A roadmap for hybrid cloud

By: Pedro Soares


There is a lot written and published on hybrid cloud adoption: how to get there, why you should do it, how much more efficient it will be and how its many features will benefit your business. But these articles usually focus on the end result. Not enough attention seems to be put on the fact that moving to the cloud needs preparation on many levels.

Understanding your workloads and applications is a must if you want to embark on any kind of transformation — for business and IT alike. You may also recognize that cloud transformation comes at a cost, and it should necessarily be a part of your criteria for a successful migration plan. But what needs to happen before all this?

Cloud migration considerations

Most companies will be adopting cloud gradually, and some applications will be at least partially run on traditional IT for a long time. Barriers or inhibitors will show up sooner or later, and they’ll probably appear in a way that can at least hinder your adoption project business case. It is essential that you ensure your current design, technology plan and operational processes are ready to overcome these obstacles.

For example, if you have a heavily segmented network with many physical boundaries due to an old audit and compliance rules, chances are you’ll be in trouble when you try to expand that same design into the cloud. Rather than dealing with this when you’re developing the cloud solution, it’s much more efficient to redesign your data center to be “cloud ready.” Doing so can significantly lower the cost of transformation and its associated risk.

Along the same lines, upgrading all your security policies to cope with a virtualized world (including compute, network and storage) will help to prepare your environment to handle multiple providers when they become necessary.

Dealing with multiple providers

In the near future, a high percentage of IT services will be provided in a standardized way, and the number of service providers a company has will increase to meet business demand. My previous blog post discusses delivery models for managing the complexity in a multi-provider environment. What must be understood is that preparing your infrastructure to handle multiple providers is also a fundamental piece that should be done ahead of time (or at least added to your overall adoption plan).

No matter how well you prepare, each provider will do things in slightly different ways. A certain level of consistency and standardization needs to be achieved so you can have a cost-effective cloud adoption. Even something as simple as an IP address change can bring down a full application, and the cycles required to identify the issue and fix the code will cost a lot (both money and project time). Such unplanned costs can really stack up if you have to move several workloads.

Do you have a hybrid cloud plan that accounts for these bumps in the road? A cloud broker can help you manage and control all these differences at the infrastructure level. In my next post, I’ll dive deeper into workload migration.

Topics: , ,

About The Author

Pedro Soares

Executive Architect, IBM

Pedro Soares is an Executive Architect working for IBM Global Technology Services. He's been creating IT solutions for more than 20 years, and mostly for IBM. Started as a Systems Engineer on Operating Systems and then moved to a more Network oriented perspective, then IT Systems Management, a bit of Security, and for the last... Read more