Seatunnel 2.3.1 部署使用
大数据相关文件可以参考我的csdn或者github
1 部署
1.1 下载解压
https://dlcdn.apache.org/incubator/seatunnel/2.3.1/apache-seatunnel-incubating-2.3.1-bin.tar.gz
下载完毕之后上传到服务器上面并解压
1
2
| # 解压到了/opt/module目录下
tar -zxvf apache-seatunnel-incubating-2.3.1-bin.tar.gz -C /opt/module
|
1.2 下载对应的connector
在apache的仓库下载相应的connector,下载时每个jar包在不同的路径下面，放到/seatunnel-2.3.1/connectors/seatunnel目录下
https://repo.maven.apache.org/maven2/org/apache/seatunnel/
1
2
3
4
5
6
7
8
9
10
11
12
13
| connector-assert-2.3.1.jar
connector-cdc-mysql-2.3.1.jar
connector-console-2.3.1.jar # 自带的
connector-doris-2.3.1.jar
connector-elasticsearch-2.3.1.jar
connector-fake-2.3.1.jar # 自带的
connector-file-hadoop-2.3.1.jar
connector-file-local-2.3.1.jar
connector-hive-2.3.1.jar
connector-iceberg-2.3.1.jar
connector-jdbc-2.3.1.jar
connector-kafka-2.3.1.jar
connector-redis-2.3.1.jar
|
配置安装seatunnel的插件
1
| vim seatunnel-2.3.1/config/plugin_config
|
调用安装脚本的时候会在maven的中央仓库下载对应的jar包，尽量少放，下载太慢了,我放了这些
1
2
3
4
5
6
7
| --connectors-v2--
connector-assert
connector-cdc-mysql
connector-jdbc
connector-fake
connector-console
--end--
|
1.3 安装seatunnel
1
| sh bin/install-plugin.sh 2.3.1
|
整个过程非常慢…应该是从maven中央仓库下载东西
⭐1.4 补充一些jar包
- 使用hive的话需要将这3个jar放入到seatunnel-2.3.1/lib目录下：
1
2
3
4
5
6
7
8
9
10
11
12
13
| hive-exec-2.3.9.jar
# 下载链接
# https://repo.maven.apache.org/maven2/org/apache/hive/hive-exec/2.3.9/hive-exec-2.3.9.jar
# 注意这里是hive-exec-2.3.9.jar，不要从你的hive的lib目录下拷贝最新的jar包，就用这个
seatunnel-hadoop3-3.1.4-uber-2.3.1.jar
# 下载链接
# https://repo.maven.apache.org/maven2/org/apache/seatunnel/seatunnel-hadoop3-3.1.4-uber/2.3.1/seatunnel-hadoop3-3.1.4-uber-2.3.1.jar
seatunnel-hadoop3-3.1.4-uber-2.3.1-optional.jar
# 下载链接
# hhttps://repo.maven.apache.org/maven2/org/apache/seatunnel/seatunnel-hadoop3-3.1.4-uber/2.3.1/seatunnel-hadoop3-3.1.4-uber-2.3.1-optional.jar
|
中间由于其他缘故我拷贝了一个hive框架/lib目录下的libfb303-0.9.3.jar放到seatunnel的lib目录下了。
- 使用
mysql的话需要将mysql的驱动拷贝过来，应该是需要8系列的mysql驱动，我这里使用的是mysql-connector-java-8.0.21.jar
2 测试样例
2.1 官方demo fake to console
seatunnel-2.3.1/config/v2.batch.config.template
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
| env {
execution.parallelism = 2
job.mode = "BATCH"
checkpoint.interval = 10000
}
source {
FakeSource {
parallelism = 2
result_table_name = "fake"
row.num = 16
schema = {
fields {
name = "string"
age = "int"
}
}
}
}
sink {
Console {
}
}
|
运行命令
1
2
| cd /opt/module/seatunnel-2.3.1
./bin/seatunnel.sh --config ./config/v2.batch.config.template -e lcoal
|
运行成功的话会可以在console看到打印的测试数据
2.2 mysql to console
我新建了一个用来放运行配置的目录/opt/module/seatunnel-2.3.1/job
mysql_2console.conf
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
| env {
execution.parallelism = 2
job.mode = "BATCH"
checkpoint.interval = 10000
}
source{
Jdbc {
url = "jdbc:mysql://hadoop102/dim_db?useUnicode=true&characterEncoding=utf8&useSSL=false"
driver = "com.mysql.cj.jdbc.Driver"
connection_check_timeout_sec = 100
user = "root"
password = "111111"
query = "select * from dim_basicdata_date_a_d where date < '2010-12-31'"
}
}
sink {
Console {
}
}
|
查询的是一张日期维表的数据
建表语句：
