Schema Registry 中的向后兼容性问题和不确定性

Question

我有一个用例，我有一个 JSON，我想生成架构并从 JSON 中记录并发布一条记录。我已经配置了值序列化器并且架构设置是向后兼容的。

第一个JSON

String json = "{\n" +

     "    \"id\": 1,\n" +

     "    \"name\": \"Headphones\",\n" +

     "    \"price\": 1250.0,\n" +

     "    \"tags\": [\"home\", \"green\"]\n" +

     "}\n"
     ;

版本 1 架构已注册。

在 avro 控制台消费者中收到消息。

第二个JSON.

String json = "{\n" +

    "    \"id\": 1,\n" +

    "    \"price\": 1250.0,\n" +
    "    \"tags\": [\"home\", \"green\"]\n" +
    "}\n"
    ;

注册架构成功。已发送消息。

现在尝试发送 JSON 1 发送成功

架构 3：

String json = "{\n" +

    "    \"id\": 1,\n" +
    "    \"name\": \"Headphones\",\n" +

    "    \"tags\": [\"home\", \"green\"]\n" +
    "}\n"
    ;

遇到这种情况的错误。 Caused by: io.confluent.kafka.schemaregistry.client.rest.exceptions.RestClientException: 正在注册的模式与早期模式不兼容；错误代码：409

How is that schema generated from 2nd JSON was registered and the third one was rejected? Although I didn't have any Default key for the deleted field? Is it that the Schema Registry always accepts the 1st evolution? (2nd schema over 1st)

模式注册表中的模式

版本 1 架构

{ "fields": [

{

  "doc": "Type inferred from '1'",

  "name": "id",

  "type": "int"

},

{

  "doc": "Type inferred from '\"Headphones\"'",

  "name": "name",

  "type": "string"

},

{

  "doc": "Type inferred from '1250.0'",

  "name": "price",

  "type": "double"

},

{

  "doc": "Type inferred from '[\"home\",\"green\"]'",

  "name": "tags",

  "type": {

    "items": "string",

    "type": "array"

  }

}
],
"name": "myschema",
"type": "record"   }

版本 2：

{ "fields": [

{

  "doc": "Type inferred from '1'",

  "name": "id",

  "type": "int"

},

{

  "doc": "Type inferred from '1250.0'",

  "name": "price",

  "type": "double"

},

{

  "doc": "Type inferred from '[\"home\",\"green\"]'",

  "name": "tags",

  "type": {

    "items": "string",

    "type": "array"

  }

}
],
"name": "myschema",
"type": "record"   }

Answer 1

让我们回顾一下向后兼容规则...https://docs.confluent.io/current/schema-registry/avro.html#compatibility-types

首先，默认值是不可传递的，因此版本 3 只会查看版本 2。

向后规则规定您可以删除字段或添加可选字段（具有默认值的字段）。我假设您的模式生成器工具不知道如何使用可选值，因此您只能删除，不能添加。

在版本 1 和版本 2 之间，您删除了有效的名称字段。

在版本 2 和即将到来的版本 3 之间，它认为您正在尝试 post 一个删除价格的新模式（这没关系}，但添加了一个必需的名称字段，这是不允许的。

Schema Registry 中的向后兼容性问题和不确定性

Backward Comaptibility issue and uncertainity in Schema Registry

avro

apache-kafka

kafka-producer-api

confluent-schema-registry