The Evolution of the Software-Defined Solution Provider
"Agile" is a poor little word. We've batted it around mercilessly these past several years. It's good to see it coming into its own now that we realize how agile methodology delivers more agility.
Those who've been around long enough will remember the early days of Cisco and its few competitors. You needed to run software to manage these new-fangled routers, but then you kind of became a sysadmin so it was cool. Then the genius idea came: Put that programming into firmware built right into the box. Much more efficient and tighter management, right? Yeah, no so much.
We spent more than a decade correcting some of our earlier errors. We pulled the intelligence back out of routers and switches, even servers and storage devices. We realized that we'd have far more flexibility if the programming were all back where it belonged, on a computer. Yes, just any computer.
Thus we've come full circle, back to what we now call "software-defined networking," which we originally called "networking," and "software-defined storage," which we used to call "storage." We've come so far.
McNealy Was Right
Scott McNealy, founding chairman of Sun Microsystems, often said quotable things. One of his best was, "Those who cannot do, can sort." Loved that one; didn't pick on teachers. But McNealy was perhaps best known for saying, "The network is the computer!" Turns out he was right.
In what may stand as the grandest example of scaling out, the cloud has become a laterally scaled computing engine consisting of a network of smaller computers. We're able to transport commands, data, everything at electron speed from place to place, computer to computer, in order to compute at unheralded speeds with unparalleled flexibility and, yes, major agility.
Of course, we've needed new software technologies to get us there, but software technology is what we do. Our industry has produced SD-WAN for software-defined wide area networking, and there's software-defined storage such as Gluster and Ceph from Red Hat by way of the Open Systems Consortium. Then there's also the big player, microservices in containers, which has completely redefined coding applications.
Welcome, Cloud-Native Software
As with anything else in our industry, we needed to give it a category name, and since it's all meant to run in the cloud, why not call it all "cloud-native" technology? So we did.
Now, this may encourage some of you to have new business cards printed with "CNSP: Cloud-Native Software Provider" under your name. You're certainly welcome to do that, as so many of our "MSP" colleagues have.
But more to the point is the idea of creating a practice that only delivers software-defined networks, completely virtual datacenters in which code moves around as quickly and effortlessly between cloud locations as it does along the internal data bus in a PC. The network itself has become the computer, just as McNealy predicted, and you can provide that environmental computing platform to your customers today.
You offer significant cost-savings as all the network adapters, routers, switches, CSU/DSUs, modems and other devices once again become cheap slaves of the overlording management software, which is at your command. You use commodity storage drives configured in resilient RAID arrays and managed by your software to do whatever you need them to do. Again, costs are slashed as functionality is moved to less expensive hardware and far more capable software.
Control feels good, no?
Redefining Your Role in Your Customer's Compute Strategy
Everyone wants to become more "sticky" to their customers. While that sounds messy, it certainly makes sense to perpetuate your role and your revenue stream. When the network, the storage, the code, even the entire datacenter are defined by the software, and that software is managed by you...well, how much more sticky can you get?
Most every problem is immediately solved with redirection or inexpensive "hot spares." Scaling becomes something the network just does naturally. Support software speaks to supervisory software to solve anomalies. When the software runs the network and you run the software, you also control your customer. Powerful stuff.
Frequent readers of this blog and my other writings may have noticed that everything is grounded in a deeper understanding of the past, which helps us better understand why things are what they are and where they're most likely going. But the most important thing we examine -- which is also the most important thing you must examine -- is the future. Stare hard into the future and decide how you want to get there. Plan for your journey. Be the architect of your own tomorrow.
Posted by Howard M. Cohen on September 08, 2021 at 9:48 AM