Cloud migration has been a buzzword around the tech and business communities for well over a decade now, with most of the emphasis being on how companies can move their architectures from on-premise environments to public cloud providers like Amazon Web Services (AWS), Google Cloud Platform (GCP), Microsoft Azure, Alicloud and others.
Yet with all of these cloud providers to choose from, what about the other aspect of cloud migration: moving or expanding your architecture from one cloud platform to another (aka cross-cloud migration or cross-cloud adoption)? There are a number of reasons why onboarding a new cloud provider can be beneficial to your business efforts, including improved resiliency, better regional coverage or vendor-specific services that better fit your development needs or provide cost optimization opportunities.
Whatever the reason for migrating between clouds, one of an organization's key goals during this initiative is controlling costs and ensuring that it gets the most value out of the money spent. When doing so, it’s important to distinguish between costs that needlessly drive up your cloud bill, and ones that increase your spend while better enabling your business to meet its objectives. For example, migrating some workloads to containers or serverless architecture may incur some short term cost increases, but in the long run it’s worth it if it leads to more efficient product development.
With that in mind, let’s look at some of the key ways that organizations can optimize their spend when migrating between public clouds.
Operational efficiency
It goes without saying that the more efficient your internal processes are, the better it is for your business’s bottom line. But how do you ensure efficiency when dealing with cross-cloud migration? The only real way is by doing your homework when it comes to the new cloud provider that you’re adopting. Determine how easy it will be to replicate workloads and services on the new infrastructure. Will your Ops or Engineering teams have to rearchitect all of your existing components, or will they be able to “lift and shift” (i.e. migrate them with minimal reconfiguration)?
Another thing to consider is how easily you’re able to avoid vendor lock-in. Evaluate your DevOps stack to determine which of your current services are vendor-agnostic, meaning that they can integrate with your new cloud provider’s infrastructure without requiring any meaningful changes. And if any of them can’t, you should consider either switching to a similar service that’s able to work across multiple cloud providers (also making any future migrations that much easier), or keeping those services on your original cloud provider and embracing a multi-cloud strategy.
Vadim Solovey is Co-Founder & Chief Technical Officer of DoiT International.
Technical expertise
Every public cloud has nuances and technical differences that can play a big role in how your environment gets configured and operated. Chances are that your personnel are already well-versed at working on your existing cloud provider’s architecture, but how will they be able to adapt to the new one? Before you commit to a new provider, you’ll want to determine how much training your teams will require in the new system and how long it will take (spoiler alert: there’s a good chance that these efforts will be more challenging and time-consuming than you think). It may also be necessary to hire additional people who have the necessary expertise.
Here is where working with a partner can be extremely beneficial, particularly when it comes to shortening the time to get migrated and fully operational on the new cloud. Not only should a partner have the necessary expertise and support capabilities to help you along the way, but they can also serve as an extension of your team throughout your relationship with the vendor, thus saving you the time and effort to hire, onboard and train new experts on your own payroll.
In addition to the technical support that a partner can offer, they can also be a huge asset when negotiating with the cloud providers due to the extensive relationship that they should already have with them.
Built-in cost savings
Many cloud providers offer built-in cost optimization functions, but taking advantage of them isn’t always easy, nor is it the same across different vendors. Commitment discounts are a prime example of this – each cloud provider has different types of commitments with varying levels of flexibility and management challenges. Some allow you to change machine types or regions without any issue, while others are inflexible and thus carry more risk. Some providers also offer a marketplace where you can sell your unused commitments to other users as a way to recoup sunk costs, but with no guarantee that you’ll be able to offload them.
Here again we see the benefit of working with an experienced partner to help manage your cost optimization efforts beyond just negotiating the deal on your behalf. The best partners can also provide software capabilities that automate things like commitment management, spot instances and anomaly detection to ensure that your monthly costs aren’t spiraling out of control. A great partner will also have vendor-specific solutions that help optimize your use of things like databases and storage, thus allowing your team to focus less on managing your cloud infrastructure and more building a great product that will actually grow your business.
