You’ve heard that old Chinese proverb that says a journey of a thousand miles begins with a single step. It’s sound advice … except if the journey you’re talking about is the one from the data center to the cloud. With cloud deployment at the center of virtually any digital transformation effort, the journey itself can have a profound impact on the successful outcomes you seek at the destination. So you have to get it right. But what does that actually mean? And how do you configure your cloud instance appropriately to ensure you don’t inadvertently sabotage your budget and performance targets?
This journey has multiple destinations—and a lot of baggage
For companies that plan for a 100% cloud deployment, the journey can be fairly straightforward if they’re simply starting fresh with cloud-native applications. But according to IDC, that’s a small minority, only 2% of companies. The vast majority—84%—will go for a hybrid approach with some portion of their IT environment (infrastructure, applications, data analytics, etc.) in the cloud and some on-premises. And therein lies the challenge.
Migrating a self-contained application from on-premises to the cloud would be a relatively simple proposition, except that’s a rare beast nowadays. We’ve spent the past 20-some-odd years decoupling application functionality and detaching infrastructure layers into smaller and smaller components and services. This approach has delivered many benefits, including increased flexibility, agility, speed to deployment, etc. And if that sounds familiar, it should because these are also the reasons why companies are moving to the cloud—to turbocharge those benefits. Today, the typical enterprise has hundreds—or even thousands—of applications, infrastructure components, and service sets that are all interconnected in a complex web of dependencies.
It’s not hard to spot the challenge: How do you keep stuff from breaking when you move some of it to the cloud? Even if you’re in the small cohort that’s going the all-cloud route, you still probably have to migrate some parts of your existing infrastructure. And in all cases, you’re not doing one massive move but in phases. Either way, you have to deal with this challenge.
The value of move groups
Clearly, you need to understand your baselines at high fidelity and map out all the dependencies between related applications, infrastructure, and services sets. This can be a massive undertaking, making advanced analytics a must-have. But that’s just the first step; the key is in figuring out what to move when. To do that, you also need an intelligent, prioritized plan for how you’re going to address those dependencies to ensure a smooth migration. This requires advanced conversation analysis so you can construct your move groups that are targeted to move together. And with high-fidelity data about various utilization attributes from your to-be-moved workloads, you can profile them to validate performance and estimate costs in the cloud. This enables you to determine the optimal candidate cloud configurations for those groups. The result is better performance and reduced risk both during and after your migration.
Prioritize intelligently—with Virtana Workload Placement
Virtana Workload Placement delivers the dependency analysis and profiling capabilities you need to establish the move groups that need to be migrated together to improve performance, reduce risk, and avoid potentially costly mistakes.
Scott Leatherman
