The Multi-Cloud Strategy: Is this a Must-Have?

The recent outage linked to CrowdStrike, which affected 8.5 million Windows devices, had a significant impact on daily life and businesses worldwide. Though less than 1% of Windows computers were affected, the consequences were far-reaching. Imagine this: 4,300 flights canceled, airports in chaos, and handwritten boarding passes being issued like it was the '80s (or before) all over again. Financial services also suffered, with disruptions in payment systems and customer access at banks worldwide. Healthcare systems were affected too, causing significant operational issues. Healthcare systems were affected too, forcing them to postpone services and revert to manual systems. This wasn’t just confined to one region—it disrupted services across the U.S., Canada, the UK, Europe, and Asia. Err... yes, business executives raised their eyebrows, seeing this as a new challenge to tackle.

I'm not getting into the technical details of the recent outage, but there's a bigger picture worth our attention. Businesses everywhere are moving to the cloud because it lets them grow globally without the heavy upfront costs. Cloud vendors have made this super easy with their flexible, pay-as-you-go pricing. And yeah, businesses are loving it.

For the past few years, I've been helping companies build scalable systems on different cloud platforms like AWS and Microsoft Azure. One of the key aspects I consider when designing cloud architecture is ensuring resiliency and business continuity. When designing solutions, I make sure they can handle outages in any availability zone or region. If one zone goes down, the application seamlessly runs from another. And if an entire region fails, the system kicks in from a different region. It’s all about keeping things running smoothly no matter what happens. For customers with serious business-critical solutions, they always consider a multi-account strategy. This means keeping their disaster recovery (DR) environment in a separate region and under a different account. I’ve noticed that cloud providers support this multi-account strategy very well.

Now, with the recent outage, C-level executives are likely rethinking their strategies and considering the need for a multi-cloud approach. This strategy is essential for ensuring that if one cloud provider experiences a disruption, the business can seamlessly continue its operations using another provider. It's about building a robust, flexible infrastructure that isn't solely dependent on one vendor. By diversifying across multiple clouds, businesses can achieve greater resilience, minimize risks, and maintain continuous service availability, no matter what challenges arise with any single cloud provider.

Leveraging a multi-cloud strategy for business-critical services requires careful planning from the ground up. When designing cloud solutions, it’s crucial to include multi-cloud readiness. Fortunately, most cloud vendors like AWS and Azure offer similar services. For example, AWS has S3 Storage, while Azure offers Blob Storage; AWS provides EC2 instances, and Azure has Virtual Machines.

Imagine you have an application that stores media files in AWS S3. To ensure it can withstand a cloud provider outage, you need to make it multi-cloud resilient by integrating it with both AWS S3 and Azure Blob Storage. This involves keeping the files in sync across both storage solutions. If one of the cloud services goes down, your application should automatically switch to the other, allowing operations to continue smoothly. This seamless transition ensures that your business remains operational and your data remains accessible, regardless of disruptions with any single cloud provider.

Setting up a multi-cloud system requires careful planning and the right tools to synchronize both data and applications across different cloud providers. Regular data replication is essential to keep storage solutions in sync, while ensuring that your applications are also capable of running seamlessly in multiple cloud environments. Your application code should be designed to interact with various cloud services, checking their availability and switching over when necessary.

Your team, including developers and architects, must be knowledgeable about multiple cloud services. When designing or developing solutions, they should be able to map these services and make applications truly multi-cloud. This often requires constant upskilling to stay current with evolving cloud technologies. Ensuring that your team is well-versed in various cloud platforms allows you to build resilient, flexible applications that can operate seamlessly across different cloud environments, thereby enhancing your overall disaster recovery and business continuity strategies.

Summary

I believe that for business leaders, a multi-cloud strategy is a key consideration. It would be ideal if cloud vendors like AWS, Azure, and GCP could interoperate and offer multi-cloud solutions, making it simpler for businesses. However, whether this is possible in today's competitive market is debatable.

Until then, businesses need to ensure their developers and solution architects fully understand different cloud platforms. They should implement applications in a way that a disruption from any cloud vendor will not affect business operations.