Edge Computing with IEF

Intelligent EdgeFabric (IEF) provides you a complete edge computing solution where cloud applications are extended to the edge. By leveraging edge-cloud synergy, you can manage edge nodes and applications remotely while still processing data nearby. In addition, you can perform O&M in the cloud, including edge node monitoring, edge application monitoring, and log collection.

Edge Computing with IEF

Figure 1 shows how IEF implements edge computing.

Enable IEF to manage edge nodes, and bind end devices to the nodes.
Install edge node software on edge nodes, and bind end devices to the nodes. Then you can use IEF to deploy applications on those edge nodes.

For details on how to manage edge nodes and bind end devices, see Node Management and End Device Management.
Develop applications, create images, and upload the images to SoftWare Repository for Container (SWR).
This step ensures that edge nodes can pull images from SWR when IEF delivers applications.

There is no sequential order between this step and step 1. Either of them can be performed first as needed.
Deploy applications.
After the edge nodes are managed and applications are developed, you can use IEF to deploy applications on edge nodes and start running your services.

Running applications can be monitored and alarms can be reported through the Application Operations Management (AOM) service.

For details on how to deploy applications, see Containerized Application Management.
(Optional) Send the data back to the cloud for further processing, and update the applications based on the results.
This step is not closely related to the use of IEF, but it is a common way to optimize applications based on processing data. You can determine whether to perform this step based on your requirements.

For example, if you deploy a facial recognition application on an edge node, you can periodically upload collected data to OBS and add the data to a training set to generate a new and better model. Then, you can package the new model into a new image and update the image to the application deployed on the edge node.