Flink流批一体计算（23）：Flink SQL之多流kafka写入多个mysql sink

docker exec -it 192d1369463a bash

bash-5.1# cd /opt/kafka_2.12-2.5.0/bin

bash-5.1# ./kafka-console-producer.sh --broker-list localhost:9092 --topic case_kafka_mysql

&gt;{"ts": "20201011","id": 8,"price_amt":211}

topic: flink_test_2

docker exec -it 192d1369463a bash

bash-5.1# cd /opt/kafka_2.12-2.5.0/bin

bash-5.1# ./kafka-console-producer.sh --broker-list localhost:9092 --topic flink_test_2

&gt;{"id": 8,"coupon_price_amt":100}

创建 数据表

mysql 建表语句

CREATE TABLE `sync_test_2` (

  `id` bigint(11) NOT NULL AUTO_INCREMENT,

  `ts` varchar(64) DEFAULT NULL,

  `total_gmv` bigint(11) DEFAULT NULL,

  PRIMARY KEY (`id`),

  UNIQUE KEY `uidx` (`ts`) USING BTREE

) ENGINE=InnoDB AUTO_INCREMENT=5 DEFAULT CHARSET=utf8mb4;



CREATE TABLE `sync_test_22` (

  `id` bigint(11) NOT NULL AUTO_INCREMENT,

  `ts` varchar(64) DEFAULT NULL,

  `coupon_ratio` double DEFAULT NULL,

  PRIMARY KEY (`id`),

  UNIQUE KEY `uidx` (`ts`) USING BTREE

) ENGINE=InnoDB AUTO_INCREMENT=5 DEFAULT CHARSET=utf8mb4;

2. 创建 数据表

创建数据源表

create table flink_test_2_1 (
  id BIGINT,
  ts VARCHAR,
  price_amt BIGINT,
  proctime AS PROCTIME ()
)
 with (
   'connector' = 'kafka',
   'topic' = 'case_kafka_mysql',
   'properties.bootstrap.servers' = '127.0.0.1:9092',
   'properties.group.id' = 'flink_gp_test2-1',
   'scan.startup.mode' = 'earliest-offset',
   'format' = 'json',
   'json.fail-on-missing-field' = 'false',
   'json.ignore-parse-errors' = 'true',
   'properties.zookeeper.connect' = '127.0.0.1:2181/kafka'
 );

create table flink_test_2_2 (
  id BIGINT,
  coupon_price_amt BIGINT,
  proctime AS PROCTIME ()
)
 with (
   'connector' = 'kafka',
   'topic' = 'flink_test_2',
   'properties.bootstrap.servers' = '127.0.0.1:9092',
   'properties.group.id' = 'flink_gp_test2-2',
   'scan.startup.mode' = 'earliest-offset',
   'format' = 'json',
   'json.fail-on-missing-field' = 'false',
   'json.ignore-parse-errors' = 'true',
   'properties.zookeeper.connect' = '127.0.0.1:2181/kafka'
 );

关键配置的说明：

json.fail-on-missing–field：在json缺失字段时是否报错

json.ignore–parse–errors：在解析json失败时是否报错

一般无法保证json格式，所以以上两个配置是比较重要的。

创建数据目标表

CREATE TABLE sync_test_2 (
                   ts string,
                   total_gmv bigint,
                   PRIMARY KEY (ts) NOT ENFORCED

 ) WITH (
   'connector' = 'jdbc',
   'url' = 'jdbc:mysql://127.0.0.1:3306/db01?characterEncoding=UTF-8',
   'table-name' = 'sync_test_2',
   'username' = 'root',
   'password' = 'Admin'
 );

CREATE TABLE sync_test_22 (
                   ts string,
                   coupon_ration bigint,
                   PRIMARY KEY (ts) NOT ENFORCED
 ) WITH (
   'connector' = 'jdbc',
   'url' = 'jdbc:mysql://127.0.0.1:3306/db01?characterEncoding=UTF-8',
   'table-name' = 'sync_test_2',
   'username' = 'root',
   'password' = 'Admin'
 );

3. 计算

一个作业中写入一个Sink或多个Sink。

说明写入多个Sink语句时，需要以BEGIN STATEMENT SET;开头，以END;结尾。

BEGIN STATEMENT SET;      --写入多个Sink时，必填。
INSERT INTO sync_test_2
SELECT
  ts,
  SUM(price_amt - coupon_price_amt) AS total_gmv
FROM
  (
    SELECT
      a.ts as ts,
      a.price_amt as price_amt,
      b.coupon_price_amt as coupon_price_amt
    FROM
      flink_test_2_1 as a
      LEFT JOIN flink_test_2_2 b on b.id = a.id
  )
GROUP BY ts;

INSERT INTO sync_test_22
SELECT
  ts,
  sum(coupon_price_amt)/sum(amount) AS coupon_ration
FROM
  (
    SELECT
      a.ts as ts,
      a.price_amt as price_amt,
      b.coupon_price_amt as coupon_price_amt
    FROM
      flink_test_2_1 as a
      LEFT JOIN flink_test_2_2 b on b.id = a.id
  )
GROUP BY ts;;
END;      --写入多个Sink时，必填。

WITH子句

WITH提供了一种编写辅助语句以用于更大的查询的方法。这些语句通常被称为公共表表达式（CTE），可以被视为定义仅针对一个查询存在的临时视图。

改写上述查询：

BEGIN STATEMENT SET;      --写入多个Sink时，必填。
with orders_with_coupon AS (
    SELECT
      a.ts as ts,
      a.price_amt as price_amt,
      b.coupon_price_amt as coupon_price_amt
    FROM
      flink_test_2_1 as a
      LEFT JOIN flink_test_2_2 b on b.id = a.id
)

INSERT INTO sync_test_2
SELECT
  ts,
  SUM(price_amt - coupon_price_amt) AS total_gmv
FROM orders_with_coupon
GROUP BY ts;

INSERT INTO sync_test_22
SELECT
  ts,
  coupon_price_amt/price_amt AS coupon_ration
FROM orders_with_coupon
GROUP BY ts;;
END;      --写入多个Sink时，必填。