外部数据源中指定的错误或不可访问的位置

Question

我正在尝试将 Azure 文件存储中的文件保存到 Azure SQL 数据库表 varbinary(max) 列（按照建议存储整个内容 in this answer）。我尝试了几次调整我的 SQL 查询，但没有成功。这是导致错误“外部数据源“my_Azure_Files”中指定的位置错误或无法访问的代码。当它调用 OPENROWSET 时：

OPEN MASTER KEY DECRYPTION BY PASSWORD = 'mypassword123'
GO

CREATE DATABASE SCOPED CREDENTIAL [https://mystorageaccount.file.core.windows.net/]
WITH IDENTITY = 'SHARED ACCESS SIGNATURE',
SECRET = 'sas_token_generated_on_azure_portal';

CREATE EXTERNAL DATA SOURCE my_Azure_Files
    WITH (
        LOCATION = 'https://mystorageaccount.file.core.windows.net/test', 
        CREDENTIAL = [https://mystorageaccount.file.core.windows.net/],
        TYPE = BLOB_STORAGE
    );

Insert into dbo.myTable(targetColumn)
Select BulkColumn FROM OPENROWSET( 
BULK 'test.csv', 
DATA_SOURCE = 'my_Azure_Files', 
SINGLE_BLOB) AS testFile;

CLOSE MASTER KEY;  
GO

我可以通过网络浏览器使用相同的 SAS 令牌和 url 路径下载 test.csv 文件。我还能够验证凭据和外部源是否已在数据库中成功创建：

+-------------------------------------------------+------------------+-----------------------------------------------------+-------------------------+------------------+---------------------------+---------------+---------------+----------------+--------------------+----------+
| data_source_id                                  | name             | location                                            | type_desc               | type             | resource_manager_location | credential_id | database_name | shard_map_name | connection_options | pushdown |
+-------------------------------------------------+------------------+-----------------------------------------------------+-------------------------+------------------+---------------------------+---------------+---------------+----------------+--------------------+----------+
| 65540                                           | my_Azure_Files   | https://mystorageaccount.file.core.windows.net/test | BLOB_STORAGE            | 05/01/1900 00:00 | NULL                      | 65539         | NULL          | NULL           | NULL               | ON       |
           

| name                                            | principal_id     | credential_id                                       | credential_identity     | create_date      | modify_date               | target_type   | target_id     |                |                    |          |
+-------------------------------------------------+------------------+-----------------------------------------------------+-------------------------+------------------+---------------------------+---------------+---------------+----------------+--------------------+----------+
| https://mystorageaccount.file.core.windows.net/ | 1                | 65539                                               | SHARED ACCESS SIGNATURE | 15/07/2020 13:14 | 15/07/2020 13:14          | NULL          | NULL          |                |                    |          |

在 Azure 门户上创建 SAS 时，我检查了所有允许的资源类型和所有允许的权限，“删除”除外。我还删除了领先的“？”从 SAS 到 SECRET 字段中使用。

我尝试了TYPE = BLOB_STORAGE 和TYPE = HADOOP 以及SINGLE_BLOB、SINGLE_CLOB 和SINGLE_NCLOB 参数的变体。

请帮我解决我的问题。

Answer 1

按照以下步骤，可以成功插入到目标表中：

1. 在生成 SAS 时，请选择允许的资源类型为“容器”和“对象”：

1. 复制 SAS 并使用以下命令：

    CREATE MASTER KEY ENCRYPTION BY PASSWORD = 'password@123'

1. 使用在开始时不带“？”生成的 SAS 令牌并创建 Scoped Credentials：

    CREATE DATABASE SCOPED CREDENTIAL MyAzureBlobStorageCredential  WITH IDENTITY = 
    'SHARED ACCESS SIGNATURE',  SECRET = 'sv=2019-10- 
    10XXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXX';

4. 创建引用您的 blob 路径的外部数据源：

    CREATE EXTERNAL DATA SOURCE MyAzureBlobStorage 
    WITH ( TYPE = BLOB_STORAGE,
    LOCATION = 'https://mystorageaccount.file.core.windows.net'
    , CREDENTIAL= MyAzureBlobStorageCredential 
    );

5. 使用 OPENROWSET 运行插入：

    Insert into dbo.test(name1)
    Select BulkColumn FROM OPENROWSET( 
    BULK 'test/test.csv',
    DATA_SOURCE = 'MyAzureBlobStorage', 
    SINGLE_BLOB) AS testFile;

6. 也可以使用批量插入：

    BULK INSERT dbo.test
    FROM 'test/test.csv'
    WITH (DATA_SOURCE = 'MyAzureBlobStorage',
    FORMAT = 'CSV');

假设表 dbo.test 已经创建