Inspur TStor3000 parallel storage system is a parallel file storage system developed and optimized for high-performance computing. It is based on the commercial version of the BeeGFS file system and implements a distributed metadata architecture to improve the scalability and flexibility of applications. TStor3000 can support the configuration of high-availability clusters. When a server fails, it can ensure that customer data is not lost and business is not interrupted. By integrating the compute node SSD resources as high-speed data buffers, the IO performance of the system can be further improved. Suitable for high-performance computing fields such as life sciences, materials chemistry, engineering simulation, meteorological science, and petroleum exploration.
Quick Spec
Table 1 shows the quick spec.
|
Model
|
Inspur TStor3000 parallel storage system
|
|
Bandwidth
|
Given a cluster, the bandwidth increases linearly with the number of processes, peaking at 13GB / s
|
|
Given the number of processes, the bandwidth increases linearly with the number of storage nodes, with a peak of 25GB / s
|
|
Hard disk support
|
Supports SSD / SATA / SAS / NLSAS, etc.
|
|
Client
|
RHEL, Fedora, SLES, OpenSuse, Debian, Ubuntu, etc.
|
|
Network support
|
FDR / EDR, Omni-Path, 1 / 10GbE, Seastar, MX, Elan, etc.
|
|
Access interface
|
Support standard POSIX file interface, support NFS / CIFS access interface
|
|
Buddy Mirror
|
Supports metadata and file content mirroring, half of the nodes fail, business is not interrupted, and data is not lost
|
|
BeeOND
|
Using compute node SSD hard drives, enabling BeeOND can increase read / write bandwidth by 4-5 times
|
Product Details
Inspur TStor3000 System provides these features:
High performance
�� Developed specifically for scenarios with demanding performance requirements, supporting multi-instance deployment of a single service to share the load of a large number of clients
�� Support RDMA / RoCE and TCP protocols;
�� File index and file specific content support distributed deployment;
�� A single data stream can reach 9GB / s;
Flexibility
�� Multiple BeeGFS services can be deployed on the same server;
�� Support to set strips for different files in different folders;
�� Support for adding servers online during business operation��
�� Support real-time creation of file system instance (BeeOND);
�� Support different Linux distributions and kernel versions;
�� Support NFS and Samba protocols;
Ease of use
�� Rigorous production environment test before release;
�� The service of the server runs in user mode;
�� Develop lightweight design for user mode, client / server does not need any kernel patches;
�� No special requirements for hardware;
�� Supports BeeGFS file system monitoring through GUI deployment management
Get More Information
Do you have any question about the Inspur TStor3000 System
Contact us now via Live Chat or Contact@huatenmax.com.
Contact@huatenmax.com
+86 136 9226 0946
