Rackable Systems RapidScale™ clustered storage solutions utilize one or more RapidScale appliances to create a single, virtual pool of storage. Data is striped across all the appliances in a storage cluster to maximize parallelism and performance.
 SA2150 |
 SA3100 |
- 3U half-depth chassis
- 9 hot-swap 750GB SATA II disk drives
- 6.75TB of disk storage capacity
|
- 3U standard-depth chassis " 16 hot-swap 750GB SATA II disk drives
- 12.0TB of disk storage capacity
|
|
The RapidScale clustered storage system employs standard Linux servers, with
a parallel version of the proven XFS file system (zXFS), and one or more RapidScale
storage appliances – to create a single, virtual pool of storage with
data striped across all appliances in a storage cluster to maximize performance.
By utilizing standard Linux file systems and management tools, RapidScale minimizes
installation effort and management costs, and fits easily into a Linux administrator’s
compute and management environment.
RapidScale Linux Client Integration |
With its true parallel architecture, RapidScale's global namespace dramatically
simplifies storage management by eliminating the “islands of data”
problem typically associated with NAS systems. Addressing the performance requirements
of compute clusters of any size is easily achievable, as RapidScale provides
near-linear scaling of performance as storage appliances are added to the cluster.
Rackable Systems can also provide RapidScale clusters in pre-configured racks
to ensure quick installation times, even in very large implementations.
RapidScale’s parallel architecture is designed to provide efficient, near-linear
scaling. It achieves this with a unique cache coherency architecture that avoids
the communications overhead associated with other cluster file system architectures
that employ a central metadata server, or impose a high level of inter-server
communication. RapidScale’s unique approach makes it possible to build
very large storage clusters that scale in both capacity and performance, while
maintaining simple management and low cost.
For applications with the most demanding uptime requirements, RapidScale HA
configurations provide two independent levels of data protection – hardware
RAID within every appliance in addition to an efficient distributed parity RAID
scheme across multiple appliances. This strategy maximizes resiliency while
preserving availability even during storage appliance outages as hot spare failover
protection and dynamic reconstruction capabilities rapidly engage to restore
all contents of the failed system.
RapidScale clusters are designed to be an ideal fit for Linux based compute
cluster environments, in a wide variety of demanding applications. RapidScale
enables these applications with a shared file system, global namespace, scalable
capacity and performance – all while providing an exceptionally resilient
and highly available data storage system.
Click here for more information about how to
purchase Rackable Systems storage solutions.