For good reason, cloud proponents are pitching a move to the cloud as a good way to save money and enhance the flexibility of IT programs. But those benefits don’t necessarily apply to all workloads.
In fact, Kathleen Casey wrote recently at TechTarget, “some applications could perform poorly or cost more in the cloud.
That means, she argues, that organizations need to ask themselves a few questions before undertaking a cloud migration. First, and perhaps most logically, it helps if organizations have a clear understanding of what they expect to get out of the cloud. Is it lower costs? Greater flexibility? A bit, or a lot, of both?
Beyond that, there are the more practical questions of determining which programs and apps should end up in the cloud.
“Assess your application,” Casey says, “and review its requirements -- for both performance and compliance -- to determine if it is a good candidate for the public cloud. Check if the application has any specific network needs or dependencies. Unless you also migrate the systems on which your app depends, including databases, latency could be an issue. Also, review an application's design before you migrate it to the cloud; those apps that frequently read and write to storage systems, for example, could end up costing you more off premises.”
Finally, Casey points to the importance of cloud migration tools. “Migration is a complex process that comes with risk and the potential for high costs,” she notes. “There are both cloud provider-native and third-party tools that can aid organizations through the process. . . . If an enterprise needs to move large amounts of data to the cloud, it can also perform an offline data migration. This requires an organization to store its data on physical disks and then ship the disks to its cloud provider. While this method might seem old-school, it can be more cost-effective than a migration via a network if you have terabytes of data.”