Managed identities for Document Translation

Managed identities for Azure resources are service principals that create a Microsoft Entra identity and specific permissions for Azure managed resources. Managed identities are a safer way to grant access to storage data and replace the requirement for you to include shared access signature tokens (SAS) with your source and target URLs.

• You can use managed identities to grant access to any resource that supports Microsoft Entra authentication, including your own applications.

• To grant access to an Azure resource, assign an Azure role to a managed identity using Azure role-based access control (Azure RBAC).

• There's no added cost to use managed identities in Azure.

Prerequisites

To get started, you need:

• An active Azure account—if you don't have one, you can create a Trial.

• A single-service Translator (not a multi-service Azure AI services) resource assigned to a geographical region such as China North. For detailed steps, see Create an Azure AI services resource.

• A brief understanding of Azure role-based access control (Azure RBAC) using the Azure portal.

• An Azure Blob Storage account in the same region as your Translator resource. You also need to create containers to store and organize your blob data within your storage account.

• If your storage account is behind a firewall, you must enable the following configuration:

  1. Go to the Azure portal and sign in to your Azure account.

  2. Select the Storage account.

  3. In the Security + networking group in the left pane, select Networking.

  4. In the Firewalls and virtual networks tab, select Enabled from selected virtual networks and IP addresses.

     

  5. Deselect all check boxes.

  6. Make sure Microsoft network routing is selected.

  7. Under the Resource instances section, select Microsoft.CognitiveServices/accounts as the resource type and select your Translator resource as the instance name.

  8. Make certain that the Allow Azure services on the trusted services list to access this storage account box is checked. For more information about managing exceptions, see Configure Azure Storage firewalls and virtual networks.

     

  9. Select Save.

     Note

     It may take up to 5 min for the network changes to propagate.

  Although network access is now permitted, your Translator resource is still unable to access the data in your Storage account. You need to create a managed identity for and assign a specific access role to your Translator resource.

Managed identity assignments

There are two types of managed identities: system-assigned and user-assigned. Currently, Document Translation supports system-assigned managed identity:

• A system-assigned managed identity is enabled directly on a service instance. It isn't enabled by default; you must go to your resource and update the identity setting.

• The system-assigned managed identity is tied to your resource throughout its lifecycle. If you delete your resource, the managed identity is deleted as well.

In the following steps, we enable a system-assigned managed identity and grant your Translator resource limited access to your Azure Blob Storage account.

Enable a system-assigned managed identity

You must grant the Translator resource access to your storage account before it can create, read, or delete blobs. Once you enabled the Translator resource with a system-assigned managed identity, you can use Azure role-based access control (Azure RBAC), to give Translator access to your Azure storage containers.

1. Go to the Azure portal and sign in to your Azure account.

2. Select the Translator resource.

3. In the Resource Management group in the left pane, select Identity.

4. Within the System assigned tab, turn on the Status toggle.

   

   Important

   User assigned managed identity won't meet requirements for the batch transcription storage account scenario. Be sure to enable system assigned managed identity.

5. Select Save.

Grant storage account access for your Translator resource

1. Go to the Azure portal and sign in to your Azure account.

2. Select the Translator resource.

3. In the Resource Management group in the left pane, select Identity.

4. Under Permissions select Azure role assignments:

   

5. On the Azure role assignments page that opened, choose your subscription from the drop-down menu then select + Add role assignment.

   

6. Next, assign a Storage Blob Data Contributor role to your Translator service resource. The Storage Blob Data Contributor role gives Translator (represented by the system-assigned managed identity) read, write, and delete access to the blob container and data. In the Add role assignment pop-up window, complete the fields as follows and select Save:

   Field	Value
   Scope	Storage.
   Subscription	The subscription associated with your storage resource.
   Resource	The name of your storage resource.
   Role	Storage Blob Data Contributor.

   

7. After the Added Role assignment confirmation message appears, refresh the page to see the added role assignment.

   

8. If you don't see the new role assignment right away, wait and try refreshing the page again. When you assign or remove role assignments, it can take up to 30 minutes for changes to take effect.

   

HTTP requests

• An asynchronous batch translation request is submitted to your Translator service endpoint via a POST request.

• With managed identity and Azure RBAC, you no longer need to include SAS URLs.

• If successful, the POST method returns a 202 Accepted response code and the service creates a batch request.

• The translated documents appear in your target container.

Headers

The following headers are included with each Document Translation API request:

HTTP header	Description
Ocp-Apim-Subscription-Key	Required: The value is the Azure key for your Translator or Azure AI services resource.
Content-Type	Required: Specifies the content type of the payload. Accepted values are application/json or charset=UTF-8.

POST request body

• The request URL is POST https://<NAME-OF-YOUR-RESOURCE>.cognitiveservices.azure.cn/translator/text/batch/v1.1/batches.
• The request body is a JSON object named inputs.
• The inputs object contains both sourceURL and targetURL container addresses for your source and target language pairs. With system assigned managed identity, you use a plain Storage Account URL (no SAS or other additions). The format is https://<storage_account_name>.blob.core.chinacloudapi.cn/<container_name>.
• The prefix and suffix fields (optional) are used to filter documents in the container including folders.
• A value for the glossaries field (optional) is applied when the document is being translated.
• The targetUrl for each target language must be unique.

Translate all documents in a container

This sample request body references a source container for all documents to be translated to a target language.

For more information, see request parameters.

{
    "inputs": [
        {
            "source": {
                "sourceUrl": "https://<storage_account_name>.blob.core.chinacloudapi.cn/<source_container_name>"
            },
            "targets": [
                {
                    "targetUrl": "https://<storage_account_name>.blob.core.chinacloudapi.cn/<target_container_name>"
                    "language": "fr"
                }
            ]
        }
    ]
}

Translate a specific document in a container

This sample request body references a single source document to be translated into two target languages.

{
    "inputs": [
        {
            "storageType": "File",
            "source": {
                "sourceUrl": "https://<storage_account_name>.blob.core.chinacloudapi.cn/<source_container_name>/source-english.docx"
            },
            "targets": [
                {
                    "targetUrl": "https://<storage_account_name>.blob.core.chinacloudapi.cn/<target_container_name>/Target-Spanish.docx"
                    "language": "es"
                },
                {
                    "targetUrl": "https://<storage_account_name>.blob.core.chinacloudapi.cn/<target_container_name>/Target-German.docx",
                    "language": "de"
                }
            ]
        }
    ]
}

Translate all documents in a container using a custom glossary

This sample request body references a source container for all documents to be translated to a target language using a glossary.

For more information, see request parameters.

{
    "inputs": [
        {
            "source": {
                "sourceUrl": "https://<storage_account_name>.blob.core.chinacloudapi.cn/<source_container_name>",
                "filter": {
                    "prefix": "myfolder/"
                }
            },
            "targets": [
                {
                    "targetUrl": "https://<storage_account_name>.blob.core.chinacloudapi.cn/<target_container_name>",
                    "language": "es",
                    "glossaries": [
                        {
                            "glossaryUrl": "https://<storage_account_name>.blob.core.chinacloudapi.cn/<glossary_container_name>/en-es.xlf",
                            "format": "xliff"
                        }
                    ]
                }
            ]
        }
    ]
}

Great! You just learned how to enable and use a system-assigned managed identity. With managed identity for Azure Resources and Azure RBAC, you granted Translator specific access rights to your storage resource without including SAS tokens with your HTTP requests.

Next steps