[ad_1]
A new buzz-phrase crossed your correspondent’s office: “Computing First Networkingâ€.
WTF is this?
According to a brief by analyst firm Gartner, it is “an emerging network solution that provides services for compute tasks by sharing computing resources from multiple edge sites” and “a new kind of solution decentralized computing that maximizes the efficiency of edge computing â€.
But why does edge computing need more efficiency? Over the past few years, we’ve been told that Edge solves all of your problems, bringing the compute closer to where the data is created so you can work on that data without having to move it to a cloud or data center. data for processing. Or pay for this schlepping.
Reality Check: Gartner believes that the compute resources at the edge may not have enough capacity to handle all the tasks, so you need to find other resources to do the work. Maybe even other resources on the edge.
At this tricky stage Reg readers will probably have thought that assigning workloads to available resources is exactly the kind of work load balancers can do.
Sorry. Gartner believes that load balancers were not designed to work with edge-y resources because the resources offered change very quickly as containers and serverless are often deployed at the edge. Or on MECs – multiple access edge computing (MEC) sites that house an edge router, servers, power and cooling.
Enter Computing First Networking (CFN), which Gartner’s Owen Chen describes as “a new kind of decentralized computing solution.” [that] Leverages both the state of computing and networking to help determine the optimal edge between multiple edge sites with different geographic locations to meet a specific edge computing demand.
CFN does this using the dynamic anycast (dyncast), which Gartner describes as “a distributed technique that follows the idea of ​​pooling resources to send service requests to the best.” [edge] site dynamically.
“Instead of measuring overall metrics such as the CPU / GPU / memory usage of an MEC site, dyncast relies on a compute module called a demon station to acquire compute state in the granularity of the application. . This allows a computation metric to be calculated that reflects the workload of a certain application deployed at the MEC site.
Gartner offers the example of enabling self-driving cars as an application for CFM.
The company imagines that MEC sites would take care of collecting and processing traffic information so that driverless cars know what to expect on the road, but also have a role of video streaming in cars to entertain their passengers. occupants. Traffic analysis is clearly the most important job a MEC can do, so Gartner imagines CFM would find other resources that are not busy and therefore can be given the job of streaming video.
And that resource will likely be there for the outlet, because in a well-wired city there will be MECs on many 5G base stations. At the end of the working day, when the CBDs and industrial parks are emptying, the CEMs there should be ready to take on work while the CEMs next to the freeways are at their peak.
The CFN appears in articles at wacky conferences, but was not featured in Gartner’s 2021 Hype Cycle for Advanced Computing. So, it’s probably not something you will need to deploy quickly. It is, however, a sign that doing Edge well won’t be as straightforward as the edge-centric servers and software overlays that have been announced every two weeks lately. ®
[ad_2]