Technical Advantages
Building “Lean/Simple” Machine Room

Distributed architecture highly integrates computing, network and storage resources. The Data Center purchases only a set of infrastructure with no need to purchase, deploy and maintain a separate server or storage equipment. It can greatly save space in the machine room and also reduces the pressure on heat dissipation, UPS and other environment equipment.

Agile Start and Flexible Extension

With three-node start, IT infrastructure can be linearly extended by adding nodes. The initial investment is low and the extension cost is predictable. It helps enterprises make the investment based on reality, simplify IT investment planning and focus on business growth.

Highly Reliable and Fault Self-recovery

There is no single fault for distributed architecture. Data protection is based on advanced replication. Data can still be used even in the case of a node fault. The cluster will automatically conduct multi-node concurrent recovery for faulty data making data recovery window period much shorter.

Protecting Hardware Investment

Software definition shields the complexity of heterogeneous devices and different hardware can be matched based on actual business demand. The software iteration allows new functions without updating hardware, which saves the replacement of resources and protects the original hardware investment.

High Performance

For distributed storage architecture, one more server means the extension of a memory controller and the cluster storage performance is linearly scalable. The high performance local storage designed for raw disk and I/O localization could give full play of SSD performance.

Single Interface, Easy Management

The computing, storage, network and other resources can be unified on the single management platform to provide end-to-end visibility and simplify the operation and maintenance work. It can provide rich system information and a visual display. Once something is wrong, it can be rapidly positioned and resolved.

Application Scenario

Large applications (ultra-high concurrent real-time trading scenarios) such as e-commerce, finance, O2O, social applications, retail and SaaS service providers are universally challenged by large user bases (millions or more), frequent marketing activities and a slow response of the core trading system database.


With the extension of PB data storage to industry monitoring , remote control and smart city and under the scenarios of smart television, medical imaging, smart home and car networking, there are many sensing and monitoring devices, a high sampling rate and large data sizes. Usually, one-years worth of data storage reaches PB even EB level. However, the traditional centralized data center server architecture and open source database plan cannot store and use such a large amount of data.


Generally, a cloud service platform is home to many pictures, documents, video data, where data volume is in the billion-trillion level. Service platforms often need to index these files into the data center and provide real-time new, modified, read and delete operations at the index level.