August 6, 2021

Migrating to the Cloud: Pocket Guide for Engineering Managers

So, you’ve gotten inundated with information about cloud computing, and now, you’re considering how your organization can get started.

No arguing it, cloud computing is changing the way companies use technology. It’s not only about the dramatic decrease in your capital expense – which is actually important in its own right, but also the impact it has on productivity.

And, by productivity, that covers the output of your developers, and the efficiency of your computing machines. This is revolutionary to how you see DevOps, and other day-to-day engineering operations.

One would expect a company as big as Twitter to enjoy seamless operations, wherever its infrastructure is hosted. But, no. It had to get on Google Cloud to see an increase in speed, flexibility, and scalability.

But just before you make critical decisions about what cloud services you need, or how to even get started, you’ve got a couple of boxes you need to check. And here’s where it starts.

Your Current Infrastructure

Typically, by using an on-premises server, you’ve got your work cut out for you, when it comes to creating and managing technology infrastructure. Now that you might be moving to the cloud, the first step is to take inventory of what powers your operations at this time.

Most likely, you’re running on physical servers, with large computers, and with operating systems installed.

Typical data center architecture

There will also be data centers where your data is stored and managed, with backups there as well. And what about your development environment? Sure containers wouldn’t come up, as they would in virtual servers, but it’s essential to take note of anything currently playing its role on your server.

Picture the current architecture. This creates the foundation for what you might need, or not need, and how the cloud comes into your infrastructure planning.

Isolate Specific Needs: Going all Out, MultiCloud, or Hybrid?

Now that you’ve pictured your current architecture, there’s a little more digging to do, and that’s to isolate areas where you’re having frictions. 

Netflix needed more efficient machine learning models, and that, among a couple of other identified frictions, led the video streaming giant to Google Cloud. So, that area you’ve been looking to improve might be the starting point of your cloud journey.

This might usually require you to have a sit-down with a cloud consultant, to go over your architecture, and suggest appropriate cloud solutions suited for your business. 

The multi-cloud option. Credit: Aviatrix

Sometimes, your cloud adoption won’t be total, meaning you probably have some infrastructure still running on-premises. That’s the hybrid option. You could also run in multiple public clouds at the same time, but it all starts with identifying what you specifically need.

We’ve helped a number of companies find this answer at Vertex, and if you need us to help, you can drop us a line here.

If, when developing new products, or just adding supplementary apps to your solutions, your time-to-market is considerably slow, most cloud providers have virtual development environments that are structured to speed up your developers’ works.

And think about the process of always getting more servers, and expanding your data centers. You could replace those with virtual machines that are deployable in minutes. 

Again on scaling, the higher your business grows with physical servers, the more you’re compelled to replicate servers to cater to these new users. That has to be pre-planned, right? So, what if you got a drastic surge in growth? The resulting infrastructural collapse could cost you those new users.

Target would have a longer tale to tell about this, no thanks to the outages they had in 2019, and another this year – all very likely tied to overwhelmed servers.

Perhaps you’re in the 40% of American businesses, who spend $100K on physical data storage now, and the cost is only growing, while not being totally justifiable. It might be time to put that load on a public cloud. 

Explore Cloud Providers

We’re in August 2021, and as of today, the global cloud industry is worth the north of $350 billion. Providers are doubling down on computing powers, trying to make physical servers less and less desirable.

And that’s exploding because the cloud really does revolutionize computing. Think of how Google Cloud’s Compute Engine could deploy VMs in minutes, helping engineers duplicate the server’s capacity infinitely. 

On GCP again, tools like Cloud Spanner would process transactional and petabytes of data at lightning speed. Plus, a lot more tools for data storage, containerization, machine learning – the list goes on.

Google Cloud infrastructure overview

So, here’s where you look at each cloud provider, so you can determine which has the right solutions for your company’s infrastructural needs. It might also be helpful here that you talk to a cloud service provider for guidance.

Vertex is a Google Cloud partner, and we’ve done multiple projects on AWS, Azure, and a couple more public clouds. You can drop us a line here, and we’d be excited to help. 

On Amazon Web Services, there are IDE and clusters of developer tools, including software development kits (SDKs) for front-end web, mobile apps, and IoT.

Then, you’ve got several developer tools on Azure. Other cloud providers like Cloudflare, IBM, VMWare, and many others would also help you move that heavy capital costs to less operational expenses.

The Planning

Here’s probably the most important part. If you’re moving to the cloud or moving across clouds, there are several variables involved. 

It needs a complete planning process that takes into account how to keep your operations running seamlessly, while you move. Also, you might want to do this in phases. Maybe, over a couple of months, you go from DevOps to core development and have storage at the end. 

Usually, you’d want to start this from where there’s the least amount of risk, even though the backup is there.

But then, providers like Google Cloud already made this easy as well. Like, for data, you could upload to the cloud directly from your data center – as huge as it may be. 

So, that’s about it. The path to the cloud might actually be much easier than you had thought.

Ready to get started? We’d love to discuss how you can do this without any hassle. Send us a message here