卡夫卡连接日期处理 debezium 生成的事件

Question

我正在使用 debezium SQL 服务器来跟踪生产基地的变化。主题已创建，CDC 工作正常，但是当尝试使用 jdbcSinkConnector 将数据转储到另一个 Sql 服务器数据库时，我遇到了以下错误。

com.microsoft.sqlserver.jdbc.SQLServerException: One or more values is out of range of values for the datetime2 SQL Server data type

在源数据库上，sql 数据类型是 timestamp2(7)。 kafka事件是1549461754650000000。模式类型是 INT64。架构名称 io.debezium.time.NanoTimestamp.

我找不到一种方法来告诉 TimestampConverter 值不是以毫秒或微单位表示，而是以纳秒表示（无论如何都不适用于微秒）。

这是我的连接器配置

{
    "name": "cdc.swip.bi.ods.sink.contract",
    "config": {
        "connector.class": "io.confluent.connect.jdbc.JdbcSinkConnector",
        "tasks.max": "1",
        "topics": "swip.swip_core.contract",
        "connection.url": "jdbc:sqlserver://someip:1234;database=DB",
        "connection.user": "loloolololo",
        "connection.password": "muahahahahaha",
        "dialect.name": "SqlServerDatabaseDialect",
        "auto.create": "false",
        "key.converter": "io.confluent.connect.avro.AvroConverter",
        "value.converter": "io.confluent.connect.avro.AvroConverter",
        "key.converter.schemas.enable": "true",
        "key.converter.schema.registry.url": "http://localhost:8081",
        "value.converter.schemas.enable": "true",
        "value.converter.schema.registry.url": "http://localhost:8081",
        "transforms": "unwrap,created_date,modified_date",
        "transforms.unwrap.type": "io.debezium.transforms.UnwrapFromEnvelope",
        "transforms.created_date.type": "org.apache.kafka.connect.transforms.TimestampConverter$Value",
        "transforms.created_date.target.type": "Timestamp",
        "transforms.created_date.field": "created_date",
        "transforms.modified_date.type": "org.apache.kafka.connect.transforms.TimestampConverter$Value",
        "transforms.modified_date.target.type": "Timestamp",
        "transforms.modified_date.field": "modified_date",
        "insert.mode": "insert",
        "delete.enabled": "false",
        "pk.fields": "id",
        "pk.mode": "record_value",
        "schema.registry.url": "http://localhost:8081",
        "table.name.format": "ODS.swip.contract"
    }
}

Answer 1

SQL 服务器连接器中缺少一个功能 - DBZ-1419。

您可以通过编写自己的 SMT 解决此问题，该 SMT 会在 JDBC 连接器处理之前在接收器端进行字段转换。

Answer 2

我忘记了 post 答案。属性 "time.precision.mode":"connect" 成功了

https://debezium.io/documentation/reference/connectors/sqlserver.html#sqlserver-property-time-precision-mode

{
    "name":"debezium-connector-sqlserver",
    "config": {
        "connector.class":"io.debezium.connector.sqlserver.SqlServerConnector",
        "key.converter": "org.apache.kafka.connect.json.JsonConverter",
        "value.converter": "org.apache.kafka.connect.json.JsonConverter",
        "key.converter.schemas.enable":"true",
        "value.converter.schemas.enable":"true",
        "database.hostname":"someHost",
        "database.port":"somePort",
        "database.user":"someUser",
        "database.password":"somePassword",
        "database.dbname":"someDb",
        "database.server.name":"xxx.xxx",
        "database.history.kafka.topic":"xxx.xxx.history",
        "time.precision.mode":"connect",
        "database.history.kafka.bootstrap.servers":"example.com:9092"
    }
}

卡夫卡连接日期处理 debezium 生成的事件

Kafka Connect date handling of debezium generated events

apache-kafka

apache-kafka-connect

debezium