本地二级索引的 PynamoDB 错误:ValueError('Table table_4 has no index: key1_index',)

PynamoDB Error for Local Secondary Index: ValueError('Table table_4 has no index: key1_index',)

我是 pynamodb 的新手,我正在尝试查询 LSI(已在 dynamodb main table 中定义)并得到以下异常。

'exception': ValueError('Table table_4 has no index: key1_index',)

用例:我只想在 hash_key 上查询 LSI,因为我已经为 key1 (range_key) 提供了默认值,并希望根据 key1 获取排序后的数据。

大规模集成电路型号

class LSI_Index(LocalSecondaryIndex):
    id = UnicodeAttribute(hash_key=True)
    key1 = NumberAttribute(default=int(time.time()), range_key=True)

    class Meta:
        index = "key1-Index"
        projection = IncludeProjection(['id', 'key1', 'key2'])

主要Table型号

class MainModel(Model):
    id = UnicodeAttribute(hash_key=True)
    key2 = UnicodeAttribute(range_key=True, default='')
    key1 = NumberAttribute(default=int(time.time()))
    key1_index = LSI_Index()   # LSI

    class Meta:
        .....

存储库方法代码

    def read_by_LSI(cls, id):
    try:
        data = []
        iterator = MainModel.key1_index.query(hash_key=id,limit=1,scan_index_forward=False, consistent_read=True)
        for item in iterator:
            data.append(json.loads(json_dumps(item)))
        return data

DynamoDB 主Table描述:

{
    "Table": {
        "AttributeDefinitions": [
            {
                "AttributeName": "id",
                "AttributeType": "S"
            },
            {
                "AttributeName": "key1",
                "AttributeType": "N"
            }
        ],
        "TableName": "table_4",
        "KeySchema": [
            {
                "AttributeName": "id",
                "KeyType": "HASH"
            }
        ],
        "TableStatus": "ACTIVE",
        "CreationDateTime": 1647447620.911,
        "ProvisionedThroughput": {
            "LastIncreaseDateTime": 0.0,
            "LastDecreaseDateTime": 0.0,
            "NumberOfDecreasesToday": 0,
            "ReadCapacityUnits": 10,
            "WriteCapacityUnits": 5
        },
        "TableSizeBytes": 779,
        "ItemCount": 7,
        "TableArn": "arn:aws:dynamodb:ddblocal:000000000000:table/table_4",
        "LocalSecondaryIndexes": [
            {
                "IndexName": "key1-Index",
                "KeySchema": [
                    {
                        "AttributeName": "id",
                        "KeyType": "HASH"
                    },
                    {
                        "AttributeName": "key1",
                        "KeyType": "RANGE"
                    }
                ],
                "Projection": {
                    "ProjectionType": "INCLUDE",
                    "NonKeyAttributes": [
                        "key2",
                        "key1",
                    ]
                },
                "IndexSizeBytes": 779,
                "ItemCount": 7,
                "IndexArn": "arn:aws:dynamodb:ddblocal:000000000000:table/table_4/index/key1-Index"
            }
        ]
   }
}

请告诉我哪里做错了,因为我无法调试问题。 提前致谢。!!

所以基本上,在 pynamodb 中定义 LSI 模型时,请确保 Meta class 属性名称与文档中提到的相同 - https://pynamodb.readthedocs.io/en/latest/indexes.html

在上面的例子中,我犯了以下错误,导致出现 ValueError('table table_4 doesn't have key1_index') Exception.

在 LSI_Model Class 的 MetaClass 中,我使用了错误的属性名称 'index' 而不是 'index_name'。我已经更新了代码,现在可以正常工作了。

class LSI_Index(LocalSecondaryIndex):
id = UnicodeAttribute(hash_key=True)
key1 = NumberAttribute(default=int(time.time()), range_key=True)

class Meta:
    index_name = "key1-Index"
    projection = IncludeProjection(['id', 'key1', 'key2'])

注: LSI Meta Class 属性应命名如下(文档中提到):

  1. index_name
  2. 投影 等等..