Parameterize MongoDB linked service Database Name in ADF

Question

I have many databases in MongoDB (nearly 20). I need to create linked services for all of them in Azure Data Factory. After creating one, I found that linked services for MongoDB are not parameterized for database name. Is there any way to create a parameterized linked service? I can't create 20 of them.

Answer 1

Yes, the MongoDB linked service database name can be parameterized in Azure Data Factory.

For MongoDB/MongoDB Atlas linked services, the database (or databaseName) property supports expressions, so it can be defined as a parameter instead of a fixed value.

High-level steps:

1. Open the MongoDB (or MongoDB Atlas) linked service in ADF.
2. Switch to the Code view (or enable Specify dynamic contents in JSON format under Advanced).
3. Add a parameters section to the linked service JSON, for example:

      {
        "name": "LS_MongoDb",
        "properties": {
          "type": "MongoDbV2",
          "parameters": {
            "DatabaseName": {
              "type": "String"
            }
          },
          "typeProperties": {
            "connectionString": "<your-connection-string>",
            "database": "@{linkedService().DatabaseName}"
          }
        }
      }
   

   For MongoDB Atlas, use type": "MongoDbAtlas" and "database": "@{linkedService().DatabaseName}".
4. Save the linked service.
5. In each pipeline (or dataset) that uses this linked service, supply the appropriate database name value for DatabaseName at runtime.

This way, a single linked service can connect to multiple MongoDB databases by passing different parameter values, avoiding the need to create 20 separate linked services.

---

References:

• Parameterize linked services in Azure Data Factory and Azure Synapse Analytics
• Copy data from or to MongoDB using Azure Data Factory or Synapse Analytics
• Copy data from or to MongoDB Atlas using Azure Data Factory or Synapse Analytics
• MongoDbV2LinkedService interface
• MongoDbLinkedService interface

Answer 2

Hi @Dhruv Singla
Good question - this is a common limitation with MongoDB in ADF.

For MongoDB linked services, the database name is not fully supported as a dynamic parameter in all cases, especially when using the UI or Key Vault setup. So your observation is correct.

Here are the practical options:

Option 1 (recommended): handle it at dataset level

• Keep one linked service (connection only)
• Create datasets per database or parameterize the collection/query
• Pass database-specific values in dataset/activity

Option 2: separate linked services

• If database name must change at connection level, then yes - multiple linked services may be required
• This is the most stable approach today

Option 3 (advanced / try if needed):

• You can try parameterizing via JSON (code view), but support is inconsistent for MongoDB
• Works in some cases, but not guaranteed

Reference:

• Parameterize linked services in Azure Data Factory and Azure Synapse Analytics
• MongoDBLinkedService.DatabaseName Property (allows expressions)
• Create a linked service to MongoDB using UI
• Copy data to or from Azure Cosmos DB for MongoDB (similar pattern)

I hope this helps. Let me know your use case - I can suggest the cleanest design for your pipeline.