Everyone who follows cloud computing agrees that we are starting to see more businesses utilise a multi-cloud strategy. The question this raises is: why is a multi-cloud strategy important from a functional standpoint, and why are enterprises deploying this strategy?
To answer this, let’s define “multi-cloud” since it means different things to different people. I personally like this one, as seen on TechTarget:
“the concomitant use of two or more cloud services to minimise the risk of widespread data loss or downtime due to a localised component failure in a cloud computing environment… a multi-cloud strategy can also improve overall enterprise performance by avoiding “vendor lock-in” and using different infrastructures to meet the needs of diverse partners and customers”
From my conversations with some cloud gurus and our customers, a multi-cloud strategy boils down to:
- Risk mitigation – low priority
- Managing vendor lock-in (price protection) – medium priority
- Optimising where you place your workloads – high priority
Let’s look at each one.
Risk mitigation
Looking at our own infrastructure at ParkMyCloud, we use AWS and other AWS services including RDS, Route 53, SNS and SES. In a risk mitigation exercise, would we look for those like services in Azure, and try to go through the technical work of mapping a 1:1 fit and building a hot failover in Azure? Or would we simply use a different AWS region – which uses fewer resources and less time?
You don’t actually need multi-cloud to do hot failovers, as you can instead use different regions within a single cloud provider. But that’s betting on the fact that those regions won’t go down simultaneously. In our case we would have major problems if multiple AWS regions went down simultaneously, but if that happens we certainly won’t be the only one in that boat.
Furthermore, to do a hot failover from one cloud provider to another (say, between AWS and Google), would require a degree of working between the cloud providers and infrastructure and application integration that is not widely available today.
Ultimately, risk mitigation just isn’t the most significant driver for multi-cloud.
Vendor lock-in
What happens when your cloud provider changes their pricing? Or your CIO says we will never be beholden to one IT infrastructure vendor, like Cisco on the network, or HP in the data centre? In that case, you lose your negotiating leverage on price and support.
On the other hand, look at Salesforce. How many enterprises use multiple CRMs?
Do you then have to design and build your applications to undertake a multi-cloud strategy from the get-go, so that transitioning everything to a different cloud provider will be a relatively simple undertaking? The complexity of moving your applications across clouds over a couple of months is nothing compared to the complexity of doing a real-time hot failover when your service is down. For enterprises this might be doable, given enough resources and time. Frankly, we don’t see much of this.
Instead, I see customers using a multi-cloud strategy to design and build applications in the clouds best suited for optimising their applications. By the way — you can then use this leverage to help prevent vendor lock-in.
Workload optimisation
Hot failovers may come to mind first when considering why you would want to go multi-cloud, but what about normal operations when your infrastructure is running smoothly? Having access to multiple cloud providers lets your engineers pick the one that is the most appropriate for the workload they want to deploy. By avoiding the “all or nothing’ approach,” IT leaders gain greater control over their different cloud services. They can pick and choose the product, service or platform that best fits their requirements, in terms of time-to-market or cost effectiveness – then integrate those services. Also, this approach may help in avoiding problems that arise when a single provider runs into trouble.
A multi-cloud strategy addresses several inter-related problems. It’s not just a technical avenue for hot failover. It includes vendor relationship management and the ability optimise your workloads based on the strengths of your teams and that CSP’s infrastructure.
By the way — when you deploy your multi-cloud strategy, make sure you have a management plan in place upfront. Too often, I hear from companies who deploy on multiple clouds but don’t have a way to see or compare them in one place. So, make sure you have a multi-cloud dashboard in place to provide visibility that spans across cloud providers, their locations and your resources, for proper governance and control. This will help you can get the most benefit out of a multi-cloud infrastructure.