“Advantages of Software-defined Storage in Petabyte-scale Environments” is the last report about Petabyte-scale storage I wrote for GigaOM Research.
Here you can read the first part of Executive Summary of the report:
Despite most vendors misusing the term “software-defined storage”, the separation of software and hardware does indeed deliver significant advantages to the end user, including faster deployment, massive scalability, and reduced costs.
Traditional data-storage systems were proprietary, custom designed, expensive, and limited in terms of software features. With those systems, the process of designing a new storage product could take years to complete. Now, thanks to faster, cheaper, and standard CPUs, storage-management code is moving up the stack. As vendors adopt standard x86 components, and separate hardware and software, storage systems can be deployed as a software package on standard x86 servers configured with local disks. Multi-petabyte scale-out storage systems in webscale and HPC environments are making their first appearances in large enterprises across all industries.
SDS does not often make sense in smaller environments, but as capacity needs grow, separating hardware from software allows the end user to tailor the best possible configuration at the minimum price and operational costs.
Full report is available here (registration required).
[Disclaimer: The reports are underwritten by Scality]