XEBS-XINFINI® Block Storage
Enterprise-grade distributed block storage for all-flash data centers
All NVMe distributed block storage based on XSEA’s extremely fast shared-everything architecture
Make full use of NVMe over Fabrics standard protocols and network technologies to facilitate the connection of multiple virtualization, database, and container platforms
Cloud and virtualization
In cloud and virtualization scenarios, it can still provide consistent data services and unified management and operation and maintenance skills. XEBS-XINFINI can provide higher redundancy and reliability guarantees, and use low snapshot costs and data reduction capabilities to save at least half the cost
Efficient search
Cross-platform/cross-bucket/cross-directory global search capabilities simplify the query process. Billion-level storage scale can provide second-level queries and improve data management efficiency
Flexible permissions
The multi-tenant management system better adapts to the enterprise personnel organizational structure. The system can provide different data access views and assign different operation permissions to different users
Business linkage
Business applications can sense data changes in a timely manner, and the system can work closely with business applications to meet different data processing needs
Product Value

High performance, high reliability and high efficiency distributed all-flash storage products

XSKY is a new generation of all-flash distributed storage system that adopts the new revolutionary eXtreme Shared-Everything Architecture (referred to as: XSEA). It can make full use of the latest standard storage protocols and network technologies. Through breakthrough methods, it solves the compromise problems of performance, reliability, scale, and cost under the traditional Shared-Nothing architecture in the past 20 years

Extremely high reliability capabilities

In terms of critical reliability, the Shared- Everything architecture significantly improves failure recovery. Whether it is a node, network or SSD failure, other nodes can quickly take over to ensure business continuity. This is because each service has access to all data without having to wait for data replication or state synchronization in the event of a failure. This design targets slow disk and sub-healthy network problems and only takes 100ms to complete failover, providing high reliability guarantee for storage.

Optimized for TLC SSD

Single-layer TLC NVMe SSD is used to build a storage pool and simplify the storage hardware structure of the cluster. Through EC (erasure coding) and data compression technology, a storage system utilization rate of more than 100% is achieved, which greatly reduces TCO costs without sacrificing performance.

End-to-end NVMe protocol support

XEBS-XINFINI products fully support NVMe/TCP and NVMe/RDMA protocol interfaces. In addition to the NVMe over Fabrics standard protocol used by clients to access storage, the NVMe over Fabrics standard protocol is also used in the storage internal interconnection network. It is a complete end-to-end NVMe design, which means that all storage nodes can efficiently access each node through NVMe over Fabrics. An NVMe SSD, thus avoiding the additional overhead caused by storage protocol conversion.

Product Advantages

The industry's first distributed primary storage using shared-everything architecture

Latency as low as 100µs
  • It supports all storage nodes and can efficiently access each NVMe SSD through NVMe over Fabrics. Achieve ultra-low latency of 100μs under 40%-60% load to meet a wider range of application load requirements
>100% storage system utilization rate
  • Through EC erasure coding and data compression technology, it can achieve more than 100% storage system disk utilization rate without sacrificing performance, greatly reducing TCO costs
100ms failover
  • When faced with slow disk or sub-healthy network problems, it can quickly failover and self-heal within 100ms to ensure stable operation of the system and reduce operation and maintenance costs
open ecology

Private cloud deep adaptation