Do you remember how your organization started using the cloud or why? Do you remember the meeting where it was discussed, who was assigned to lead the migration strategy, and what was in the detailed plan they produced?
Of course you don’t, because it didn’t happen like that. It started with a request to IT for connecting already deployed cloud workloads into the corporate network. “How long? HOW long?”
Now your organization has 50 different applications deployed in the cloud across multiple regions – and that’s just the ones IT knows about. Fire up a VPC in AWS, deploy an application, let’s say a better solution for stock inventory in your company’s stores. Everyone loves it and the CIO agrees that it should be rolled out across the entire business.
“Great work, Carla, but you’re going to need to talk to SecOps and the networking team.”
No problem, we’ll build a VPN tunnel to the cloud. This stuff is easy. Cloud is great.
We’re going to need some more instances of the app. Fire up a few more VPCs.
“Hey, networking! We need some connectivity over here.”
Which is fine, until colleagues on the other side of the country start asking for the app. We’re now three thousand miles from the data center. Maybe we need to think about a colocation facility to handle the west coast operation. Come to think of it, why are we running all these connections out of data centers at all? Wouldn’t it make more sense to put the firewalls and load balancers in the cloud too? Then we could just have one big pipeline to the cloud.
Now we want to bring our far east operation on board. Their preferred provider for cloud is Azure. We can do the same thing with them, right?
“Hey, networking! We’re going to need more connectivity.”
They say it’s going to take a while to figure out. Apparently Azure does VPN a bit differently. They’re not sure how easy it’s going to be to insert firewalls. We might need different tools to manage both environments. Visibility could be a problem. It could take a while. Networking has promised to get back to us.
Now take the scenario we’ve just described and repeat it for your next application or cloud provider and the one after that. You begin to see how cloud journeys that start in hope end in despair, with resources everywhere and no uniform way of managing them.
The common threads running through every organization’s experience are that networking was always an afterthought, and that development teams underestimated the complexity of the environment when applications started to scale and proliferate.
The good news is that while this experience is common, it’s not inevitable!