改一下就OK, 百度一下fork: Cannot allocate memory 就有很多解决方案

后来发现不是进程数的问题。
可能是for的数字太大
我还是改为java写个jar包上来

package org.example;

import java.util.Random;

/**
 * Hello world!
 *
 */
public class GenerateLog
{
    public static void main( String[] args ) throws InterruptedException {
        int len = 100000;
        int sleepMilis = 1000;
//        System.out.println("<生成id范围><每条数据停顿时长毫秒>");
        if(args.length == 1){
            len = Integer.valueOf(args[0]);
        }
        if(args.length == 2){
            len = Integer.valueOf(args[0]);
            sleepMilis = Integer.valueOf(args[1]);
        }
        Random random =  new Random();
        for(int i=0; i<10000000; i++){
            System.out.println(i+"," + random.nextInt(len) );
            Thread.sleep(sleepMilis);
        }
    }
}

java版升成数据脚本

[root@hadoop102 bin]# cat generateLog.sh 
#!/bin/bash

#for i in {1..100000};do echo $i,$RANDOM >>  /opt/module/logs/mockLog.log; sleep 1s ;done;
nohup java -jar /opt/software/iceberg-learning-1.0-SNAPSHOT.jar >  /opt/module/logs/mockLog.log 2>&1 &

2. 通过flume把制造的数据，写入到kafka

一开始，想通过linux管道写到kafka,发现 管道

2.1 flume把日志直接写到kafka的配置文件

[root@hadoop102 conf]# cat file-to-kafka.conf 
a1.sources = r1
a1.channels = c1
a1.sources.r1.type = TAILDIR

a1.sources.r1.positionFile = /var/log/flume/taildir_position.json
a1.sources.r1.filegroups = f1
a1.sources.r1.filegroups.f1 = /opt/module/logs/mockLog.log

a1.channels.c1.type = org.apache.flume.channel.kafka.KafkaChannel
a1.channels.c1.kafka.bootstrap.servers = hadoop101:9092,hadoop102:9092,hadoop103:9092
a1.channels.c1.kafka.topic = behavior_log
a1.channels.c1.kafka.consumer.group.id = flume-consumer

#绑定source和channel以及sink和channel的关系
a1.sources.r1.channels = c1

2.2 准备启动flume agent的脚本

[root@hadoop102 bin]# cat flume.sh 
#! /bin/bash

case $1 in
"start"){
        for i in hadoop102
        do
                echo " --------启动 $i flume-------"
                ssh $i "source /etc/profile;cd /opt/module/flume; nohup  flume-ng agent --conf conf --conf-file conf/file-to-kafka.conf  --name a1 -Dflume.root.logger=INFO,console  >/opt/module/flume/logs/file-to-kafka.log   2>&1 &"
        done
};;
"stop"){
        for i in hadoop102
        do
                echo " --------停止 $i flume file-to-kafka.conf -------"
                ssh $i "ps -ef | grep file-to-kafka.conf  | grep -v grep |awk '{print \$2}' | xargs -n1 kill"
        done

};;
esac
[root@hadoop102 bin]# 

2.3 检查是否写入到kafka

[root@hadoop103 ~]# kafka-console-consumer.sh --topic behavior_log --bootstrap-server hadoop101:9092,hadoop102:9092,hadoop103:9092 
413,97153
416,77212
414,27686
412,33012
415,84019
419,28548
418,39725
421,38533
417,39665
420,9306

3.把kakfa的行为数据入湖

3.1.定义kafka表

启动flink-sql

 sql-client.sh embedded -j /opt/software/iceberg-flink-runtime-0.12.1.jar  -j /opt/software/flink-sql-connector-hive-2.3.6_2.12-1.12.7.jar  -j /opt/software/flink-sql-connector-kafka_2.12-1.12.7.jar  shell 

3.1.1 使用csv格式读取看看

create table kafka_behavior_log
(
  i BIGINT,
  id BIGINT
) WITH (
  'connector' = 'kafka',
  'topic' = 'behavior_log',
  'properties.bootstrap.servers' = 'hadoop101:9092,hadoop102:9092,hadoop103:9092',
  'properties.group.id' = 'rickGroup7',
  'scan.startup.mode' = 'earliest-offset',
  'format' = 'csv'
)
select * from kafka_behavior_log；
读取不到实时数据

3.1.2 使用raw格式读取看看

create table kafka_behavior_log_raw
(
log STRING
) WITH (
  'connector' = 'kafka',
  'topic' = 'behavior_log',
  'properties.bootstrap.servers' = 'hadoop101:9092,hadoop102:9092,hadoop103:9092',
  'properties.group.id' = 'rickGroup7',
  'scan.startup.mode' = 'earliest-offset',
  'format' = 'raw'
)

