It seems that nowhere is the use of clustered NAS going to be more prevalent than in the cloud. They seem tailor made for each other because cloud-based services have the need for massive scaling and moderate performance while being very cost effective. Clustered NAS solutions seem to fill that bill.Interestingly, a difference between local and cloud-based network-attached storage systems is the requirement on local NAS systems for CIFS and NFS protocol support. This isn't the case with cloud-based storage. While more and more cloud service providers are adding NFS and CIFS support, the protocols aren't considered ideal for the bandwidth-sensitive Internet. These protocols give way to HTTP, FTP, or webdev, for example.
There have been interesting solutions to resolve the gap between NFS/CIFS and the reality of the Internet. Nirvanix, for example, has rolled out a NFS/CIFS gateway that is installed at the local data center and acts as a protocol converter to its cloud storage service. Other than some support concerns about local hardware, this provides the user with a more familiar transfer protocol interface while providing transfer efficiency that a cloud storage service desires because of the thinner bandwidth and latency of the Internet.
The cloud storage providers have a key decision to make as it relates to clustered NAS, and the jury seems to be out on the best approach. Do you develop the technology in-house or use current technology? Amazon and Nirvanix are examples of companies that are leveraging their own IP and then offering a turnkey service. Other companies, such as Cleversafe or ParaScale, are providing the clustered NAS technology to infrastructure providers or even standalone organizations to create a private cloud. This allows these types of companies to focus on their core competencies and add value in other ways, not having to re-invent the storage wheel.
The distributed nature of the cloud services also plays into the hands of clustered NAS quite well, since most of these solutions have replication built into the OS. Replication isn't an add-on like in traditional storage; the capability is basically a part of the operating system itself. Clouds need this distribution of data for both protection of that data in the event of a site failure or in the distribution of that data for streaming applications. For example, if you had a movie posted to the Internet, it could be streamed from the closest, least-busy set of nodes.
In the next entry we will dive a little deeper into the performance and scale differences between tightly coupled clusters and loosely coupled clusters and how those differences might determine what type of solution you select for clustered storage, be it local or cloud.
Be sure to register for our upcoming webinar, Demystifying Primary Storage Data Reduction.
Track us on Twitter: http://twitter.com/storageswiss.
Subscribe to our RSS feed.
George Crump is founder of Storage Switzerland, an analyst firm focused on the virtualization and storage marketplaces. It provides strategic consulting and analysis to storage users, suppliers, and integrators. An industry veteran of more than 25 years, Crump has held engineering and sales positions at various IT industry manufacturers and integrators. Prior to Storage Switzerland, he was CTO at one of the nation's largest integrators.