This project contains an example of how to populate an index with 2 data sources - 1) Cosmos DB (SQL API), and 2) Azure Blob Storage.
Create the following Azure Services within your Azure subscription:
Set the corresponding service tier ("Free" sku may suffice depending on how much data you may have); accept all defaults.
Create a general purpose, V2 Azure storage account. Set the "Replication" to "Locally redundant storage (LRS)" as there is no need for geo-redundancy for a POC.
Within the storage account, create a container to house your documents.
Create a Cosmos DB collection. Note that each document must contain a property named metadata_storage_path whose value is equivalent to the base64 encoded fully qualified path of its corresponding blob file. This is what links the Cosmos DB data to the blob storage files.
Import the Postman collection.
The collection relies on the following Environment Variables:
| Variable | Description |
|---|---|
| searchAccount | your search account name |
| searchApiVersion | 2019-05-06 |
| searchApiKey | primary API key for search account |
| cosmosDbConnString | copy "Connection String" value from Cosmos DB account. Append the value with ";Database=[database-name]" |
| storageConnString. | storage account connection string value |
Execute the Postman scripts in order starting with "1.0 Create idx-demo Index".
Test the index from the Azure Portal