When you hear the words “volume, variety, and velocity,” you probably think of big data. While those three words certainly apply to big data use cases and technologies, they also apply in a general sense to all data creation and flows. And that presents a big challenge to data protection architectures that were not crafted to deal with a huge diversity of workloads and data types. That’s especially true within the context of ongoing information infrastructure transformation, most notably cloud and server virtualization.
Sepaton, a backup and recovery company, designed its new VirtuoSO platform specifically to meet the increasing volume, variety and velocity challenges that apply to data protection. I'll provide a close look at VirtuoSO and how it works. But first, let's examine the trends that have forced architectural changes in data backup.
IT Transformations
The first transformation of IT, which focused on the digitization of business workflow and business automation (notably online transaction processing systems), used backup software from disk to tape as the primary means to achieve basic data protection. The second transformation of IT added digitization to the human experience, starting with employees using social, communication and productivity tools such as email and word processing. The Web and mobile computing expanded consumers' access to cyberspace.
Data protection was retrofitted to try to accommodate these new processes, even though original business data was block-oriented structured data and most of the new data is file-oriented semi-structured or unstructured information. A new, important architectural change has been the adoption of disk-to-disk data protection, such as the use of a virtual tape library (VTL) that effectively replaces tape with hard disks as the first tier of backup data protection. The major driver of that change involved the incorporation of deduplication as integral to the backup process.
The third transformation of IT involves machine-created data, such as sensor-based information, a trend often referred to as the “Internet of Things.” This raises an additional challenge for data protection.
How VirtuoSO Works
VirtuoSO is a scale-out network-attached storage (NAS) product, meaning it's file-oriented (i.e., it uses the standard CIFS and NFS protocols) rather than block-oriented, like a VTL. That also means VirtuoSO is separate and distinct from Sepaton’s block-oriented VTL. However, Sepaton said VirtuoSO will eventually support VTL.
VirtuoSO contains all the necessary hardware (including storage) and software required for backup/restore, which is why the company calls it a data protection platform. The word “appliance” might also be used here, as it serves the single-purpose, dedicated mission of data protection. Sepaton describes VirtuoSO as managing all data types individually but within a single pool of storage (think of this as analogous to server virtualization consolidation).
[Read about Asigra's new pricing model, which focuses on how much data needs to be recovered rather than how much storage is used in "Shifting The Backup And Recovery Focus To Recovery."]
Although Sepaton does not refer to it in this way, I believe VirtuoSO can be considered a converged heterogeneous infrastructure for data protection, because it tightly combines HP servers, Hitachi Data Systems (HDS) storage, data backup/restore software (such as CommVault, NetBackup, Networker, Tivoli Storage Manager, and Veeam), and its own software into a single system. Consequently, it promises to provide the benefits of converged infrastructure -- centralizing the management of IT systems to consolidate systems, increase resource utilization rates and lower costs.
VirtuoSO focuses on being able to efficiently work with all types of data. This is important because the increased diversity of data types typically favor file versus block orientations. That requires the use of a file system abstraction (as there is a need for file services that provide a richer metadata environment), finer granularity on managing particular data and greater flexibility in managing data, including data movement in a virtual environment.
By the way, note that understanding and managing blocks is still essential to this process, since everything is still composed of blocks (to use a chemistry/physics analogy, think of “molecular” objects made up of “atomic” blocks).
Sepaton has built VirtuoSO around its OptiScale architecture, which is based on the Sepaton OptiScale File System (SOFS), an object store with a file system presentation layer. This reflects the movement toward object storage, which is a key trend in the storage arena. The advantage of an object store is that it can be designed to manage an essentially unlimited amount of storage.
Now, how file system abstraction can play a key role in an object store may not be intuitive, but it does. In this case, VirtuoSO has a POSIX-compliant (Portable Operating System Interface) file system abstraction on top of the object store that works with remote access methods that include NFS and CIFS at first and others later on.
Next Page: Deduplication And Object Management