Scalable File I/O
| Lustre
Lustre® is a scalable, secure, robust, highly-available cluster file system. It is designed, developed and maintained by
Cluster File Systems, Inc The central goal is the development of a next-generation cluster file system which can serve clusters with 10,000s of nodes, petabytes of storage, move 100s of GB/sec with state of the art security and management infrastructure.
Lustre runs today on many of the largest Linux clusters in the world, including Voltaire's installation at the Tokyo Institute of Technology (Japan's No. 1 supercomputer; No. 7 supercomputer on Top500 list, June 2006) with Sun Microsystems. The supercomputer consists of next generation Sun Fire® x64 (x86, 64-bit) servers using over 10,000 AMD Opteron processor cores
connected by multiple Voltaire Grid Director™ ISR 9288 switches. The Voltaire switches also deliver high-performance connectivity to more than 1 Petabyte of Sun and NEC storage.
The latest version of Lustre is available from Cluster File Systems, Inc. Public open source releases of Lustre are made under the GNU General Public License. These releases are found here, and are used in production supercomputing environments worldwide.
Tokyo Institute of Technology Installation
650 Sun Multi-core Nodes - Each with 2 DDR HCAs
8 Voltaire Grid Director™ ISR 9288 switches
System is Stretched over 3 floors using 300 optical links
1PB InfiniBand storage with Lustre for File I/O and iSER for Block I/O
Lustre OSTs - Sun storage servers with InfiniBand
Legacy NAS storage connected via IP to Voltaire InfiniBand routers at 32 Gbps throughput