从 Kafka 消费 CDC 数据（变更捕获，需 Upsert 语义）
用 kafka 连接器 + 主键 + 处理函数 模拟 Upsert，示例：
CREATE TABLE `KAFKA_TEST_0002` (
`LGL_PERN_CODE` VARCHAR COMMENT 'LGL_PERN_CODE',
`LBLTY_ACCT_NUM` VARCHAR COMMENT 'LBLTY_ACCT_NUM',
`ACCT_NM` VARCHAR COMMENT 'ACCT_NM',
`CUST_NUM` VARCHAR COMMENT 'CUST_NUM',
`NAT_CODE` VARCHAR COMMENT 'NAT_CODE',
-- 声明主键（用于 Upsert 去重）
PRIMARY KEY (`LBLTY_ACCT_NUM`) NOT ENFORCED
) WITH (
'connector' = 'kafka', -- 恢复为 kafka 连接器
'topic' = 'KAFKA_TEST_0002',
'properties.bootstrap.servers' = '10.57.48.38:21007,10.57.48.37:21007,10.57.48.36:21007',
'properties.group.id' = '7a074dd07bfb4d4da39eb0f5773b952b',
'scan.startup.mode' = 'earliest-offset',
'format' = 'debezium-json', -- 适配 CDC 格式
'debezium-json.ignore-parse-errors' = 'true',
'debezium-json.schema-include' = 'true',
'properties.security.protocol' = 'SASL_PLAINTEXT',
'properties.kerberos.domain.name' = 'hadoop.124dba82_3b54_0125_81e4_110652049a41.com',
'properties.sasl.kerberos.service.name' = 'kafka'
);
-- 如需 Upsert 输出，再通过 Sink 写入 upsert-kafka
CREATE TABLE KafkaUpsertSink (
`LBLTY_ACCT_NUM` VARCHAR,
`LGL_PERN_CODE` VARCHAR,
`ACCT_NM` VARCHAR,
PRIMARY KEY (`LBLTY_ACCT_NUM`) NOT ENFORCED
) WITH (
'connector' = 'upsert-kafka', -- Sink 侧使用 upsert-kafka
'topic' = 'sink_topic',
'properties.bootstrap.servers' = '...',
'key.format' = 'json',
'value.format' = 'json'
);
-- 业务逻辑：从 Kafka 读 CDC 数据，处理后 Upsert 写入
INSERT INTO KafkaUpsertSink
SELECT LBLTY_ACCT_NUM, LGL_PERN_CODE, ACCT_NM
FROM `KAFKA_TEST_0002`;