0

我正在使用 kafka-connect 将 MySQL 表中的行流式传输到 kafka 主题中。这很好用。

然后我创建一个表:

CREATE TABLE mytable (id INT, email VARCHAR, gender VARCHAR, first_name VARCHAR, last_name VARCHAR) WITH (KAFKA_TOPIC='mysql-my-table', VALUE_FORMAT='AVRO', KEY='id');

这也有效,我可以通过以下方式确认:

LIST TABLES; 
DESCRIBE EXTENDED mytable;

我明白了mytable

问题是当我执行

SELECT * FROM mytable;

然后我没有得到任何结果并且提示没有响应,我必须按下ctrl+c才能重新获得控制权。

可能是什么问题呢?

4

1 回答 1

0

经过一段时间尝试不同的事情并阅读文档后,我发现了问题。

id我的主题中消息的字段类型为 ? INT并且根据 KSQL TABLES 文档,它们需要是类型VARCHAR

所以我按照这里描述的步骤来解决这个问题,现在一切正常。

所以步骤如下:

-- Create a stream on the original topic
CREATE STREAM users_with_wrong_key_format (userid INT, username VARCHAR, email VARCHAR)
WITH (KAFKA_TOPIC='users', VALUE_FORMAT='JSON');

-- Derive a new stream with the required key changes.
-- 1) The CAST statement converts the key to the required format.
-- 2) The PARTITION BY clause re-partitions the stream based on the new, converted key.
CREATE STREAM users_with_proper_key
WITH(KAFKA_TOPIC='users-with-proper-key') AS
SELECT CAST(userid as VARCHAR) as userid_string, username, email
FROM users_with_wrong_key_format
PARTITION BY userid_string;

-- Now you can create the table on the properly keyed stream.
CREATE TABLE users_table (userid_string VARCHAR, username VARCHAR, email VARCHAR)
WITH (KAFKA_TOPIC='users-with-proper-key',
        VALUE_FORMAT='JSON',
        KEY='userid_string');

-- Now you can create the table on the properly keyed stream.
CREATE TABLE users_table (userid_string VARCHAR, username VARCHAR, email VARCHAR)
WITH (KAFKA_TOPIC='users-with-proper-key',
    VALUE_FORMAT='JSON',
    KEY='userid_string');
于 2018-07-20T11:55:28.567 回答