Introduction to Azure Resource Graph

Azure cloud is a container of resources.

Your resources can be spread across multiple subscriptions and you may have resources from various resource providers (e.g Disks, VMs, Networks, etc..) 

How can you get an aggregated or nested query result of your azure resources? Yes, you can do it with Azure PowerShell with some effort. Is there any easier and rich alternative.

Azure Resource Graph will come to your rescue.

What is Azure Resource Graph?

• Azure Resource Graph itself is a resource provider
• Resource Graph periodically scan Azure Resource Manager (ARM) for any changes to resources and imports those changes to its internal database
• Scheduled changes of resources are notified to Azure Resource Graph (Change Notification)
• This is a free service (hence a user quota is applied)
• This can be considered as a powerful inventory system for Azure resources.

Okay, What is the usage of Resource Graph?

Inquire information

As I mentioned earlier, this is an inventory of Azure resources. We can query even across multiple subscriptions.

In order to query, you need to navigate to Azure Resource Graph Explorer

We can use Kusto Query Language (KQL) to query.

As you can see there is a link to pin query results to a dashboard

Since this is very powerful, Azure uses Resource Graph to populate information in Azure Portal.

You can load the Graph query by navigating to Open query link

        

Another important feature of Resource Graph is that we can scope our queries to different levels. Following are the options available

• Directory
• Management Group
• Subscription

Once your query is completed, you can save queries in private or shared galleries' and consume whenever necessary.

Change Tracking

Whenever there is a change of a resource, it will be notified to Azure Resource Graph. We can use Application Change Analysis feature of Azure to check the change history of a resource.

Application Change Analysis uses Azure Resource Graph in order to obtain history of changes

  

Take actions based on IOT telemetry with IOT Hub, Event Grid and Logic App

This article is a continuation of my previous Post

With this article I'll explain how to notify users when a metric has reached a threshold value. In this example I will fire an email when the temperature exceeds 25c.

For this example also I'm using Raspberry Pi Azure IoT Online Simulator. (Kudos to the developers !!) 

In order to accomplish the task I need to cover following area

• Configuration of IOT Hub to receive telemetry (Covered in previous post)

• Creation of logic app to filter out relevant telemetry and notify

• Connect IOT hub to Event Grid which uses WebHook to transfer IOT telemetry

Will start with the first part. Following are the steps I used create the logic app

Creation of Logic App

Step 1: Select when a HTTP request is received trigger

Click on Use sample payload to generate the schema link

Since we are expecting the telemetry payload, provide the following. You can get more information Here

{
    "body": [
        {
            "id": "a68df6bc-0fa3-16b1-fbfb-0f00c7677708",
            "topic": "/SUBSCRIPTIONS/AFC2F241-D12F-45C5-9791-72A633F51345/RESOURCEGROUPS/CONTOSO-RESOURCE/PROVIDERS/MICROSOFT.DEVICES/IOTHUBS/CONTOSO-HUB",
            "subject": "devices/iot-test",
            "eventType": "Microsoft.Devices.DeviceTelemetry",
            "data": {
                "properties": {
                    "temperatureAlert": "false"
                },
                "systemProperties": {
                    "iothub-connection-device-id": "iot-test",
                    "iothub-connection-auth-method": "{\"scope\":\"device\",\"type\":\"sas\",\"issuer\":\"iothub\",\"acceptingIpFilterRule\":null}",
                    "iothub-connection-auth-generation-id": "637795588618116304",
                    "iothub-enqueuedtime": "2022-02-05T14:02:38.4880000Z",
                    "iothub-message-source": "Telemetry"
                },
                "body": "eyJtZXNzYWdlSWQiOjkwLCJkZXZpY2VJZCI6IlJhc3BiZXJyeSBQaSBXZWIgQ2xpZW50IiwidGVtcGVyYXR1cmUiOjI2LjY2NDI1NDk5NDIyNzg1NywiaHVtaWRpdHkiOjY0LjQ2NDk5OTA1NjMwMDY3fQ=="
            },
            "dataVersion": "",
            "metadataVersion": "1",
            "eventTime": "2021-02-05T14:02:38.488Z"
        }
    ]
}

Step 2: This payload is retrieved as an array. We need to get the first item from the array in order to process

First we will initialize a variable and set value from the output from our POST Body.

