Protecting data is not simply a matter of keeping it secure. Resiliency is essential. Users need to be able to access data even in the event of a failure. To ensure this, cloud storage providers such as Amazon S3 and Microsoft Azure use an architecture known as RAIN, or Redundant Array of Inexpensive/Independent Nodes. We thought we would discuss the technique briefly because it reveals a great deal about the strengths and weaknesses of the cloud, and where the Nasuni Filer fits into the picture.
With RAIN architecture, independent servers in the cloud make complete copies of your data. This data is protected because it is copied from machine to machine in the cloud and the servers check that each copy is perfect. If one of those servers fails, your data does not disappear. The others detect the loss and make additional copies through a process called regeneration.
Most storage systems use a different architecture, known as RAID, or Redundant Array of Inexpensive/Independent Disks. Dell’s EqualLogic, HP’s LeftHand and EMC’s Clariion all deliver RAID protection to businesses through a storage area network, or SAN. The RAID method does something similar to RAIN, but at the disk or machine level.
The advantage with the RAIN architecture of the cloud is that it is much more scalable: Protection is happening at the server level, not down at the disk level. The RAIN method is also more reliable. An entire node could fail, but you would still have access to your data because it would be replicated on additional nodes.
Yet there is a drawback. None of the clouds allow you to modify the data in the cloud and maintain the same filenames – not even Amazon’s new versioning system. The clouds are fantastic as archives, but since they do not allow you to modify an object in play, they are limited as business tools.
This is one of the reasons we identified a need for an appliance like the Nasuni Filer. If you save a Word document that you have been working on, you expect the filename to remain the same, and you expect the data associated with that filename to be the new data, not the old text. You want to be able to write and overwrite as you would with a typical file server.
At Nasuni, we merged the familiar world of the file server with the protection of the unlimited cloud. The Filer has a built-in archive — snapshot technology eliminates the need for backup. Your working set is stored in the local cache, so that users enjoy local-like performance, and you can always go back through the archive to look for previous versions of a file. You can easily retrieve these older versions from the cloud and roll back unwanted changes.
This post is not meant to be a criticism of the RAIN architecture. RAIN means your files will not be lost if a server goes down. It allows the clouds to protect your data. But the Filer capitalizes on this strength while also delivering the convenience and familiarity of a typical file server. It marries the two worlds, allowing our customers to enjoy the best of both.
Amazon S3, Google Cloud and Microsoft Azure are growing at an incredible rate. See how they compare in the 2015 State of Cloud Storage infographic.
Share this blog post: