Edge device IoT deployment and maintenance

Edge devices, connected devices, IoT devices and embedded Linux are all referred to an embedded system running inside a product that is connected to the internet and responsible for delivering the mission.

To deploy and maintain a fleet of edge devices from any kind, there are a few best practices that may help us avoid product recalls and customers angry.

Before sharing a few simple tips that would make your deployment to be easy much safer, other questions may come up:

What is really mean to deploy edge device fleet to production?

Edge devices should build and design to survive. As we may need to deploy and maintain thousands or even millions of edge devices from any kind – embedded Linux, RTOS or any other IoT computer, you wouldn’t like to recall some of them or to find your self supporting each one of them on a daily basis. For that reason, edge devices build to survive and need to have the correct mechanism and technologies for reducing possible issues.

Why maintaining edge devices remotely is so important?

So, with the understatement of what can happen to the product fleet, and the future risks that may need to be solved, We can guess how important it may be to have the ability to manage and maintain IoT edge devices remotely. Having the ability to deploy software updates, patches, and monitor all fleet can ensure a stable and safer deployment that will keep our project successful.

Here are a few tips that will make your edge device product deployment and maintenance much better:

1. Manage edge devices in groups – Thousands of edge devices running all over the world in different environments and areas can be very hard to maintain. Groups can help us arrange the product fleet by different deployment status and manage them in accordance.

Here is a great example of product fleet possible arrangement that will make all much easier to maintain, 3 Groups:

Test group – include identical edge devices like the ones in production.

Production group – includes part of the IoT edge devices that already in the field, and

Beta group – small production device fleet group that receives new upgrades before the production group and after running the upgrade on the test group.

2. Before going to production, make everything run as a service – Service detention may depend on the operating system and the specific software the edge device is running, but, the important basic thing to pay attention here is to have a separate entity which will be responsible to re-run our product application in any case of software error or crash. In embedded Linux based edge devices, we can achieve that using systemd service manager. Here is a great guide explaining how to move your application to systemd service easily.

3. Keep your IoT edge device product 100% available remotely – Nothing can ensure us that our application will run without software bugs or product issues that we will have to fix as soon as possible. To fully be confident with moving to the production stage, the edge device will need to have a separate service which responsible for keeping our edge devices completely available over the internet and in any situation.

Here at JFrog Connect, we provide an IoT edge device management and maintenance platform that will bring you back the confidence in deploying thousands of edge devices to production, to the field, far from human hands.