Storage ParaStor200 Distributed Parallel Storage System

ParaStor200 distributed parallel storage system is independently developed by Sugon’s R&D. It is also a product of Sugon’s many years experience researching into HPC, and massive data storage and processing. ParaStor200 is definitely the right way to solve performance, reliability and scalability issues for HPC storage.

ParaStor200:

Originated from HPC

Driving and

Promoting HPC

The Evolution of HPC Storage
Such as NFS. Serious performance and scalability bottlenecks for large system or I/O intensive applications
RAID + I/O nodes + parallel file system like Lustre: Single-point failure Poor fault-tolerance
ParaStor200: Storage server based and hardwaresoftware integrated distributed parallel storage. More reliable and scalable

Product Features

Integration

  • Hardware-software integration
  • MMassive storage for unstructured data
  • Single and global namespace
  • Easy Management for distributed data
  • Multiple logical file systems support

High Performance

  • Both for single-stream and aggregate I/O throughput
  • Optimization for different file sizes
  • InfiniBand RDMA support
  • Tunable for different I/O workloads

Reliability

  • Full redundancy, no single-point failure, for missioncritical applications
  • Data protection: replication or N+M erasure coding
  • Automatic data rebuilding on disk failure

High Performance

  • Online capacity expansion
  • Automatic load balancing after expansion
  • Scale-Out, linear growth in both capacity and performance