FAQs and Support
Copy link to clipboard

Support

If you have any questions about our product and usage, please review our existing documentation on our Developer Portal or consult the FAQs section below. If you cannot find what you are looking for, please contact our [email protected] or reach out in our community Discord.

FAQs

Here are the most frequently asked questions about EdgeXR's product and offering. If you have questions and don't see them here, please submit them to [email protected] and we will do our best to publish your questions along with answers in a timely fashion. Please do check back periodically as we continuously update our FAQ page.

General usage

What is a Cloudlet?

From a developer's perspective, a cloudlet is a data center hosted by one of EdgeXR's partners, which is typically a Telecom Company like Deutsche Telekom or Telus. These cloudlets are available for developers to deploy their application instances and application instances can be deployed across multiple cloudlets. You can see a list of all cloudlets that are available for deployment within the EdgeXR console under the Cloudlets section.

What is an MCC-MNC?

MCC-MNC is actually two acronyms. MCC stands for Mobile Country Code and MNC stands for Mobile Network Code. Together, both codes create a unique ID for a telecom operator in a specific country or region. This value is very useful as it lets EdgeXR label cloudlets with a MCC-MNC. Additionally, from a user's cellular connection, a client application can get an MCC-MNC ID for their telecom provider. This creates a mapping from user's device to their network carrier, which is one of the requirements to serve the best cloudlet to a user. Here is a full list of MCC-MNC.

What is a Distributed Matching Engine (DME)?

The Distributed Matching Engine (DME) is the first entry point for all client applications that want to connect to one of their deployed application instances. Typically, each DME is associated with an MCC-MNC and can be accessed via https://(mcc-mnc).dme.cloud.edgexr.org:38001. Based on the MCC-MCC, the DME will maintain a list of all application instances that have been deployed to cloudlets with the given MCC-MNC value. There are also regional DMEs that are set up to maintain a list of all application instances for the given region.

In order to use the DME to connect to a deployed application instance, please refer to our Connect Client Applications to Deployed Cloudlets Guide.

What is a Flavor?

A Flavor is a set of compute resources that are provided to your application instances like RAM, vCPUs, Disk Space, and vGPUs. Various cloudlets will support various flavors depending on what compute hardware has been installed. You can see a list of all flavors that are available within the EdgeXR console under the Flavors section.

EdgeXR Edge-Cloud Console

I just created an account on the EdgeXR Edge-Cloud Console, but my account is currently locked. How do I unlock it?

The EdgeXR platform is currently in private beta, which is why we initially lock all newly created accounts. To expedite the process in unlocking your account, please contact [email protected] and send us a small summary of your company and how edge can help your use case.

How do I reset my password?

On the login screen, select Forgot Password?, enter your email address, and then select Send Password Reset Email. You will receive an email from [email protected] with a link to reset your password.

If you have more questions around the account creation process, please take a look at this guide which documents all the features of the EdgeXR Edge-Cloud Console.

What types of deployment tools are supported?

At this time, EdgeXR supports 4 main deployments: VM, Docker, Kubernetes, and Helm. To learn more about deploying each, please refer to our console guide.

What versions of helm, docker, docker-compose, and k8 yaml do you support?

docker: 20.10.12
docker-compose: 1.24.1-mobiledgex1
k8s: 1.18.2
helm: 3.4.2

What are my resource usage metrics?

Monitoring data with the Edge-Cloud Console assumes the application was deployed via the Console, and this data can therefore not be observed outside the Console.

Different metrics are available for the different deployment methods.

Using the EdgeXR SDK to find/connect to a cloudlet will help provide additional metrics including devices connected.

The Alert Event framework that is under development is designed to provide alerts on certain events (low disk, CPU, etc).

How do I determine if there are sufficient resources to deploy my app to this cloudlet?

Currently, overall metrics cannot be retrieved from console or API, and will require trial and error from the user’s part.

What can I do if I run out of resources for a cloudlet?

The primary solution is to reduce usage either in what is being requested in terms of flavor or to remove other applications from the cloudlet.

How do I upgrade/patch my application?

Update the application first, which will then allow you to update the application instances built off that application.

What happens if the underlying server(s) which supports my application faults or fails?

Attempt to redeploy the application, assuming there are available resources.

How do I manually scale out my application on a specific cloudlet?

Manually deploy more instances; consider usage of the autoscale functionality. This will likely be made easier by deployment of the event trigger framework.

Can I autoscale my application on a specific cloudlet?

Use the autoscale functionality built into the product.

What if my application has a stateful component?

Docker volumes are currently supported.

What are the best practice patterns for deploying my application?

Nothing currently.

What are anti-patterns that I need to be aware of so that I can avoid them?

Nothing currently.

Do I need to architect my application to help the DME manage my traffic, or is that all handled by the DME itself?

DME currently runs in two modes: mode one is strictly as a matching engine, and mode two matches based on performance tests. On the client side, you can utilize the EdgeXR SDKs to call the DME APIs.

Can I integrate with my CI/CD process pipeline?

Follow this working example using Github Actions.

What flavors are available on a given cloudlet?

Currently, these are not available via the API, only the console.

This is going to be dependent on the operator in terms of the flavors they are offering. We currently do have an ability to request new flavors which routes through cloud operations.

SDK Usage

What are the currently supported SDK languages?

The EdgeXR SDK is written in 4 frameworks:

If you need to support a different client side framework i.e. web, you can also use our REST APIs.

I am getting an DmeDnsException from the SDK on my device with the message "Generated mcc-mnc.dme.cloud.edgexr.org hostname not found". How should I handle this exception?