Over the past couple of years, I’ve traveled to quite a few cool places to talk with IT pros and corporate leaders. This past week, I went to Japan, where I met with company leaders from NEC Global, NTT, Canon, Konica, ASICS and others. In past years, I’ve also met with SoftBank, one of the largest internet service providers in the world. They own Sprint in the United States, as well as other carriers worldwide.
I noticed that all of these companies, especially NEC, NTT and SoftBank, were looking for ways to upgrade the skills of their employees. Each one is ready for the cloud, but in such a fast-moving field – particularly in regard to the internet of things (IoT) – they’re very interested in finding ways to make their workers even more cloud-ready. They also need to upskill their employees so they can support new devices that help manage intellectual property and gather IoT-based data quicker than ever before.
All of these organizations were interested in cloud security. Too often, their partners simply tried to take older network applications and services and simply replicate them on the cloud. The result was even less secure networks that were very efficient at getting the job done, but also in letting hackers enter company systems. A security pro that I know over at ASICS, Shigekazu Tanimoto, told me how it’s vital for companies to change their perspective. They simply can’t afford to migrate the same old practices they’ve been following from their old, installed network services to the cloud. It’s a different world – one that will punish old behaviors, procedures and approaches. That’s not the only problem, though.
Another problem was that as corporations started moving to the cloud, they experienced network latency. Why? Because too many organizations thought they could just purchase a new router or two, as they had before, and enjoy the benefits of the cloud. That just doesn’t work. Today’s networks are as different from those of 10 years ago, as the old Tokyo Tower (left) is different from the Tokyo Skytree (right).
Figure 1: On the left is the Tokyo Tower, taken from the back of the taxi taking me to NEC's headquarters. On the right is the Tokyo Skytree, taken the evening I met with Canon.
I love the old Tokyo Tower, but the Tokyo Skytree was built to help house all the devices necessary to make today’s mobile and cloud computing work in Tokyo. In cloud computing, as in architecture, we need modern approaches to handle issues such as latency, as well as the security problems that keep occurring.
As I talked with company leaders in Japan, I noticed that many of them were discussing edge networking devices that help cloud and IoT devices process information faster. I even saw a confidential document that described some of these devices. They all said they needed individuals who could help connect and manage these devices securely because their customers needed better support.
Several of these companies have also been tasked with creating robust, resilient networks for the 2020 Tokyo Olympics. They then talked about the challenge of having to create these networks quickly, where none had existed before.
Figure 2: Some of the military folks I met at AFCEA.
I found this discussion interesting, because about a week before I was in Tokyo, I heard similar examples from members of the U.S. military. I was lucky enough to talk with Army colonels and techie corporals at the AFCEA CERTS conference in Augusta, Georgia. Each of these folks told me how they were responsible for quickly creating secure, resilient networks in the field. They need a training program on what the cloud and IoT means in 2018. They need workers who understand security, edge networking, virtualization and modern cloud implementations.
Figure 3: The AFCEA certs committee, discussing security and cloud needs.
It was fascinating, because both the corporations in Japan and many U.S. military personnel around the world needed the same thing: individuals who understood how modern cloud computing works with the advent of hyper-mobile devices, IoT and the need for quickly deployed networks.
With IoT and newer cloud approaches, we don’t live in a typical client/server or even Google-like world. We’re further decentralizing/federating our computing and networking capability than even the traditional Amazon cloud model. Yet everyone is interested in workers who can help install, configure and manage devices that help gather and collect IoT data in intermediate stages, as shown in Figure 4.
Figure 4: IoT data, edge networks and core network services
I think a lot of people feel that they still live in a world like the late 1990s, or even early to mid-2000s, where we pretty much followed the client/server model. Sure, that model still exists, just in the same way mainframes are still around – and will be for years. But with today’s IoT-based model, connections don’t just go from several clients through some firewall and then to a server or database. The connections are much more numerous and diverse.
Now that IoT and newer cloud approaches are with us, we don’t live in a typical client/server or even Google-like world. We’re further decentralizing/federating our computing and networking capability than even the traditional Amazon cloud model. Yet everyone is interested in gathering data and collecting it in intermediate stages.
IoT devices are far, far too numerous to use the traditional client/server model. So, before all of that IoT data is passed to a company’s data center or web server, it’s first collected and even partially processed by intermediary, or edge, devices, such as:
- Connection Devices: If we’re lucky, we live in a 4G LTE world. But 5G networks are on their way and will be mesh oriented. Instead of the huge cell towers that we all have seen, 5G connection devices will be far less conspicuous. Many won’t be much bigger or taller than a cupboard in your kitchen. Some will be smaller than that. Others may be a bit bigger, but they certainly won’t be the size and type of a typical 4G LTE tower. These devices, along with their older cell tower counterparts, are responsible for grabbing data from the IoT sensors and devices we use in our daily lives.
- Edge Switches: These gather data from millions of IoT devices and funnel it to intermediary edge devices.
- Microclouds: Traditionally, cloud services have been provided from large, power-hungry, heat-generating data centers, such as Facebook’s in Prineville, Oregon. Traditional data centers will remain necessary, but some typical data center functions are being replaced by much smaller centers that aren’t much larger than a typical server rack. Some are the size of a server closet or even an automobile. Yet, they’re capable of processing data and turning it into information that can companies can use. These devices collect data from IoT devices, partially process it and then forward it to larger data centers and servers.
- App Devices: These units bring essential cloud-based applications closer to the customer. They reduce latency and make customers happier.
- Content Delivery Service Provider (CDSP) Devices: These systems and units help provide faster delivery of streamed content, as well as webpages.
A friend of mine who works for Western Digital uses a great analogy to illustrate today’s cloud. IoT devices are like drops of rain. There are tens of millions of them at any given instant. But eventually those drops of rain gather into puddles, pools, streams and ponds. Each of those streams and ponds has little ecosystems of their own. These puddles, streams and ponds are very similar to the edge devices being deployed today. These edge devices then communicate with larger data centers and companies, which are larger rivers and oceans. It’s a simple analogy, but it helps me start understanding how the industry is creating and deploying edge networks.
The new version of CompTIA Cloud+ (CV0-002) validates the expertise needed to maintain and optimize cloud infrastructure services. Learn more about the new Cloud+, which launches this month.