Step 3: Again we will initialize a variable and use previously created variable and get the first element

Step 4: We will use the Parse JSON activity, and parse Body_Param variable. We use following JSON body to generate the schema

{
    "id": "a68df6bc-0fa3-16b1-fbfb-0f00c7677708",
    "topic": "/SUBSCRIPTIONS/AFC2F241-D12F-45C5-9791-72A633F51345/RESOURCEGROUPS/CONTOSO-RESOURCE/PROVIDERS/MICROSOFT.DEVICES/IOTHUBS/CONTOSO-HUB",
    "subject": "devices/iot-test",
    "eventType": "Microsoft.Devices.DeviceTelemetry",
    "data": {
        "properties": {
            "temperatureAlert": "false"
        },
        "systemProperties": {
            "iothub-connection-device-id": "iot-test",
            "iothub-connection-auth-method": "{\"scope\":\"device\",\"type\":\"sas\",\"issuer\":\"iothub\",\"acceptingIpFilterRule\":null}",
            "iothub-connection-auth-generation-id": "637795588618116304",
            "iothub-enqueuedtime": "2022-02-05T14:02:38.4880000Z",
            "iothub-message-source": "Telemetry"
        },
        "body": "eyJtZXNzYWdlSWQiOjkwLCJkZXZpY2VJZCI6IlJhc3BiZXJyeSBQaSBXZWIgQ2xpZW50IiwidGVtcGVyYXR1cmUiOjI2LjY2NDI1NDk5NDIyNzg1NywiaHVtaWRpdHkiOjY0LjQ2NDk5OTA1NjMwMDY3fQ=="
    },
    "dataVersion": "",
    "metadataVersion": "1",
    "eventTime": "2021-02-05T14:02:38.488Z"
}

Step 5: The telemetry field is in "body" field. Now we need to retrieve it. Again we will initialize another variable to hold the field

Step 6: Now we need to work with the telemetry string. Which is also a JSON. In order to do it first initialize a new variable. We need this to hold the later decoded value

Step 7: Our telemetry field is base64 encoded. for an example we get a value like this

"body": "eyJtZXNzYWdlSWQiOjkwLCJkZXZpY2VJZCI6IlJhc3BiZXJyeSBQaSBXZWIgQ2xpZW50IiwidGVtcGVyYXR1cmUiOjI2LjY2NDI1NDk5NDIyNzg1NywiaHVtaWRpdHkiOjY0LjQ2NDk5OTA1NjMwMDY3fQ=="

We can use an online base64 decoder to check

We can use expression to to decode 

Step 8:  Then we will parse the telemetry JSON string.

We can use following to generate the schema

{
  "messageId":90,
  "deviceId":"Raspberry Pi Web Client",
  "temperature":26.664254994227857,
  "humidity":64.46499905630067
}

 

Step 9: Now comes to the interesting part. Now we will evaluate the value of the temperature. If it is larger than 25, we need to send an alert. We use If/Else branch to evaluate the condition

Step 10: We can use an email service to send an email. I use Gmail provider. We can construct the body of the email as we wish

That's all we have to do from Logic App side. Let's save it

Connect IOT Hub to Event Grid and use WebHook to connect to Logic App

Step 1: Navigate to IOT Hub and click on Events. Then click on Event Subscription

Step 2: Create Event Grid Subscription and Select 

• Device Telemetry event type
• WebHook for the endpoint

What would be the url of our endpoint. Easy! We will navigate to our logic app and HTTP request action. Our endpoint URL is there

We will specify that URL for the WebHook

That's it.

Let's try this out

We can check from our logic app overview to see if it succeeded

Great!! Let's see if I got any email alert

That's all. Hope I covered everything here :)