How fast did we get here? It wasn’t that long ago that every enterprise looked at cloud as an amorphous “thing to do in the future.” And in the blink of an eye, every enterprise is racing to the cloud.
Even by tech standards, the adoption rate is incredible. Sure, the luddites out there inevitably talk about lack of security, lack of control, lack of visibility, and yes, lack of ease for the infrastructure guys. While that implies that on-premise applications are more secure, more documented, and have more visibility, a recent spate of spectacular outages tells a different story. But, on that last point, the lack of ease part…they may have something. Let’s dig into that. Let’s talk about unleashing the power of the cloud.
It’s time to learn from the server/app guys. Puppet, Chef, and Ansible—or any number of open-source alternatives—of the world made it easy for server and application guys to deploy hundreds and thousands of servers with a push of a button. But what about us networking guys? Are we still banging away on keyboards? Using CLI? How can we compete with scripted, regimented, and automated processes? SDN is, well, Still Doing Nothing for enterprise customers. But there is hope and it’s called SD-WAN. We’ll come back to that topic.
Visibility is a key component to ensure we’re getting the performance expected from cloud and I would argue that if it’s baked in from the get-go, it’s easier to have visibility. This is if you collect telemetry at the user, as it traverse the network, enters the cloud, and within the app in the cloud. If you can do that automatically and programmatically, and without human latency, then end-to-end visibility is a foregone conclusion. This is where detail matters. You cannot measure this just from on-premise. You must measure this from the cloud and report back to an on-premise monitoring solution. You must measure the application performance at scale and with high-definition resolution. Averaging or sampling transaction times will often hide the problem, never mind finding the problem. Check out this awesome session at Velocity from Jon Hodgson to see what I mean. Therefore, it is critical to end the tyranny of silos. Cloud is the forcing function that removes silo’d workflows. Desktop, network, system administrators and application folks have to come together to deliver the application on time, every time and at scale. This is what cloud is all about. Cloud is what enables infrastructure teams to think and act like a P&L and not a cost center. Infrastructure is no longer a sunk cost. When cloud is done right, infrastructure can be just as agile and profitable as any P&L.
But, that agility is not possible without SD-WAN. How hard is it right now to send users at the branch direct to Internet for trusted sites? Trusted sites and apps like Office 365, Microsoft Azure, salesforce.com, and Amazon Web Services? Why back haul users if you can securely send them straight to the cloud? Turns out, with the right SD-WAN solution, you can do just that—and with the click of a button. But more importantly, the right SD-WAN solution will allow you to collect telemetry as the application leaves the premise. It will report telemetry from the cloud. It will overcome the added latency that cloud can bring to the equation. I would argue that with the right SD-WAN solution, you’ll have better visibility than ever. I’ll cover this in a follow up blog. So make that leap of faith to the cloud—I’ll have the safety harness ready for you as you make that leap.