Data access and availability are key functional capabilities for data storage systems. Organizations rely on data to be available despite any number of mishaps that can affect data. These mishaps include damage from negligence, disgruntled employees, white collar espionage, curiosity, disasters, and computer system failures. Beyond protecting data, snapshots have the value of being able to let data users make changes without harming the underlying permanent data storage. These transient changes to data include creating versions, making temporary work copies, simulating new conditions, and providing temporary datasets for IT and staff uses.
Snapshots are useful for avoiding version skew when backing up volatile data sets, such as tables in databases or the folder store of a mail server. Snapshots may be either a duplicate or a copy of the dataset it represents. A "version" of a persistent dataset is effectively a snapshot.

The Value of Snapshots is based on the Value of Data

Organizations have a significant investment in their data storage resources. They create new uses and hence value from their data daily. The role snapshots play in those value creation activities has evolved. Originally snapshots were a simple technique for protecting data. Now snapshots have become an indispensable means for individual end users. They can manipulate, create, replicate, and associate their data in new and novel ways. Snapshots provide a means for making data conform to new use scenarios without requiring permanent change tot eh data itself. Should the new uses prove productive and worthy of long-term retention, snapshots can be converted to stable data storage. Productivity drives the use cases for storage. Storage becomes an even more valuable resource when supported by advanced storage management tools such as snapshots. 

Increase availability to critical Logical Drives

Xyratex RAID offers an optional snapshot feature specifically designed to ensure that user operations are not disrupted by snapshot functions and which enable high levels of production data utilization. Our snapshot is designed for users whose data availability cannot be disrupted. Point-in-time images of logical drives are saved for near instantaneous roll-back of updates. Xyratex has created a snapshot solution that is compatible with round-the-clock processing as it stages data for operations such as backup, data mining/analysis and work distribution. A momentary suspension of processing allows application data to synchronize to a known state preparing the snapshot volume for use.

A snapshot logical drive is a virtual point-in-time image of a source logical drive. It is the logical equivalent of a complete physical copy, but is created much more quickly and requires less disk space. Snapshot logical drives appear and function as standard logical drives. They are host addressable and can be read or copied to create a real copy of a point-in-time. Snapshot is an integral component of the code that runs in the RAID controller rather than on a host – maximizing performance while ensuring full availability to data as applications continue to process. Through its copy-on-write technology, snapshot preserves data in its original form in the Overwrite Data Area. This functionality asks for minimal dedication of storage capacity, typically a mere 10 percent to 20 percent of source logical drive, enabling it to generate several snapshots within the space required for a single mirror.

One feature that is a real differentiator for Xyratex is our ability to enable customers to configure where the copy-on-write data resides. Storage administrators can size and place snapshots according to the capacity and performance needs of the application. A SAS array could use SATA drives for the copy-on-write data. A RAID6 array can be snapped to a RAID0 with fewer drives. These enable the best mixing of resources in tiered storage environments that optimize cost, efficiency, data availability, and protection. Xyratex ensures data integrity with snapshots by coordinating snapshot initiation with host buffers management of in transit data images to the RAID controller.

Fast recoveries to point-in-time versions of logical drives are accomplished via “snap-back” operations. A snap-back reverses all the updates made to all of the data stored on a source logical drive to the point-in-time the snapshot was established. Users can quickly back-out erroneous changes and recover critical data.

Technical Specification