Everyone’s journey is going to be unique. There are many principles to apply while identifying your cloud migration strategy; you cannot expect to pull something off the shelf and run an out-of-the-box process or tool and have everything you need.
What to Consider When Creating Your Strategy
1. Know What You Have
Don’t assume that the spreadsheet with all your applications, servers, and systems is accurate. Don’t assume you understand how a system is being used and know everything it talks to. Have a qualified partner inventory your environment and assess its real usage with specialized tools. This gives you a data-driven, unbiased picture of what you have, what you didn’t know you had, how everything integrates, and what its real-world usage looks like.
This is also a great time to do an Optimized License Assessment (OLA) to understand what licensing you have and make adjustments in the follow-on phases. It’s a great way to optimize costs and avoid license problems in an audit.
2. Understand and Organize Your Workloads
Now that you have a good picture of what’s really in your environment and how it is used, you can group things. Group interdependent workloads and prioritize them. Cull zombie servers. Incorporate optimized licensing configurations. Document each of these workloads’ business constraints. Interview stakeholders, listen to their pains, understand their constraints.
3. Identify an Approach for Workloads
This is where you need a qualified partner who can help you look down the road a bit. Migrating to the cloud isn’t just switching data centers. It can involve any combination of migration approaches – rehosting (lifting and shifting), re-platforming, repurchasing, refactoring, retiring, or retaining. Which should you choose? Should you take a phased approach? Why?
The 6 R’s of Cloud Migration
So you know what you have, you’ve understood and organized your workloads and now it’s time to choose which approach you are going to use.
1. Rehost
Rehosting is commonly known as “lift and shift.” Many people envision this approach when they embark on their journey – which is no surprise since it is the quickest. They want to run what they have today, no changes, in the cloud.
2. Replatform
Replatforming otherwise known as the “lift, tinker, and shift” approach is running on what you have today – but making a few tweaks. In this approach, you may be deciding that your interface needs to be replaced or certain parts of the system can be cost-optimized.
3. Repurchase
Repurchasing can often mean replacing on-premise systems with cloud services, or purchasing something that runs less expensively in a cloud environment.
4. Refactor
This approach will be your go-to when you need to heavily rearchitect or rewrite a system to get the most out of it in the cloud. It could involve rebuilding your system as a cloud-native application for cost optimization and ultimate agility purposes.
6. Retain
Sometimes it’s just easier retaining a system on-premise in its current environment. It could be a legacy application with a short lifespan, a system that is highly coupled to an on-premise, physical device or maybe you’re going to go multi-cloud and leave that system where it is while you spin up a new cloud on another vendor platform to leverage its strengths. The important thing is you understand what you have and WHY you’re taking the action best for you.
5. Retire
Culling that zombie server, and retaining – which means just leaving it. Some systems, like your on-premise door access system, may need to stay
What's Next?
Now it’s time for you to decide what approach (or combination of approaches) is going to be the best for your systems.