Multi-Cloud: Worst Practice or the Future of Public Cloud?

Corey Quinn, cloud economist at The Duckbill Group, recently argued that multi-cloud is "the worst practice to be avoided by default". Not everyone agrees.

The author of the Last Week in AWS newsletter defines multi-cloud as running the same workload across multiple cloud providers in an agnostic way. He suggests that the approach is promoted by smaller cloud providers that cannot have a significant market share and third-party vendors that build tools for multiple cloud providers. The key issue of multi-cloud approaches is the lowest common denominator:

If you treat all of those environments as being the same thing, that means that every additional service that is any higher-order than those baseline primitive offerings is closed to you.

Multi-cloud can as well limit negotiating leverage for enterprises and he explains that the cost of managing multiple providers is often underestimated:

Before you go whole hog into a second cloud provider, first spin up an active-active environment across two regions in your current provider. With complete service and API compatibility between those regions, it should, by your theory, be a piece of cake. Come back and talk to me after you have done that, and we will see how "simple and straightforward" this really is.

Not everyone agrees with Corey Quinn's analysis. According to the 2020 State of the Cloud Report published by Flexera, most companies embrace multi-cloud, with 93 percent of enterprises having a multi-cloud strategy.

Debanjan Saha, general manager of data analytics at Google Cloud, and Eyal Manor, vice president of engineering and product at Google Cloud, recently explained why a multi-cloud approach is the future:

We are seeing so many benefits in how people use information, with more to come, as multi-cloud computing moves from a better way to access data to a platform for using information better. All these multi-cloud capabilities are designed to offer customers flexibility and choice to run their applications in the most efficient manner (...) giving customers maximum access and decision-making power to their most valuable technology asset: their data.

They explain that the definition of multi-cloud itself is open to debate:

The term multi-cloud can mean different things to different users and customers. For some customers, multi-cloud refers to leveraging multiple public cloud technologies at once, others refer to it as using public cloud in parallel with traditional non-cloud systems, and still others mean using multiple public clouds simultaneously for different workloads.

Another common reason to embrace multi-cloud is to avoid lock-in to a single vendor. Open source advocate Peter Zaitsev writes that "the new reality is a technology universe where firms have a more distributed approach".

Corey Quinn acknowledges the lock-in issue but concludes that skills in the engineering team are more important than differences in APIs, and companies do not want to hire generalists:

What happens when you announce a global migration from AWS to Oracle Cloud? Well, to start, at least a third of your engineering staff will quit to go work down the road with their existing skill sets on a platform that other companies are using more deeply.