Attending the first-ever Edge Field Day provided me with a fresh perspective on how service consumption on the edge is evolving. New technologies are bringing secure, reliable options to the edge to solve unique challenges. Throughout my career, I’ve worked on a variety of edges, including those that required helicopters or snowcats to reach, telecoms with thousands of sites, and hotels that sail around the world (as well as your typical enterprise with hundreds of locations).

What is the Edge and why should I care about it?

For most of us in IT, the edge was the part of the infrastructure that was closest to your environment’s north or south boundary. Usually this was either the ‘Internet edge’, ‘campus / user edge’ or ‘WAN edge’. In certain industries however, the edge means something wildly different.

Whether you are in manufacturing, telecom, public infrastructure, retail, or other such verticals, your edge may look significantly different. You may have hundreds or thousands of sites, or your site may not be a physical location at all! Each of these has its own unique challenges that traditional IT is not used to designing for. These can include environmental, mobility, connectivity, security and other considerations that are unique from Datacenter and IT closet requirements.

What are the challenges?

Consider that you have a resort style hotel that requires IT systems to support hotel, hospitality, entertainment, and other such functions. Now imagine that the hotel is on a ship, sailing all over the world. All of the sudden your edge has requirements that make waiting on next business day hardware replacement seem quaint. Other industries face different challenges that could include being exposed to a corrosive environment due to chemicals or the elements.

Example (edge) Edge cases:

• How do I power everything I need at the site when there might not be outlets? (hint: you use POE whenever possible)
• When I need to get replacement hardware to the edge ‘site’ and it is in another country for less than a day, what things are required?
• If dial-up seems fast by comparison to my 800+ ms latency connection how do I get things patched?

Where should I start?

The very first thing you need to do when looking at creating an edge design is to focus on the requirements. That list might look different from your typical technology assessment requirements although there may be some crossover. The business questions should come before the technology questions. Starting there will give you a good idea of how the infrastructure will need to be architected.

Try starting with the following questions:

• What is the impact to the business if my technology stack is down (not everything needs 99.999% availability)?
• Which component services must be protected and prioritized?
• Are there any compliance or regulatory controls I need to address?
• Is physical security something I need to take precautions against?
• Are there special requirements for getting on site equipment or personel on site?
• Do I need to be able to recover things locally without any on site support?