Business Case for All-Flash Arrays When Integrating Flash into the Cloud

The use of cloud-based services is growing rapidly in today’s 3rd Platform computing environment. Datacenters looking to craft the most cost-effective infrastructure to service legacy workloads like relational database and enterprise applications while accommodating next-generation applications in the mobile computing, social media, and big data and analytics areas are increasingly moving to hybrid cloud environments.

When choosing AFAs over hyperconverged architectures to build cloud infrastructure, purchasers typically cite one or more of the following:
Economics. Better storage efficiency technologies that drive higher data reduction ratios (and hence a lower effective cost per gigabyte [GB]), combined with the efficiencies overall of highly flash-optimized platforms, drive a compelling total cost of ownership (TCO) argument.
Maturity. The network-based array model is more familiar to many administrators and offers capabilities that are proven over longer periods of time than newer, hyperconverged platforms with workloads that require extremely high reliability and availability.
Data services functionality. These platforms offer a better set of features necessary to reliably manage dense, multitenant infrastructures.
Degraded mode performance. For workloads that require predictable performance, AFAs can support an infrastructure that suffers no performance impact in the event of underlying failures. For hyperconverged offerings, while supporting transparent recovery, performance will vary depending on the type of underlying failure.
With their FlashArray, vendors like Pure Storage meet these requirements and are not just selling these solutions to enterprises; these vendors have a surprisingly large percentage of their businesses coming from service providers and other organizations looking to cloud infrastructures to reliably deliver better economics and increased agility.