LUSTRE
High-Performance and Supercomputing File System
High Performance and Scalability
For the world's largest and most complex computing environments, the Lustre™ file system redefines high performance, scaling to tens of thousands of nodes and petabytes of storage with ground breaking I/O and meta-data throughput.
Key Features
Scalability
Extending your file system is simple - with Lustre's object-based storage architecture, clients and storage can be added as needed using LAN or WAN connectivity. Lustre installations can be scaled from dozens to tens-of-thousands of clients with storage capacities ranging from gigabytes to petabytes.
Speed
We know of nothing that is faster than Lustre. Lustre's high performance is achieved through fully parallel I/O between clients and servers. Aggregate performance of up to hundreds of GB/s ? and soon, TB/s ? can be achieved.
Availability
Redundant servers and storage fail-over ensure that Lustre is highly available with no single point of failure.
Lustre in Use
Lustre technology powers over 60 percent of the top 100 supercomputers. Lustre is the first and foremost production-tested, object-based Linux cluster file system and is used in some of the world's largest commercial, university, research, and government environments.
Lustre has high performance capabilities and open licensing, it is often used in supercomputers. At the present time, 15 of the top 30 supercomputers in the world have Lustre file systems in them, including the world's fastest TOP500 supercomputer, Titan.
Performance Results
- File I/O as a percent of raw bandwidth: ~90%
- Achieved single object storage server I/O: >6 GB/sec
- Achieved single client I/O: >3 GB/sec
- Achieved aggregate I/O: 1,000 GB/sec (at LLNL)
- Meta-data transaction rate: 24,000 ops/sec
- Maximum file size/ maximum file system size: 2 PB / > 512 PB
- Maximum number of clients: > 26,000