This repository contains everthing you need to deploy Thingsboard on-top of Nutanix Karbon Platform Services.

This deployment will let you quickly get up and runing with both Thingsboard and Karbon Platform Services. The deployment consists of two parts, Thingsboard deployment and a Kafka to Thingsboard bridge.

This step assumes you already have acces to Karbon Platform Services. If not and you want to try it out sign up at https://my.nutanix.com

Sign in to Karbon Platform Services portal. Open the Projects section in the menu on the left. Click on Create. Give your project a name and assign a user.
Click Next. Assign a service domain using Add service domain.
Click Next.
Enable services for our project.
This project is making use of

• Kubernetes Apps
• Functions and Data Pipelines
• Ingress Controller - Traefik
• Data Streaming - Kafka, NATS

Click on Create.

Next we add a new data source. This is where the data enters the platform solution. We will create a new category for our incoming data. Open the Categories section in the menu and click Create.
Give the new category a name and click on Add value.
In my example I use IncommingData as name and add a value of weather

Once we have our category setup we can add the data source.

Open the Data sources and IoT sensors section of the menu. Click on Add Data source.
Select sensor, Give the data source a Name, a associated Service domain, select MQTTas protocol. Click Generate Certificates to generate .X509 certificates for device authentication. Download the certificates before clicking Next. In the next step, add the MQTT Topic your sensor publishes it's data on.

Give the field a name. In my example my sensor publishes on the MQTT topic weather\data and I name the field weatherData.

Click on Next.
Assign the incoming data to the category we created earlier in this step.

Click on Create.

Next step is to connect our data source to our kafka queue using a Data pipeline.
Select the project we created in step 1 from the drop-down at the top of the menu on the left. Click on Functions and Data Pipelines in the menu and select the Data Pipelines tab.
Click on Create

Give the Data pipeline a Name
Under Input select Data source and then IncommingData and weather category from step 2.
Under Transformation select None. Under Output select Publish to service domain,
select Kafka as Endpoint type and enter data as Endpoint Name. This will create a Kafka topic called data automatically.

It is important to name the endpoint data as this is referenced in the kafka consumer.

Click on Create down in the right corner.

The next and final step is to deploy the kubernetes applications.

Select Kubernetes Apps from the menu. Click on Create.

Enter a name for our application.
Select your Service Domain. Click on Next down in the right corner.

Paste or upload the thingsboard.yaml file in the code area.

Click on Create down in the right corner.

This deployment will now deploy a single instance of Thingsboard using the image thingsboard/tb-postgres. This image includes a postgres database. The deployment creates a Persistent volume claim to persist data to the Service Domain. A ingress route is created that points to the Thingsboard instance. Once the deployment is healthy you can access Thingsboard using http:https://IPTOYOURSERVICEDOMAIN/

This is a simple python application that consumes data from the Kafka service and inputs the data to Thingsboard using Thingsboards REST Api. It is available as an image (voxic/kttb) on dockerhub.

Before we can deploy the kafka consumer we need to set up a user in Thingsboard for API autentication.

Login to you Thingsboard instance (using http:https://iptoyouservicedomain). Defult username and password is:
User: [email protected]
Password: sysadmin

Create Tenant

Click on Tenants

Click on + out to the right. Fill in tenant information and click on Add.

Creat user

Click on Manage Tenant Admins on the right side

Click on + out to the right.
Fill in user information, make sure to leave the option Activation method to Display activation link.
Click on Add.
Copy the Activation link Open a New Browser window or a Incognito tab and paste your activation link. Change localhost:8080 to the IP of your Service Domain.

Remove the :8080 part as well.

Create a password for the user.

Create device
You should now be logged in as the Tenant Administrator that you just created. Last step is to create a device in Thingsboard. Click on Devices in the menu on the left.
Click on + out to the right and then Add new device.
Name your device and click on Add down in the right corner.

We are now ready to deploy the Kafka consumer and start pushing data to ThingsBoard.

Head to the Karbon Platform Services portal. Go to you project created in step 1 using the drop down menu on the left.

Select Kubernetes Apps from the menu. Click on Create.

Enter a name for our application.
Select your Service Domain. Click on Next down in the right corner.

Paste or upload the kafkaToTB.yaml file in the code area. Scroll down the last section of the yaml and change the rows:

            - name: TB_USER
              value: tbusername <---
            - name: TB_PASSWORD
              value: tbpassword <---

To the user credentials you created.

Click on Create down in the right corner.

The solution is now ready to accept data from a sensor.

The Kafka consumer expects data to arrive as a JSON object with the following structure

{
    "device":"dev001",
    "measurement":"temperature",
    "value":"0.3",
    "ts":1614801607246
}

If you don't have acces to sensors or other data sources I have created this simulator https://github.com/voxic/weatherstation-simulation

To send data to the Service domain, publish data to the MQTT broker running on the Service Domain and the configured topic in Step 2. Autentication to the MQTT broker is done using the Certificates downloaded in Step 2.