I’m hacking with a customer today who is using Python and needs to upload images to Azure IoT Hub using the File Upload API. It took us a while to figure it out and ran into a bunch of issues.
Here’s how we got it working.
- Create an Azure IoT Hub
- Register an Azure IoT Hub Device.
- Copy the Device’s connection string, you’ll need it later.
After you created your IoT Hub, you need to enable file uploads on the Azure Portal. I’m sure you can do via cli as well, but I don’t have time to research that right now.
1. Open the Azure Portal and navigate to the IoT Hub you created.
2. Click Messaging -> File Upload
3. Create a new or select an existing Standard storage account.
Make sure you create or use a Standard storage account. For some reason, file upload doesn’t work with Premium storage accounts. I’m following up with the team on that.
4. Create a container and click Select.
Make sure you click the Select button to select that container to use in File Upload.
5. On the File Upload blade, Turn on "Receive notifications for uploaded files"
6. Click Save
Make sure you click Save and that you see your container listed in the Storage container box.
If you are running this on a Raspberry PI, then you will likely need to build the SDK locally.
1. Clone the Python SDK
0. Open a command prompt.
1. Navigate to the
2. Open a command prompt and execute the following command:
Replace [device connection string] with the device’s (NOT iotowner) connection string that you copied in step 1 above.
If everything is setup correctly, you will see the following message:
Here’s the meat of the code for this sample.
And here’s the code that calls that method:
If you want to upload files, you need to read them into memory first like this:
Please reference the following article to learn how to handle the File Upload Notification event on the server-side. This is helpful when you want to kick-off a backend process when a file has been uploaded.
If you see the following error…you will need to open up the sample file in your IDE and add a comma after ‘size’ like this:
1. You don’t have a storage account associated with your IoT Hub.
2. You are using a “Premium Storage” Account. Solution: Change your storage account to Standard Storage.
3. You didn’t click Save on the File Upload Blob. Make sure the File Upload pane has your Storage container listed.
This is due to a bug in the Azure IoT Python SDK. The issue is being tracked on GitHub here. In the meantime, you can use the sample code found here: https://github.com/jongio/azure-iot-rest. You will find a Python sample that shows you how to upload files via the Azure IoT REST APIs.
4. ImportError: libboost_python-py27.so.1.55.0: cannot open shared object file: No such file or directory
This is due to the SDK being built with references that you don’t currently have on your machine. Follow the instructions here to build the SDK on the machine you will be running it on. I got it work on a Raspberry Pi 3 / Stretch.
Make sure you use the --recursive flag when you clone.
You can use Azure Storage Explorer to view the uploaded file.
If you go to the storage account that you associated with IoT Hub in Azure Portal, you’ll see an Open in Storage Explorer button.
Drill down to your container and you’ll see your file.
If you want a direct URL to the file it will look like this:
You’ll need to replace your storage name, container name and then build the rest of the path based on the parameters you passed to the upload method when you uploaded the file.
Hope this works for you. Let me know if you run into any issues.