Continuous Data Availability and Business Continuity

With its Fully Active operations, YottaYotta can provide a zero RTO and RPO, or RT0/RP0™ Continuous Data Availability in the event of a complete array failure, a GSX 3000 node failure, or even a network failure. In the event of a compete array failure, the GSX 3000 will provide fail-over of active I/O to Site 2 – with a zero RTO and RPO. Array failures typically result in many hours of downtime. With the GSX 3000, these failure cases are completely hidden from hosts and applications, and impose no requirement for implementing host-level cluster failover, promoting passive storage volumes in Site 2, or implementing a subsequent application check prior to resuming production operations. In the event of a GSX 3000 node failure, redundant nodes will take on the required workload. In the event of a link failure, operations continue uninterrupted in Site 1, and once the link is restored, data mirroring operations will resume.

GSX 3000 Continuous Data Availability during Storage Array Failure

Even in the event of host failure or a complete site failure, the GSX 3000 can offer RTOs and RPOs within minutes. In such cases, a host-based cluster failover service (e.g., Microsoft Cluster Server, HP ServiceGuard, Veritas Cluster Services) is initiated to switch over the application from Site 1 to Site 2. This may result in a brief interruption in data access (e.g., in the order of several minutes); however, because the GSX 3000 provides a Fully Active data image at the surviving site, once servers have been successfully launched in Site 2, they will have immediate access to the active data image – without requiring that a passive storage volume be promoted to production in Site 2.

GSX 3000 Business Continuity during Site Failure

In addition to reduced RTO and RPO, as opposed to traditional active/passive data replication approaches, another key benefit of our solution is that it allows organizations to understand the true operational status of the Disaster Recovery (DR) copy – because they have the option of using it in everyday production. This approach significantly reduces the logistical overhead and labor intensiveness associated with traditional DR testing.

Finally, in case of a logical failure event (e.g., data is simultaneously corrupted at all sites due to a virus or administrative error) or in any other situation that requires an administrator to revert to an uncorrupted version of the data image, the GSX 3000 leverages existing disaster recovery and point-in-time utilities on hosts and storage systems. For example, the GSX 3000 can leverage Point-in-Time (PiT) functionality on heterogeneous storage arrays to provide multiple distributed snapshots, guaranteeing global consistency for each snapshot. This provides administrators the capability to roll back to one or more known application-consistent states – providing customers complete high-availability and data protection for both physical and logical failures.