The Doppler Quarterly Spring 2019 - Page 67

Many organizations have application sets that can work independently with regard to their associated data. Our goal then is to locate data where it makes sense based on the application (and geographic) need. While it may not be as straightforward as simply replicating all the data everywhere, it may make more sense in the long run, both operationally and financially. And of course, this “right data in the right place” approach is all part of planning, analysis and creating a data portability strategy that makes sense for your particular needs. Creating a Data Portability Strategy Now that we have looked at several scenarios related to data portability in the cloud, let us consider the steps that should be taken before moving the first byte out of the data center or from one public cloud to another. 1. Determine the level of risk Risk assessment and risk mitigation are sound prac- tices in any business endeavor but especially in IT operations. The number of things that can go wrong can be staggering, and planning for those things can often take weeks or months. 2. Plan for data movement It is convenient to assume that data will move into the public cloud and stay there, but it is not very real- istic. And you still have to move that data to begin with, which takes time. You have to plan for data movement not only into the cloud, but between clouds and back out again in the event you decide (or are forced) to move. Using standard data formats, data abstraction layers and service-based architec- tures can alleviate (but not eliminate) some of these concerns. 3. Design applications for the cloud Monolithic applications often require monolithic data sources and the associated access processes for those sources. The growing use of microservices and their associated data sets allow for breaking apart massive data stores into manageable chunks, which then (potentially) make migrating those services and data between clouds much easier and less time-consuming. Conclusion Application workload portability continues to be a chal- lenge for organizations seeking to leverage multiple cloud platforms. Regardless of the reason for choosing portability as a goal, the associated data portability issues will dwarf any application stack issues. And the bigger the data, the more difficult it will be to gain portability, simply due to the time and cost associated with moving that data around. Start now by examining your applications and their data, and carefully consider the real costs of portability against the perceived value. Multicloud operations are achievable and can provide real value to your organization. Data porta- bility can be planned for and managed in order to minimize costs and maximize benefits. And your cloud strategy will no longer be weighed down by the data it is handling. SPRING 2019 | THE DOPPLER | 65