While unified storage has always been a desirable goal, making it actually work in the enterprise has always been problematic. Unified storage can work perfectly well in a small-to-medium (SMB) environment, but once an organization needs access to data storage that scales, unified storage systems generally have not been able to rise to the challenge.
However, this week Caringo is moving to bring unified storage to the enterprise with the release of Caringo Swarm7, a unified storage system that takes advantage of a new distributed index residing in RAM that locates and accesses objects in the system without incurring an IOPs overhead.
As one of the early proponents of unified storage, Caringo has been developing unified storage systems for years. Adrian Herrera, senior director of marketing for Caringo, says now the company wants to help IT organizations deploy unified storage at scale. To that end, Caringo Swarm7 can now be deployed across a single cluster consisting of petabytes of data, says Herrera.
Capable of supporting application workloads that require access to file, block and object-based storage systems, says Caringo, Swarm7 deploys on Linux running on standard x86 servers. As such, Herrera says, Caringo Swarm7 is an instance of software-defined storage that enables IT organizations to unify the management of data storage at a fraction of the cost of systems based on proprietary storage controllers.
While enterprise IT organizations have generally resisted unified storage, the amount of data that needs to be managed is fundamentally changing the economics of storage. The degree to which each of those organizations decides to embrace unified storage may vary, but going forward it’s almost certain that most enterprise IT organizations will deploy some form of unified storage that will be managed at a significantly higher level of software abstraction than traditional storage systems are today.