Identify the system components and how these communicate. This holistic view of our system will help us determine what processes we want to keep on-premise, retire, or update to fit into the cloud.

What to do
• Identify dependencies.
• Qualify workloads.

What you can’t miss
• Change the mindset. Pay attention to how you visualize your role and responsibilities as CIO. Moving to the cloud means doing things differently: from managing access policies to rethinking your tools and security.
• Examine what you take for granted. Migration can affect even simple processes. Don’t ignore them. These can pile up and become a problem if not addressed from the beginning.

Measure the best approach by considering the system information, business objectives, and budget. Also, define what the company expects from the migration: E.g. increasing network capacity or minimizing expenses.

What to do

Cloud migration's most common options are:

Re-hosting. Migrating to the cloud without modifying the system or with little modifications. E. g. AWS Application Migration Service and AWS Server Migration Service.

Re-purchasing. Moving to a new product altogether.

Re-platforming. Modifying the software to optimize it for the cloud. It doesn’t necessarily mean making deep changes to the software architecture.

Refactoring. Adjusting software to architecture level to match cloud features and tools.

What you can’t miss
• Present the budget properly. Buying capacity differs vastly from buying hardware. Hence, when figuring out and submitting the budget make sure to acknowledge this difference. Also, sometimes companies decide to keep their systems unaltered for the time being due to the expense it would be to update them or retire them altogether. It is up to you as CIO to decide what’s best.
• Go for new business strategies. While cloud migration allows businesses to scale and work more efficiently, it also offers new functionalities and unheard business opportunities and products to users and clients. As such, CIOs have to present these as the key to the migration decision.

Implement the target system, that is, the database and workload.

What to do
• Choose a provider.
• Start with a trial. Make sure that you’re in the right direction by starting on a small path and moving from there.
• Use migration tools. To automate the process, you can rely on tools. Providers often have their own: e.g. AWS Server Migration Service, Azure Migrate and Google Workspace Migrate for GPC migration.
• Test. Test how your database and applications work in the new platform.

What you can’t miss
• Trust a partner. To ensure a smooth and swift migration, consider hiring a partner specialized in the provider of your choice.
• Learn how to deal with capacity. The cloud offers a somewhat limitless capacity. Still, don’t overbuy capacity. Learn first how your system works on the cloud. Make sure to include capacity into the cloud migration proposal for speeding up infrastructure buys approvals.
• Keep an eye on your application. In case you’re only migrating your database, bear in mind that you may also need to test if your applications still work as expected with it or if they need adjustments.
• Rely on tools. Aside from migration tools, cloud tools can help you educate on what the cloud looks like. Don’t limit yourself to the platform's tools, but look out also for third parties.

Once migration is over, what’s left is to ensure the system has successfully done so.

What to do
• Monitor. Ensure it is sustaining high availability. Performance monitoring tools can help to visualize the database and the system status.
• Backup.
• Optimize. Learn from feedback what can be improved and implement cloud tools to do so like automatic scaling for micro services.

What you can’t miss
• Learn about scalability. Scalability can help you increase revenue, but also save expenses.
• Optimize your limits. Cloud providers have limits so once you are cloud-enabled learn how your operation works and what can be cut down to ensure you go over them and have to pay an extra. Minimize your data footprint as well.