Flink SQL> select * from kafka_behavior_log_raw;

能读取出实时数据

3.2 定义iceberg表

建catalog,表

CREATE CATALOG hive_catalog6 WITH (
  'type'='iceberg',
  'catalog-type'='hive',
  'uri'='thrift://hadoop101:9083',
  'clients'='5',
  'property-version'='1',
  'warehouse'='hdfs:user/hive/warehouse/hive_catalog6'
);

 database 各个catalog是共享的
use catalog hive_catalog6;
create database iceberg_db6;

create table `hive_catalog6`.`iceberg_db6`.`behavior_log_ib`(
 log STRING
)

3.3 kafka表入到iceberg

切换回default_catalog; （因为kafka表在这个catalog创建的）

use catalog default_catalog;
insert into  `hive_catalog6`.`iceberg_db6`.`behavior_log_ib` select log from kafka_behavior_log_raw

4.观察生成的数据

结论：与checkpoint的时间间隔生成数据文件

生成4个文件

[root@hadoop102 ~]# hadoop fs -ls -R  /user/hive/warehouse/hive_catalog6/iceberg_db6.db/behavior_log_ib/
drwxr-xr-x   - root supergroup          0 2022-01-19 18:17 /user/hive/warehouse/hive_catalog6/iceberg_db6.db/behavior_log_ib/data
-rw-r--r--   2 root supergroup      10165 2022-01-19 18:15 /user/hive/warehouse/hive_catalog6/iceberg_db6.db/behavior_log_ib/data/00000-0-34111aa0-6c6c-4cf0-ba4f-6c13fb278c94-00001.parquet
-rw-r--r--   2 root supergroup        569 2022-01-19 18:16 /user/hive/warehouse/hive_catalog6/iceberg_db6.db/behavior_log_ib/data/00000-0-34111aa0-6c6c-4cf0-ba4f-6c13fb278c94-00002.parquet
-rw-r--r--   2 root supergroup        569 2022-01-19 18:17 /user/hive/warehouse/hive_catalog6/iceberg_db6.db/behavior_log_ib/data/00000-0-34111aa0-6c6c-4cf0-ba4f-6c13fb278c94-00003.parquet
-rw-r--r--   2 root supergroup          0 2022-01-19 18:17 /user/hive/warehouse/hive_catalog6/iceberg_db6.db/behavior_log_ib/data/00000-0-34111aa0-6c6c-4cf0-ba4f-6c13fb278c94-00004.parquet
-rw-r--r--   2 root supergroup      19899 2022-01-19 18:15 /user/hive/warehouse/hive_catalog6/iceberg_db6.db/behavior_log_ib/data/00001-0-330ad02b-7a62-42dd-9b64-982469a3b48a-00001.parquet
-rw-r--r--   2 root supergroup        676 2022-01-19 18:16 /user/hive/warehouse/hive_catalog6/iceberg_db6.db/behavior_log_ib/data/00001-0-330ad02b-7a62-42dd-9b64-982469a3b48a-00002.parquet
-rw-r--r--   2 root supergroup        670 2022-01-19 18:17 /user/hive/warehouse/hive_catalog6/iceberg_db6.db/behavior_log_ib/data/00001-0-330ad02b-7a62-42dd-9b64-982469a3b48a-00003.parquet
-rw-r--r--   2 root supergroup          0 2022-01-19 18:17 /user/hive/warehouse/hive_catalog6/iceberg_db6.db/behavior_log_ib/data/00001-0-330ad02b-7a62-42dd-9b64-982469a3b48a-00004.parquet
drwxr-xr-x   - root supergroup          0 2022-01-19 18:17 /user/hive/warehouse/hive_catalog6/iceberg_db6.db/behavior_log_ib/metadata
-rw-r--r--   2 root supergroup        998 2022-01-19 18:12 /user/hive/warehouse/hive_catalog6/iceberg_db6.db/behavior_log_ib/metadata/00000-61699181-b792-411f-a920-5cd2b931e598.metadata.json
-rw-r--r--   2 root supergroup       2096 2022-01-19 18:15 /user/hive/warehouse/hive_catalog6/iceberg_db6.db/behavior_log_ib/metadata/00001-6ace3ab3-88fd-4c07-9d6a-c68d966dfd65.metadata.json
-rw-r--r--   2 root supergroup       3225 2022-01-19 18:16 /user/hive/warehouse/hive_catalog6/iceberg_db6.db/behavior_log_ib/metadata/00002-b48d08cc-4eeb-442c-b590-db6420ffed35.metadata.json
-rw-r--r--   2 root supergroup       4354 2022-01-19 18:17 /user/hive/warehouse/hive_catalog6/iceberg_db6.db/behavior_log_ib/metadata/00003-1e64aa46-1346-40c2-bb8f-50c5d4b8beda.metadata.json
-rw-r--r--   2 root supergroup       5831 2022-01-19 18:16 /user/hive/warehouse/hive_catalog6/iceberg_db6.db/behavior_log_ib/metadata/46a441c4-fa76-4102-93bc-a8f844342d9c-m0.avro
-rw-r--r--   2 root supergroup       5831 2022-01-19 18:17 /user/hive/warehouse/hive_catalog6/iceberg_db6.db/behavior_log_ib/metadata/a9c597aa-5b1e-418c-a4c3-77266c6fef98-m0.avro
-rw-r--r--   2 root supergroup       5838 2022-01-19 18:15 /user/hive/warehouse/hive_catalog6/iceberg_db6.db/behavior_log_ib/metadata/f7587144-2769-4128-9251-20634cfc3e10-m0.avro
-rw-r--r--   2 root supergroup       3909 2022-01-19 18:17 /user/hive/warehouse/hive_catalog6/iceberg_db6.db/behavior_log_ib/metadata/snap-1810035882406080361-1-a9c597aa-5b1e-418c-a4c3-77266c6fef98.avro
-rw-r--r--   2 root supergroup       3792 2022-01-19 18:15 /user/hive/warehouse/hive_catalog6/iceberg_db6.db/behavior_log_ib/metadata/snap-6209374798665286118-1-f7587144-2769-4128-9251-20634cfc3e10.avro
-rw-r--r--   2 root supergroup       3867 2022-01-19 18:16 /user/hive/warehouse/hive_catalog6/iceberg_db6.db/behavior_log_ib/metadata/snap-7052328324703984043-1-46a441c4-fa76-4102-93bc-a8f844342d9c.avro

