Hybrid multi-cloud is the new normal. Eighty-nine percent of companies already adopt multiple clouds, and cloud spending is anticipated to grow by twenty-nine percent in the next year, according to the Flexera 2022 State of the Cloud Report.
Most companies now utilize cloud services from one or more public cloud service providers. Top-used features include cloud data warehouse, database-as-a-service, container-as-a-service, caching, and batch processing. And arguably, there are many potential benefits of multi-cloud — you can reap the best-of-breed technologies from various sources. It reduces vendor-lock-in and enables more flexibility. And cloud-agnostic open-source solutions like Kubernetes have made realizing multi-cloud more feasible.
Yet, of course, as with most technology decisions, multi-cloud is a double-edged sword. Distributing your enterprise across multiple clouds inherently adds much more complexity. It’s easier for costs to balloon and harder to manage and enact standardization across clouds. Ungoverned multi-cloud adoption could negatively impact security and the ability to apply consistent policies across your computing infrastructure.
In this analysis, we’ll weigh the strengths of multi-cloud against its weaknesses. We’ll consider when multi-cloud makes sense and gauge if multi-cloud is actually worth it.
Benefits of Multi-Cloud
Avoid vendor lock-in. A multi-cloud strategy gives the flexibility to sever a dependency on a specific cloud provider and shift to another one. This grants more mobility for computing workloads.
Customize computing with fit-for-purpose clouds. Multi-cloud provides the freedom to choose the cloud services most suitable for your use case. Selecting the best-in-class service provider for the job can increase performance and agility.
Portability could bring cost savings. Multi-cloud could — this is a big could — reduce cloud infrastructure costs. This would require keenly optimizing savings based on which type of workload you’re using and the varying resource capacities per cloud.
Adds redundancy. In the case of a major failure, the ability to switch over to another cloud service provider could be beneficial to retain stability. Such backup and recovery plans would boost a risk management footing.
Empower teams with choice. Multi-cloud empowers developers to choose the computing environment that best matches their needs, skillsets, or preferences. This is becoming more common within some companies as full-cycle developers oversee their applications throughout development and into deployment and ongoing maintenance stages.
Keeps status quo after mergers and acquisitions. Multi-cloud conditions might arise purely out of circumstance due to mergers and acquisitions or ungoverned adoption. In these cases, it’s often easier to continue to use multiple providers than refactor everything into a single cloud.
Meeting compliance and regulatory requirements. International enterprises must deal with data compliance laws like the General Data Production Regulation (GDPR) or California Consumer Privacy Act (CCPA). As some regulations decree that data must be stored in particular locations, it might be beneficial to support multiple clouds in various geographies to support these region-specific laws.
Anti-monopoly. Lastly, from a larger economic standpoint, multi-cloud helps avoid a cloud monopoly. We’ve already seen this unfold — what was once the sole territory of AWS has evolved as other incumbents like Azure, Google Cloud, Oracle Cloud Infrastructure, and IBM Cloud gain market share. This is positive for the consumer as it leads to enhanced choice in the market and more competitive prices.
Drawbacks of Multi-Cloud
Cybersecurity repercussions of multi-cloud. Ineffective multi-cloud management can bring severe cybersecurity repercussions. For one, more surface area increases the chance of cloud-native threats such as misconfigurations, insecure default states, and leaked credentials.
Policy management is decoupled. Furthermore, a multi-cloud setup makes it so that you no longer have one consistent policy for identity and access management in the cloud. Centralized role-based access control for multi-cloud requires a platform-agnostic policy enforcement agent.
Increased cloud spending. Although the argument is sometimes made that multi-cloud can decrease spending through savvy optimizations, it’s more likely to increase cloud spend. You may also miss out on the discounts of sticking with a single provider.
Navigating inconsistencies between clouds. Multicloud complexity is bound to create more headaches than sticking to a single cloud. And portability between clouds isn’t frictionless, as similar cloud functions behave differently from cloud to cloud.
Might require additional managed services. Orchestrating multi-cloud operations is a difficult task that may require specialists. Having the ability to centrally manage multiple clouds will require an abstraction layer.
Fractured cultural knowledge. When you’re working with multiple technology platforms, it can be more challenging to retain cultural knowledge and disseminate best practices throughout an organization.
When Multi-Cloud Does and Doesn’t Make Sense
So, knowing the benefits and potential downsides, is multi-cloud actually worth it? This was a question I recently posed on Twitter and Linkedin. And, the answers really depend on who you ask and what their situation is.
Perhaps a better viewpoint is: When is multi-cloud a good idea? With that in mind, let’s consider when it makes sense to adopt multi-cloud, and when it doesn’t.
Adopt multi-cloud if:
* You want to leverage the strength of various cloud providers.
* You want increased performance and flexibility.
* You want true disaster recovery.
* You desire portability and are comfortable using open-source abstractions.
* You are a platform provider and want to make your tool available to customers using any cloud provider.
Don’t adopt multi-cloud if:
* You want simplicity.
* You are hoping to decrease operational expenses.
* You don’t have the infrastructure team or know-how to manage the added complexity.
* You are a small organization with a minuscule cloud footprint.
Tell It to Me Straight — Is Multi-Cloud Worth It?
I’ll end with a cautious “yes” that multi-cloud is worth it. Although, the question is probably moot. Because, like it or not, we’re already in a multi-cloud world, and we’re not looking back. Research from Spectro Cloud found that for Kubernetes alone, 78% have production clusters in multiple environments, with AWS and Azure being the top two choices.
“No company starts with the goal of being ‘multi-cloud,’ but that’s just the reality,” said Saad Malik, CTO, and Co-Founder of Spectro Cloud. “The reality is that legacy decisions, contract terms, internal politics, organizational inertia, mergers and acquisitions, and other factors will be much more powerful than the ‘on paper’ advantages and disadvantages.”
In the coming years, the adoption of multi-cloud is expected to rise substantially. So, the best we can do is build best practices to mitigate the complexities.
“Most large enterprises I’ve seen are multi-cloud, at least in part due to acquisitions and legacy infrastructure,” wrote James Hirst, Co-Founder at Tyk. “Everyone has a consolidation strategy, with clean lines, single supplier, efficiencies of scale. Then they meet reality. Life’s messy, business is messy, embrace it, and have a strategy to handle it.”
Want more cybersecurity insights? Subscribe to the Cybersecurity as a Business Enabler channel: