blob-trigger-cosmosdb-out-binding

blob-trigger-cosmosdb-out-binding (Python)

Sample	Description	Trigger	In Bindings	Out Bindings
cosmosdb-trigger-cosmosdb-in-binding	Azure Functions Blob Storage Trigger Python Sample. The function gets image data from Azure Blob Trigger, gets tags for the image with Computer Vision API (Azure Cognitive Services), and store the tags into Azure Cosmos DB by leveraging CosmosDB output binding	Blob Storage	NONE	CosmosDB

• blob-trigger-cosmosdb-out-binding (Python)
  • Configurations
    • Create Computer Vision resource
    • Create Blob Storage account & Container
    • Create Cosmos DB Account and DB & Collection
  • How to develop and publish the function
    • Local development
    • Publish the function to the cloud
  • Test Request
    • HTTP Request body format
  • Response body format

Configurations

As specified in functions.json, you need Azure Storage account for triggering functions and Cosmos DB Account to store data using Cosmos DB output binding

{
  "scriptFile": "__init__.py",
  "bindings": [
    {
      "name": "myblob",
      "type": "blobTrigger",
      "direction": "in",
      "path": "upload-images/{name}",
      "connection": "MyStorageConnectionString"
    },
    {
      "direction": "out",
      "type": "cosmosDB",
      "name": "doc",
      "databaseName": "testdb",
      "collectionName": "testcol01",
      "leaseCollectionName": "leases",
      "createLeaseCollectionIfNotExists": true,
      "connectionStringSetting": "MyCosmosDBConnectionString",
      "createIfNotExists": true
    }
  ]
}

Create Computer Vision resource

First, create a Computer Vision resource

COGNITIVE_RESOURCE_GROUP="rg_cognitive_test"
REGION="eastus"
COGNITIVE_ACCOUNT_NAME="mycompvision001"

echo "Create Resource Group: $COGNITIVE_RESOURCE_GROUP"
az group create --name $COGNITIVE_RESOURCE_GROUP --location $REGION

echo "Create Cognitive Resource for Computer Vision: $COGNITIVE_ACCOUNT_NAME"
az cognitiveservices account create \
  -n $COGNITIVE_ACCOUNT_NAME \
  -g $COGNITIVE_RESOURCE_GROUP \
  --kind ComputerVision \
  --sku S1 \
  -l $REGION \
  --yes

Then, Get Computer Vision API Key and endpoint. You'll use the values in later step:

COMPUTER_VISION_API_ENDPOINT=$(az cognitiveservices account show -n $COGNITIVE_ACCOUNT_NAME -g $COGNITIVE_RESOURCE_GROUP --output tsv |awk '{print $1}')
COMPUTER_VISION_API_KEY=$(az cognitiveservices account keys list -n $COGNITIVE_ACCOUNT_NAME -g $COGNITIVE_RESOURCE_GROUP --output tsv |awk '{print $1}')
echo "API Endpoint: $COMPUTER_VISION_API_ENDPOINT"
echo "API KEY: $COMPUTER_VISION_API_KEY"

Create Blob Storage account & Container

Create an Azure Storage Account

### RESOURCE_GROUP="rg-testfunctions"

RESOURCE_GROUP=$COGNITIVE_RESOURCE_GROUP 
REGION="eastus"
### STORAGE_ACCOUNT="teststore"
STORAGE_ACCOUNT="mjsteststore"
az storage account create --name $STORAGE_ACCOUNT \
    --location $REGION \
    --resource-group $RESOURCE_GROUP \
    --sku Standard_LRS

Create a container in the storage you've created

# Get Storage Key
ACCESS_KEY=$(az storage account keys list --account-name $STORAGE_ACCOUNT --resource-group $RESOURCE_GROUP --output tsv |head -1 | awk '{print $3}')

az storage container create  \
    --name "upload-images" \
    --account-name $STORAGE_ACCOUNT \
    --account-key $ACCESS_KEY

