ClusterStor™ 3000
ClusterStor - Engineered for High Performance Computing
ClusterStor is an integrated Lustre® storage solution – engineered for high performance computing. ClusterStor consolidates controllers, storage, operating system, and the Lustre file system creating a new optimal scale-out storage platform for The Future of HPC Data Storage. This fully realized solution provides the industry’s greatest reliability, scalability, superior performance and ease of management.
Click here to download the ClusterStor 3000 Datasheet.
Click here to download the ClusterStor 3000 Brochure.
Click here to download the white paper: Xyratex ClusterStor 3000 - The Future of HPC Storage.
ClusterStor 3000 is designed for OEMs and end-users who want to build petascale computing systems and solutions for high performance computing in areas such as scientific research, simulation, climate modeling and oil and gas exploration.
ClusterStor Architecture
ClusterStor solutions provide record breaking storage performance as a direct result of a fresh ground-breaking architectural design and engineering approach that meets the specific needs of high performance computing and its ever expanding scale.
The ClusterStor architecture is not merely an exercise of integrating separate storage hardware and file system software components but rather is the outcome of merging file system DNA into a high performance storage solution.
ClusterStor is not a packaged mix of different off the shelf components from different vendors. ClusterStor is a new high performance computing storage element that has been designed and built from the sheet metal all the way up through to the file system. Four high level solution components (storage hardware, operating system, management user interface, and file system) have been bonded in a synergistic manner to create the most reliable, best performing, and easiest to manage HPC storage in the world today. RAID is specifically tuned for Lustre file system performance, management has been consolidated with a view across the entire storage cluster with an ability to rapidly provision, manage or diagnose any system errors.
Performance and Scale-Out
ClusterStor introduces a scale-out storage architecture especially designed to match and take full advantage of the performance and capacity scaling capabilities of the Lustre file system. ClusterStor 3000 is a rack-scale Lustre storage solution that includes high-performance Scalable Storage Units (SSU) with dense capacity and processing technology that stores 1.5 petabytes of usable file system capacity in in a single rack and half the space of traditional solutions. ClusterStor’s massive performance and capacity scaling is delivered simply by the addition of a SSU.
ClusterStor 3000 scale-out storage architecture scales from terabytes to tens of petabytes and from 2.5 gigabytes per second to 1 terabyte per second.
Ease of Management and Deployment
A new paradigm in management is delivered through ClusterStor Manager. ClusterStor Manager consolidates management of the storage hardware, RAID, operating system, and the Lustre file system into a single, easy-to-use, administrator user interface. From initial installation, creation of file systems, ongoing monitoring and management, to expansion, and software upgrades, ClusterStor Manager brings the entire solution to a single pane of glass.
ClusterStor Manager includes a web client interface hosted from the Cluster Management Unit (CMU) and interfaces with distributed ClusterStor Manager component services. This distributed and tightly integrated architecture brings a new user interface to the Lustre system administrator.
Xyratex’ ClusterStor is shipped pre-configured. In tandem with ClusterStor Manager, deployment is a snap. Instead of spending weeks getting your Lustre environment iinto production, you can be productive with ClusterStor in less than four hours.
Reliability and Serviceability
ClusterStor 3000 has been designed from the ground up as a highly available Enterprise class storage solution with resiliency engineered into all primary subsystems.
All meta-data and management server operations are resident on the redundant Cluster Management Unit. SSUs include active/active integrated server modules, redundant and independent system interconnects and Xyratex’ unique Metis™ Power Protection technology.
ClusterStor Manager is tightly integrated into the ClusterStor 3000 system stack; from Xyratex storage and embedded server modules, RAID operations, operating system health, through to the Lustre file system and the entire storage cluster.
Software and firmware upgrades across an entire ClusterStor 3000 system is executed through ClusterStor Manager removing the headache and risks that come with large complex Lustre implementations.
Lustre is a registered trademark of Oracle Corporation.