Flink SQL 读取Kafka写入Mysql

示例demo链接：

https://www.lmlphp.com/user/8087/article/item/368437/

github链接：

https://github.com/wuchong/flink-sql-submit?spm=a2c4e.10696291.0.0.3ebb19a4uhktCo

踩坑注意：
1、kafka输入的事件戳格式为：“2017-11-26T01:00:00Z”，其他格式如："2019-11-26 13:30:30"无法正常解析。
2、尾部的分号别忘记，缺少会报sql解析错误，不注意会很懵逼。

我的Sql(调试通过)：

CREATE TABLE hot_word_in (
word_name VARCHAR,
ts TIMESTAMP
) WITH (
‘connector.type’ = ‘kafka’,
‘connector.version’ = ‘universal’,
‘connector.topic’ = ‘hot_word_topic01’,
‘connector.startup-mode’ = ‘earliest-offset’,
‘connector.properties.0.key’ = ‘zookeeper.connect’,
‘connector.properties.0.value’ = ‘localhost:2181’,
‘connector.properties.1.key’ = ‘bootstrap.servers’,
‘connector.properties.1.value’ = ‘localhost:9092’,
‘update-mode’ = ‘append’,
‘format.type’ = ‘json’,
‘format.derive-schema’ = ‘true’
);

CREATE TABLE hot_word_sink (
word_name VARCHAR,
word_count BIGINT,
ts VARCHAR
) WITH (
‘connector.type’ = ‘jdbc’,
‘connector.url’ = ‘jdbc:mysql://134.175.107.12:3306/flink_test’,
‘connector.table’ = ‘hot_word_sink’,
‘connector.username’ = ‘root’,
‘connector.password’ = ‘jack’,
‘connector.write.flush.max-rows’ = ‘1’
);

INSERT INTO hot_word_sink
SELECT
word_name,
count(1) as word_count,
DATE_FORMAT(ts, ‘yyyy-MM-dd HH:mm:00’) as ts
FROM hot_word_in
GROUP BY word_name, DATE_FORMAT(ts, ‘yyyy-MM-dd HH:mm:00’);