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
| CREATE DATABASE dim_db DEFAULT CHARACTER SET utf8 DEFAULT COLLATE utf8_general_ci;
drop table if exists dim_db.dim_basicdata_date_a_d;
create table if not exists dim_db.dim_basicdata_date_a_d
(
`date` varchar(40) comment '日期',
`year` varchar(40) comment '年',
`quarter` varchar(40) comment '季度（1/2/3/4）',
`season` varchar(40) comment '季节（春季/夏季/秋季/冬季）',
`month` varchar(40) comment '月',
`day` varchar(40) comment '日',
`week` varchar(40) comment '年内第几周',
`weekday` varchar(40) comment '周几（1-周一/2-周二/3-周三/4-周四/5-周五/6-周六/7-周日）',
`is_workday` varchar(40) comment '是否是工作日（1-是,0-否）',
`date_type` varchar(40) comment '节假日类型（工作日/法定上班[调休]/周末/节假日）',
`update_date` varchar(40) comment '更新日期'
);
|
可以自己插入几条数据试试
运行命令
1
2
| cd /opt/module/seatunnel-2.3.1
./bin/seatunnel.sh --config ./job/mysql_2console.conf -e local
|
2.3 hive to console
创建一张hive表
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
| CREATE database db_hive;
drop table if exists db_hive.dim_basicdata_date_a_d;
create table if not exists db_hive.dim_basicdata_date_a_d
(
`date` string comment '日期',
`year` string comment '年',
`quarter` string comment '季度（1/2/3/4）',
`season` string comment '季节（春季/夏季/秋季/冬季）',
`month` string comment '月',
`day` string comment '日',
`week` string comment '年内第几周',
`weekday` string comment '周几（1-周一/2-周二/3-周三/4-周四/5-周五/6-周六/7-周日）',
`is_workday` string comment '是否是工作日（1-是,0-否）',
`date_type` string comment '节假日类型（工作日/法定上班[调休]/周末/节假日）',
`update_date` string comment '更新日期'
);
|
自行插入几条数据
创建配置文件hive_2console.conf
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
| env {
execution.parallelism = 2
job.mode = "BATCH"
checkpoint.interval = 10000
}
source{
Hive {
table_name = "db_hive.dim_basicdata_date_a_d"
metastore_uri = "thrift://hadoop102:9083"
}
}
sink {
Console {
}
}
|
这里我使用的hive连接方式是jdbc访问元数据，所以metastore_uri = "jdbc:hive2://hadoop102:10000"也可以正常使用。
hive-site.xml修改配置文件,有可能你已经配置好了
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
|
<property>
<name>hive.metastore.uris</name>
<value>thrift://hadoop102:9083</value>
</property>
<property>
<name>hive.server2.thrift.bind.host</name>
<value>hadoop102</value>
</property>
<property>
<name>hive.server2.thrift.port</name>
<value>10000</value>
</property>
|
运行命令
1
2
| cd /opt/module/seatunnel-2.3.1
./bin/seatunnel.sh --config ./job/hive_2console.conf -e local
|
2.4 mysql to hive
创建配置文件
dim_basicdate_mysql_2hive.conf
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
| env {
execution.parallelism = 2
job.mode = "BATCH"
checkpoint.interval = 10000
}
source{
Jdbc {
url = "jdbc:mysql://hadoop102/dim_db?useUnicode=true&characterEncoding=utf8&useSSL=false"
driver = "com.mysql.cj.jdbc.Driver"
connection_check_timeout_sec = 100
user = "root"
password = "111111"
query = "select * from dim_basicdata_date_a_d"
}
}
sink {
Hive {
table_name = "db_hive.dim_basicdata_date_a_d"
metastore_uri = "thrift://hadoop102:9083"
}
}
|
运行命令
1
2
| cd /opt/module/seatunnel-2.3.1
./bin/seatunnel.sh --config ./job/dim_basicdate_mysql_2hive.conf-e local
|