Create Cosmos DB Account and DB & Collection

Create a Cosmos DB Account

COSMOSDB_ACCOUNT_NAME="mjsazfuncv2db"
#RESOURCE_GROUP="RG-azfuncv2"
RESOURCE_GROUP=$COGNITIVE_RESOURCE_GROUP 
az cosmosdb create \
    --name $COSMOSDB_ACCOUNT_NAME \
    --kind GlobalDocumentDB \
    --resource-group $RESOURCE_GROUP

Create Database and Collection in the Cosmos DB that you've created

# Get Key
COSMOSDB_KEY=$(az cosmosdb list-keys --name $COSMOSDB_ACCOUNT_NAME --resource-group $RESOURCE_GROUP --output tsv |awk '{print $1}')

# Create Database
DATABASE_NAME="testdb"
az cosmosdb database create \
    --name $COSMOSDB_ACCOUNT_NAME \
    --db-name $DATABASE_NAME \
    --key $COSMOSDB_KEY \
    --resource-group $RESOURCE_GROUP

# Create a container with a partition key and provision 400 RU/s throughput.
COLLECTION_NAME="testcol01"
az cosmosdb collection create \
    --resource-group $RESOURCE_GROUP \
    --collection-name $COLLECTION_NAME \
    --name $COSMOSDB_ACCOUNT_NAME \
    --db-name $DATABASE_NAME \
    --partition-key-path /name \
    --throughput 400

# Create a container for leaves
# 'leaves' need to be a single collection partition
# Please see also: https://github.com/Azure/azure-functions-core-tools/issues/930
LEASES_COLLECTION_NAME="leases"
az cosmosdb collection create \
    --resource-group $RESOURCE_GROUP \
    --collection-name $LEASES_COLLECTION_NAME \
    --name $COSMOSDB_ACCOUNT_NAME \
    --db-name $DATABASE_NAME \
    --throughput 400

How to develop and publish the function

Local development

func host start

Publish the function to the cloud

Publish the function to the cloud

FUNCTION_APP_NAME="MyFunctionApp"
func azure functionapp publish $FUNCTION_APP_NAME --build-native-deps --no-bundler

Add Functions App Settings

FUNCTION_STORAGE_CONNECTION="*************"
COSMOS_DB_CONNECTION="***************"
az webapp config appsettings set \
  -n $FUNCTION_APP_NAME \
  -g $RESOURCE_GROUP \
  --settings \
    ComputerVisionSubscription=$COMPUTER_VISION_API_KEY \
    ComputerVisionApiEndpoint=$COMPUTER_VISION_API_ENDPOINT \
    MyStorageConnectionString=$FUNCTION_STORAGE_CONNECTION \
    MyCosmosDBConnectionString=$COSMOS_DB_CONNECTION

Test Request

HTTP Request body format

HTTP Request body must include the following parameters:

{
    'permission': '<Signed permission for shared access signature (Required)>',
    'container': '<Container name to access (Required)>',
    'blobname': '<Blob object name to access (Optional)>'
    'ttl': '<Token time to live period in hours. 1hour by default (Optional)>'
}

The following values can be used for permissions: a (Add), r (Read), w (Write), d (Delete), l (List) Concatenate multiple permissions, such as rwa = Read, Write, Add

Sample Request Body

 {
    'permission': "rl",
    'container': "functiontest",
    'blobname': "sample.png"
    'ttl': 2
 }

Response body format

HTTP response body format is:

{
    'token': '<Shared Access Signature Token string>',
    'url' :  '<SAS resource URI>'
}

Sample Response Body

{'url': 'https://testfunction.blob.core.windows.net/functiontest/yokawasa.png?sig=sXBjML1Fpk9UnTBtajo05ZTFSk0LWFGvARZ6WlVcAog%3D&srt=o&ss=b&spr=https&sp=rl&sv=2016-05-31&se=2017-07-01T00%3A21%3A38Z&st=2017-07-01T23%3A16%3A38Z', 'token': 'sig=sXBjML1Fpk9UnTBtajo05ZTFSk0LWFGvARZ6WlVcAog%3D&srt=o&ss=b&spr=https&sp=rl&sv=2016-05-31&se=2017-07-01T00%3A21%3A38Z&st=2017-07-01T23%3A16%3A38Z'}