生成5个数据文件，和相应的metadata

[root@hadoop102 ~]# hadoop fs -ls -R  /user/hive/warehouse/hive_catalog6/iceberg_db6.db/behavior_log_ib/
drwxr-xr-x   - root supergroup          0 2022-01-19 18:18 /user/hive/warehouse/hive_catalog6/iceberg_db6.db/behavior_log_ib/data
-rw-r--r--   2 root supergroup      10165 2022-01-19 18:15 /user/hive/warehouse/hive_catalog6/iceberg_db6.db/behavior_log_ib/data/00000-0-34111aa0-6c6c-4cf0-ba4f-6c13fb278c94-00001.parquet
-rw-r--r--   2 root supergroup        569 2022-01-19 18:16 /user/hive/warehouse/hive_catalog6/iceberg_db6.db/behavior_log_ib/data/00000-0-34111aa0-6c6c-4cf0-ba4f-6c13fb278c94-00002.parquet
-rw-r--r--   2 root supergroup        569 2022-01-19 18:17 /user/hive/warehouse/hive_catalog6/iceberg_db6.db/behavior_log_ib/data/00000-0-34111aa0-6c6c-4cf0-ba4f-6c13fb278c94-00003.parquet
-rw-r--r--   2 root supergroup        574 2022-01-19 18:18 /user/hive/warehouse/hive_catalog6/iceberg_db6.db/behavior_log_ib/data/00000-0-34111aa0-6c6c-4cf0-ba4f-6c13fb278c94-00004.parquet
-rw-r--r--   2 root supergroup          0 2022-01-19 18:18 /user/hive/warehouse/hive_catalog6/iceberg_db6.db/behavior_log_ib/data/00000-0-34111aa0-6c6c-4cf0-ba4f-6c13fb278c94-00005.parquet
-rw-r--r--   2 root supergroup      19899 2022-01-19 18:15 /user/hive/warehouse/hive_catalog6/iceberg_db6.db/behavior_log_ib/data/00001-0-330ad02b-7a62-42dd-9b64-982469a3b48a-00001.parquet
-rw-r--r--   2 root supergroup        676 2022-01-19 18:16 /user/hive/warehouse/hive_catalog6/iceberg_db6.db/behavior_log_ib/data/00001-0-330ad02b-7a62-42dd-9b64-982469a3b48a-00002.parquet
-rw-r--r--   2 root supergroup        670 2022-01-19 18:17 /user/hive/warehouse/hive_catalog6/iceberg_db6.db/behavior_log_ib/data/00001-0-330ad02b-7a62-42dd-9b64-982469a3b48a-00003.parquet
-rw-r--r--   2 root supergroup        671 2022-01-19 18:18 /user/hive/warehouse/hive_catalog6/iceberg_db6.db/behavior_log_ib/data/00001-0-330ad02b-7a62-42dd-9b64-982469a3b48a-00004.parquet
-rw-r--r--   2 root supergroup          0 2022-01-19 18:18 /user/hive/warehouse/hive_catalog6/iceberg_db6.db/behavior_log_ib/data/00001-0-330ad02b-7a62-42dd-9b64-982469a3b48a-00005.parquet
drwxr-xr-x   - root supergroup          0 2022-01-19 18:18 /user/hive/warehouse/hive_catalog6/iceberg_db6.db/behavior_log_ib/metadata
-rw-r--r--   2 root supergroup        998 2022-01-19 18:12 /user/hive/warehouse/hive_catalog6/iceberg_db6.db/behavior_log_ib/metadata/00000-61699181-b792-411f-a920-5cd2b931e598.metadata.json
