Just like a horse and cart was an adequate form of transport back in the 19th century – albeit tediously slow and rather dangerous – as time went on, we developed a need for speed, reliability, and more storage from the vehicles transporting us around.
This transition from clunky horse and cart to virtually silent electric vehicles is really no different from the transition currently taking place in the data center industry, although this transition is happening decidedly quicker.
We are now producing data at an insatiable rate, which rather than a snowball effect, is now more akin to an avalanche at this juncture.
Due to said avalanche, more devices will (and are already starting to) send data to the ‘Edge’ and the cloud. Yet, for this to work, we’ll need to rely on more than one type of approach, i.e., a hybrid one, to make it happen.
This reliance we have on constant connectivity and communication is not only boosting the amount of data being generated but, let’s not forget, this data must also be communicated and stored, in an oft repeated cycle.
Combined with the proliferation of 5G technology (throwing bidirectional traffic into the mix) and an influx in remote working, when looking to the future, could the traditional enterprise data center have had its day?
We spoke to Shad Sechrist, data center solutions engineer at Belden, to find out whether hybrid Edge could be the answer.
Traditional Edge vs hybrid Edge
Hybrid Edge is a term many people see differently. Comparing the two, traditional Edge is generally company owned; it might be sitting in a colocation facility or somewhere else, reporting back to a typical centralized data center location.
To achieve faster response times, Edge computing essentially repackages traditional, on-premises computing and moves latency sensitive applications (like multimedia streaming and telehealth) closer to the people and devices using them.
For the Edge to fully flourish however, it needs a hybrid Edge infrastructure solution to help it along, a literal support network if you will. This layer 2 support from the cloud (both public and private) marries together with Edge computing to form hybrid Edge data centers, which need interconnections with each carrier possible, allowing the site to fulfil all applications, no matter where they are.
Hybrid Edge can be tricky to get your head around, so take this analogy. Imagine hybrid Edge computing as the hub and spokes of a wheel. Although each spoke is in a different location around the outside of the wheel (or, the ‘Edge’) each one connects to the same location.
If you consider all the spokes to be Edge data centers sitting closer to end-users, hybrid computing allows you to use third party services to connect the components of the Edge to the wheel’s other
components – this creates a ‘mesh’ network instead of a hub-and-spoke network.
This mesh network means hybrid Edge facilities can deliver latency sensitive data close to users and integrate with core or centralized applications in public clouds or corporate data centers. And instead of only supporting outward-bound data, they also support bidirectional data, a key part of 5G proliferation.
“In my view, there are two ‘flavors’ of the Edge,” says Sechrist. “One of them, I call the ‘metro version’, which is stuff usually sitting at the bottom of a 5G tower that needs to be really close to the tower, and really close to the point of mobility where the information is coming in.”
“Then there’s hybrid Edge. That’s the Edge solution with compute storage and networking, that’s the cloud solutions that might be utilized, as well as myriad other solutions like backup, storage, and compute power, which can all be in different locations but still serving the same purpose.”
Keeping your options open
Where a traditional Edge infrastructure solution might be company owned, the good news is, a hybrid Edge infrastructure doesn’t have to be, and better still, it provides a number of options.
Hybrid Edge opens up a wider scope of tools managed by third parties, whose expertise is to manage these platforms. Not all companies have the talent to manage such a complex usage solution, but with hybrid Edge, it’s a veritable pick and mix of the skills and knowledge you need. For example, storage could be something different i.e Wasabi or NetApp, providing you with much needed flexibility.
By putting all this information in lots of different places, it provides the freedom to utilize different resources to get the job done, rather than the onus being placed on owning all the equipment – an arguably unnecessary expense, not to mention responsibility.
What is driving the Edge?
In short, we are. Much like when shopping for a new car, or even a new internet provider, whichever vendor can deliver us what we want the fastest, wins our custom, and it's exactly the same concept as an end-user. Whoever can satisfy our insatiable demand for data the quickest, will ultimately come out on top.
It’s also worth noting that hybrid Edge data centers tend to be ‘lights out’ or unmanned, so a great deal of this data delivery is happening via machine-to-machine communication, which also drives the need for a hybrid Edge network.
And thanks to an increase in remote working, online streaming apps, and an ever-growing portion of our daily lives being lived out online, we’ve all grown accustomed to a unique brand of instant gratification that has seeped into all aspects of our lives, for better or worse. And if companies can’t keep up, we’re not interested.
“Think about Netflix,” says Sechrist. “I have a seven-year-old daughter, she’ll be watching Netflix, all of a sudden, it buffers and she flips out, ‘it’s not working, I hate this thing.’ And I have to explain to her it’s not your device, it’s the network, or Netflix that is having the issue.”
“Online gaming is another great example of this need for speed. If you’re competing against someone else and you’re even a fraction slower than the next guy, it’s not going to work out well for you.”
And it’s not just entertainment where Edge computing can give us the upper hand. Using Wall Street trading as an example, surely those closer to the network will have an unfair advantage over those that aren’t? Which in a Wall Street environment, could equate to millions of dollars.
Sechrist explains, “To make sure everything is equal and that everyone has the same amount of time, the closer you are to the network, you’ll have more cable curled on top of your cabinet for access, so everyone has exactly the same run life of connectivity, no matter where they are in the data center. So, there is no advantage from a time standpoint for any one person depending upon location to the data center.”
Even gaming companies have had to artificially introduce latency into their software, to ensure a more balanced and consistent experience for all.
And beyond the use cases stated above, the Edge can be responsible for people’s lives. Take autonomous vehicles for instance. The self-driving portion of the vehicle will of course be managed by on board computing, but that data still needs to be transmitted to a data center and back to the vehicle. Imagine if that information was being sent to a traditional data center who knows what distance from the source; the vehicle is told to turn, but that information arrives two seconds too slow, and now it’s too late.
It all comes back to the fact we move things to the Edge, in essence, to get something from A to B that little bit quicker. Bringing that compute, storage, and networking closer to the user,
ultimately giving us a much better, safer, more streamlined experience.
Customers want what they want, and they want it now (if not sooner) and they don’t care how you get them there. Most people are blissfully unaware that the instant gratification to which we’ve grown so accustomed, is only made possible via Edge computing. Ultimately, most of us are living life on the Edge without even realizing it.
For more information, check out Belden's whitepaper, 'Planning now for the future: Hybrid Edge data centers.'