5 Things to Consider When Creating Edge Compute Workloads

Edge compute brings the capabilities of cloud close to the end-user or end-device while also addressing the challenges of bandwidth, latency, and resiliency. Moving computing and storage to the edge speeds up the performance, and thus we see many enterprises adopting edge to complement existing cloud infrastructure.

StackPath Edge Computing is a fully managed environment that allows container and Virtual Machine (VM) workloads to run in our global edge locations. With no underlying hardware to manage, you specify the workload requirements such as instance CPU/RAM sizing, where you want to deploy your images, how many instances to run globally, and what disk resources you need. Then specify your image, and we do the rest. StackPath ensures that the number of instances specified per location are always running, manage the reliability of the deployment, and deal with updating the images to the latest version. You need only to focus on what’s running inside the containers or VMs.

Here are the five things you should consider when creating edge compute workloads:

Which Virtual Machine type and size should you select?

Choosing the right VM image for your workload is an important factor for a successful deployment. First, consider the performance that you need for your application and then select the appropriate operating system and size.

StackPath supports a variety of VM images and sizes, including five different specs ranging from 1 core and 2GB of RAM to 8 cores and 32GB of RAM, and operating system images (CentOS, Ubuntu, and Debian) that are optimized and readily available for use. For a detailed list of all available VM image types and sizes, see the Virtual Machine Images support page.

Where are your end-users located?

Though similar, edge VMs differ from cloud VMs by the structure of the network they run on. Edge VMs are located as close as possible to the end, reducing response time and improving end-user experiences. StackPath has 45 edge locations all over the world located next to IXPs to reduce latency. Faster, more flexible and scalable, data reaches the StackPath platform up to 2.6 times faster than central cloud provider data centers and returns data 37 milliseconds faster than competing cloud provider, Google.

The purpose of using edge VMs is to reduce latency. Thus, it is crucial to consider where your users are located. If you have a concentration of users in a specific region, it will benefit you to deploy your application in a point of presence (PoP) closest to that region.

Do you predict a massive rise in traffic?  Consider using autoscaling.

Having too many concurrent users is a good problem to have, but it can be tricky to ensure you have enough resources available when an unexpected spike of traffic occurs. StackPath allows you to auto-scale your workloads across PoP’s so you can serve your users worldwide and ensure that your localized traffic spike is served without any degradation of service.

Learn more about how auto-scaling works and how you to configure this feature when creating your workload.

Will your application be used globally? Consider using anycast.

If your audience is trying to access your web service in several locations around the world, StackPath’s Anycast feature will help you get the best delivery performance to any region by leveraging our private network. It can help provide your audience with a faster, uninterrupted experience as their connections get seamlessly routed to the server closest to them. Using anycast will also help lower administration costs since it will enable web services routes around slow or failed nodes.

Learn more about anycast and how you can add a global anycast IP address to your workload.

How can you save on data transfer fees?

The cost in bandwidth can be tremendous with large cloud service providers when the number of devices transmitting data at the same time grows. Edge-computing hardware and services help solve this problem by being a local source of processing and storage for many of these systems. With StackPath, network traffic pricing is free for any ingress to an instance and free egress from one instance to another within the same location. Take advantage of our distributed edge compute network and save on surprise data transfer costs from large cloud service providers.

Get Started Today

If you’ve never used Virtual Machines before and want to learn more, contact us with any questions and we will be happy to help!