-rw-r--r--   2 root supergroup       2096 2022-01-19 18:15 /user/hive/warehouse/hive_catalog6/iceberg_db6.db/behavior_log_ib/metadata/00001-6ace3ab3-88fd-4c07-9d6a-c68d966dfd65.metadata.json
-rw-r--r--   2 root supergroup       3225 2022-01-19 18:16 /user/hive/warehouse/hive_catalog6/iceberg_db6.db/behavior_log_ib/metadata/00002-b48d08cc-4eeb-442c-b590-db6420ffed35.metadata.json
-rw-r--r--   2 root supergroup       4354 2022-01-19 18:17 /user/hive/warehouse/hive_catalog6/iceberg_db6.db/behavior_log_ib/metadata/00003-1e64aa46-1346-40c2-bb8f-50c5d4b8beda.metadata.json
-rw-r--r--   2 root supergroup       5483 2022-01-19 18:18 /user/hive/warehouse/hive_catalog6/iceberg_db6.db/behavior_log_ib/metadata/00004-5cd0ffd8-7336-46ac-a280-c29dbc0faabb.metadata.json
-rw-r--r--   2 root supergroup       5831 2022-01-19 18:16 /user/hive/warehouse/hive_catalog6/iceberg_db6.db/behavior_log_ib/metadata/46a441c4-fa76-4102-93bc-a8f844342d9c-m0.avro
-rw-r--r--   2 root supergroup       5836 2022-01-19 18:18 /user/hive/warehouse/hive_catalog6/iceberg_db6.db/behavior_log_ib/metadata/a995dc51-f98f-4030-9daf-973ea2cb8106-m0.avro
-rw-r--r--   2 root supergroup       5831 2022-01-19 18:17 /user/hive/warehouse/hive_catalog6/iceberg_db6.db/behavior_log_ib/metadata/a9c597aa-5b1e-418c-a4c3-77266c6fef98-m0.avro
-rw-r--r--   2 root supergroup       5838 2022-01-19 18:15 /user/hive/warehouse/hive_catalog6/iceberg_db6.db/behavior_log_ib/metadata/f7587144-2769-4128-9251-20634cfc3e10-m0.avro
-rw-r--r--   2 root supergroup       3909 2022-01-19 18:17 /user/hive/warehouse/hive_catalog6/iceberg_db6.db/behavior_log_ib/metadata/snap-1810035882406080361-1-a9c597aa-5b1e-418c-a4c3-77266c6fef98.avro
-rw-r--r--   2 root supergroup       3947 2022-01-19 18:18 /user/hive/warehouse/hive_catalog6/iceberg_db6.db/behavior_log_ib/metadata/snap-3397835470520958512-1-a995dc51-f98f-4030-9daf-973ea2cb8106.avro
-rw-r--r--   2 root supergroup       3792 2022-01-19 18:15 /user/hive/warehouse/hive_catalog6/iceberg_db6.db/behavior_log_ib/metadata/snap-6209374798665286118-1-f7587144-2769-4128-9251-20634cfc3e10.avro
-rw-r--r--   2 root supergroup       3867 2022-01-19 18:16 /user/hive/warehouse/hive_catalog6/iceberg_db6.db/behavior_log_ib/metadata/snap-7052328324703984043-1-46a441c4-fa76-4102-93bc-a8f844342d9c.avro

生成文件特点： 每隔一分钟生成一个数据文件， 与checkpoint的时间间隔一致

生成大量的metadata文件

5. count条数

Flink SQL> select count(*) from  `hive_catalog6`.`iceberg_db6`.`behavior_log_ib` ;
[INFO] Result retrieval cancelled.

1510

不断刷没什么变化，直到刷到下一个15288，比上一个大60

只按R刷新

5. 大量小文件如何处理？

每分钟生成一份文件，三个小文件（数据文件,metadata,快照文件），一天24小时60分钟/小时=1440分钟，14403个文件=4320个文件。
一天生成4000多个小文件如何进行合并？

---

总结