M any companies are getting caught up in the hype of moving to the cloud, and in their initial pursuit they discover some of the hidden issues and costs that are otherwise not obvious. There are many services of great value in the public cloud: software, storage, infrastructure, and more, and the development of these services has triggered a rush to the cloud. However, just because we can outsource these services doesn’t mean that we always should, as noted in a recent article.
Cloud is not an “all-or-nothing” proposition; while there are dangers, it can be very wise to move functions that truly cannot be predicted to the public cloud. And even though public cloud enables flexibility, it may over time actually limit flexibility. Once substantial migration to the cloud takes place, it can become harder to internalize or share those processes depending on how much of your ecosystem exists on the same public cloud infrastructure.
In addition to concerns with long-term flexibility of cloud deployments, there are also considerations related to investment efficiency. A key variable for determining whether or not to employ public cloud is the frequency that data and content needs to be accessed outside of the cloud. If the cloud is serving a workforce that frequently needs to retrieve data to perform localized work, then the public cloud can be much more expensive than a scenario where the data is cold or the workforce relies completely on public cloud-based workflows to perform their jobs.
When considering a cloud solution, begin by defining your inputs, outputs, and service levels — such as performance and data durability — and then find the solution that best serves your needs. You should explore all of the options for non-cloud solutions as long as they meet your service-level agreements (SLAs). Rarely does anyone ask, “What technology is inside that public cloud service?”
Intelligent, easy-to-deploy, non-cloud solutions are available today, providing easy access by presenting the appropriate data interface (NAS, object, SAN, etc.) to users and applications, and can intelligently manage data across tiers of storage – just like what might happen in a cloud infrastructure. These storage tiers can range from flash, performance or capacity disk, nearline or offline tape, and even the public cloud. With these new powerful solutions, users can choose the storage technologies that best fit their data during its entire life, and manage data seamlessly with little or no IT or user intervention required. With a tiered storage approach a single on-premise solution can deliver storage for active workflows and for archive, while offering easier access than public cloud alternatives, and because it is on premise, performance and cost profiles may also be significantly better.
Given the new class of solutions on the market today, there are powerful alternatives to being swept up in the rush to the cloud, delivering the best, and most cost-effective solution for modern data and workflow challenges. For large or active datasets, on-premise alternatives to cloud are clearly worth considering.