SEPATON: Playing A Key Role In Enterprise-Class Disaster Recovery
One of the significant technologies for improving data protection for basic risk management that has emerged and evolved during the past decade is disk-based backup. At the enterprise-level especially, disk-based backup is typically represented by virtual tape libraries (VTLs).While smaller companies have been able to use disk-to-disk and some limited-capacity VTLs, only a few VTL technologies provide the performance and scalability that enterprise organizations need. One of the companies to lev
January 5, 2010
One of the significant technologies for improving data protection for basic risk management that has emerged and evolved during the past decade is disk-based backup. At the enterprise-level especially, disk-based backup is typically represented by virtual tape libraries (VTLs).While smaller companies have been able to use disk-to-disk and some limited-capacity VTLs, only a few VTL technologies provide the performance and scalability that enterprise organizations need. One of the companies to leverage this enterprise market demand that has become an established market leader in the VTL space is SEPATON.
SEPATON has made its mark with S2100-ES2 VTL, a data protection appliance designed with enterprise-class scalability and performance in mind. On top of that, SEPATON's core VTL technology is designed to enable the addition of software modules that are fully integrated into its operation. One such module provides a strong data deduplication capability, another provides bandwidth-optimized remote replication.
The initial use of VTLs was primarily in operational recovery, or recovery that occurs at a local site. Although that process could be required due to a physical failure (such as two nearly simultaneous disk failures in a RAID-protected disk array), the more likely trigger event is a logical failure, such as one due to a hacker, virus, database corruption or simply an inadvertent deletion of a file. In addition to its main purpose as an operational recovery technology, a VTL provides some additional benefits, such as reducing the time of a backup window.
VTL vendors have now turned their attention to providing broader disaster recovery support and SEPATON is no exception to that trend. Essentially, data at a local site also has to be available at a remote location designated as a disaster recovery (DR) site. The DR site springs into action when the local site is unable to perform its basic functions for an extended period of time and assumes the responsibility for running production applications.
One of the traditional methods of protecting data at a DR site has been remote mirroring. This is useful for restarting critical applications at the DR site if downtime is critical for a business such as a Web-based retailer that needs to be up and running all the time. However, due to the high cost of remote mirroring, it is hard to justify for an application that really doesn't require such high availability. Even for those applications that might benefit, remote mirroring only offers protection from physical problems, as logical problems would be quickly propagated from the local site to the remote site. Every time they backup data, enterprises face the challenge of moving large volumes of data over their network to their remote site. For most organizations, remote replication is too slow and too costly to be feasible. As a result, most continue to backup data to physical tapes and truck them to an off-site location. This process is highly manual, risky and slow.An effective and less costly approach to protecting local site data at a DR site is through electronic replication that leverages deduplication to optimize bandwidth utilization. With this type of electronic replication, a copy of data is made from a local VTL to a remote VTL over an IP-based network. In SEPATON solutions, data that is being backed up to the primary VTL at the local site is examined at the object and byte level to identify what is duplicate and what is new data. SEPATON's DeltaRemote replication software enables it to backup data locally while sending a copy of only the new data and "pointers" to duplicate data to the remote site.
The remote site is mirrored to the local site, both maintain an intact version of the most recently backed up data ready for fast restores. The key to making this process financially acceptable is to minimize the amount of bandwidth necessary to move the data from one site to the other. The process of sending only changed data and the necessary pointers significantly reduces the amount of bandwidth that would otherwise be needed, making remote replication between enterprise VTLs economically attractive for the first time.
A number of disaster recovery strategies and scenarios are possible, such as sending tapes to a third party site or using a co-location facility, which complements owning a remote data center. However, owning/managing multiple data centers tends to be the province of larger organizations that can afford both the facilities and personnel to manage them. SEPATON recognizes a number of key characteristics required for enterprise-class data recovery, including:
Enterprise-class performance and capacity requirements planning. SEPATON lays down the gauntlet to competing VTL vendors with its DeltaScale grid architecture which enables companies to add capacity and or performance as they need it without needing to buy a new system. Companies can start with a single node and add multiple nodes for more performance. It also enables them to start with a single shelf of disk for storage capacity and add disk shelves as needed to store up to 1.6 PB of data. With deduplication and compression, that translates to large enough to meet all, or virtually all, of the demands that even a very large enterprise requires.
The ability to meet the key service level agreement (SLA) goals for recovery time objective (RTO). Being able to restore your data quickly is essential, as the old "time-is-money" chestnut affirms. SEPATON's DeltaCache Recovery feature of their DeltaStor deduplication software uses a technique called forward referencing to maintain the most recent backup in a format that does not require reassembly.
Effective disaster recovery testing. Disaster recovery testing is vital, but in strictly tape-based systems the time and personnel required for testing may be so demanding that testing is ignored, performed less frequently than recommended, or only partially completed (such as for critical applications). A VTL can significantly simplify the testing process. In addition, SEPATON's DeltaRemote management console enables IT administrators to track a wide range of replication metrics to test and tune the efficiency of restores. That can be a big help in simplifying the testing process.
Even though SEPATON is "no tapes" spelled backwards, the company recognizes that many of its customers need disk and tape storage to complement each other. For example, a backup application can create physical tapes either at the local or remote site using a SEPATON VTL, an important feature, since it helps enable higher levels of data protection than can be provided economically by disk-based backup alone. That said, I believe VTLs will become more prominent as one of the front lines of defense in disaster recovery scenarios.
Before that happens, however, VTLs have to go through some tests. The first is of economic viability. Using a VTL for disaster recovery requires that data be replicated from a local to a remote site with minimal use of bandwidth, i.e., not breaking the bank paying the network vendor. SEPATON passes this test but the addition of a VTL increases costs, which have to be justified by meeting a second test.
That test includes being able to meet the SLA for RTO, as well as the ability to scale to meet expanding needs. Each organization also has to evaluate the value of being able to do this job as compared to the cost (on a TCO basis) of adding the remote VTL. This is Business 101. But with the increasing importance of business continuity that justification should continue to get easier. SEPATON passes this second test, as well.That would seem to be enough but a further test offers extra credit points. And that test determines whether a VTL solution improves operational productivity and reduces human errors. SEPATON's innovative combination of automation and management reporting tools, which help to improve process and cost of disaster recovery testing, enables the company to pass this extra credit test with flying colors.
All in all, the proven value of VTLs in disaster recovery is likely to increase the use of these solutions in organizations including the largest enterprises. SEPATON's innovative approach should place it front and center for companies considering these necessary functionalities.
About the Author
You May Also Like