xFusion Hyperverged Infrastructure solution - FusionOne HCL

Efficient and Simplified IT Architecture for Rapid Digital Transformation

FusionOne HCI is a hyperconverged infrastructure solution designed by xFusion based on its over 15 years of experience in virtualization and intelligent management. It efficiently integrates computing of diverse ecosystems and self-developed operating systems, delivering virtualization across multiple scenarios, including computing, storage, networking, intelligence, and security. With comprehensive support for critical databases, massive storage, desktop cloud, AI, cloud-native applications, data protection, and branch applications, our solution enables customers to swiftly build a simplified and efficient IT infrastructure.

The FusionOne HCI solution delivers integrated hardware and software and allows customers to manage software and hardware in a centralized manner on the unified management platform FusionOne Center.

 

Computing Virtualization
Computing resource virtualization, elastic hybrid scheduling of VMs and containers, pooling of XPU heterogeneous computing resources, and full lifecycle management

Distributed Storage
Fully distributed architecture, on-demand storage expansion, and unified storage of blocks, objects, and files, improving resource utilization

Unified Management Platform
Unified management of multiple resource pools including branches, edges, and VMware, realizing global information visualization and providing Computing-as-a-Service

 

Extraordinary Value in Simplicity, Experiencing the Benefits of HCI

Fast Construction of Distributed Resource Pools in Multiple Scenarios

The FusionOne HCI solution has been extensively deployed across a wide range of industries and application scenarios. It empowers you to build hyperconverged infrastructures of various architectures tailored to your business needs, enabling the rapid creation of distributed resource pools in key areas such as core production, agile operations, and development/testing environments

Server virtualization
- xFusion-developed virtualization platform, satisfying various requirements such as cost-effectiveness and heterogeneous resource pools

Private cloud
- Lightweight and open private cloud infrastructure that provides elastic expansion and production readiness
similar to public clouds DR (disaster recovery)
- Flexible DR solution, ensuring high service reliability for the local, same-city active-active, and remote
active/standby DR

Artificial intelligence
- From single structured data to massive unstructured data, and from general computing to heterogeneous
computing combinations, with unified support for AI applications and data

Desktop virtualization
- Innovative virtualization platform with stable and efficient distributed storage, making virtual desktops smoother, easier to manage, and more cost-effective and scalable

Branch offices and edge
- Simple and efficient infrastructure with multi-cluster management software, simplifying the O&M of multiple branch factories and stores
 

Four Core Values of a New Hyperconverged Experience

New ecosystem: from single computing to diversified computing
- Multi-ecosystem computing: compatible with x86, ARM, GPU, and NPU with comprehensive virtualization
- Dual virtualization engines: self-developed and VMware virtualization, unified management, and fast migration
New capability: from single virtualization to integration of compute, storage, network, intelligence, and security
- Multi-scenario virtualization: VM & containers, block & object & file, and SDN network
- Training and inference integration: xGPU engine pooling, 1% resource slicing and reusing, and out-of-the-box
New management: from single cluster to unified management of multi-computing clusters
- Integrated management: software-hardware synergy, network editing and visualization, intelligent prediction, and automatic O&M
- Computing-as-a-Service: unified management of multicomputing clusters and E2E service experience
Always online: triple reliability assurance & dual security hardening
- Triple reliability assurance: carrier-class components | VM HA & data replicas | emergency DR
- Dual security hardening: host & data & application & network security | antivirus snapshot

Integrated Hardware and Software Delivery for Ready-to-Use Data Centers

With an optimized deployment mode, the FusionOne HCI solution supports cluster deployments from 2 to 256 nodes. The solution uses xFusion-developed server OS and conducts kernel-level hardware tuning to deliver superior performance. Based on your business requirements, you can configure advanced functions such as containers, AI, DR, cloud management, backup, and network security to quickly build a robust computing
foundation for your enterprise.
 

System Hardware Specifications

The FusionOne HCI solution provides large-capacity, balance, high-performance, and compute nodes, and allows enterprise customers to rapidly build an optimal infrastructure based on their storage capacity, computing resource density, I/O devices, and scalability requirements.

Data Center Virtualization Applications

Healthcare Data Center Virtualization

 

Enterprise Branch Applications

 

Transportation Scenario

 

Cloud Native Applications

Cloud Native Scenarios

Efficient Resource Utilization and Agile Deployment
➢ Reuse of VM storage and network resources optimizes resource utilization.
➢ Wizard-based 3-step configuration and visualized configuration of container images enable agile deployment.
Open Evolution
➢ An open plug-in store is provided, where users can autonomously adapt and upload AI, CI/CD, IoT, and other ecosystem plug-ins, enabling continuous business evolution.

Native Deployment
➢ Hybrid deployment of VMs and containers, as well as native deployment of containers on bare metal, meets various base requirements.

Secure Environment
➢ The underlying VM environment, with more secure isolation, integrates VM HA and cross-node deployment, improving cluster reliability.

Key Software Features

Computing Virtualization

-  Dynamic Resource Scheduler (DRS), supporting load balancing under flexible policies
- Intelligent memory overcommitment, improving VM density by 50%
- Dual-engine containers, driving unified and efficient management
- SDN-based network virtualization and E2E network resource management, realizing centralized management and rapid network deployment
- vGPU/video memory slice of any size, decoupling service load from GPUs and CPUs
- Secure data deletion, preventing information disclosure
- VM HA and drive snapshots, providing high reliability in data centers
- Connection with third-party SAN storage, enabling volumes connected to external storage to be used as VIMS
storage for creating VMs

Distributed Storage

- Symmetrical distributed architecture, supporting concurrent I/O operations of multiple nodes
- Unified storage with a set of nodes supporting block, file, and object storage
- Innovative dynamic algorithms for rapid data reconstruction (1 TB in 10 minutes)
- Quality of service (QoS) capability depending on bandwidth and input/output operations per second (IOPS)
- Automatic thin provisioning of storage resources, providing more virtual storage resources than physical storage
resources for applications
- Virtual snapshots of user volume data for subsequent data restoration
- Two-level cache accelerating AI training by 70%

O&M Management

Hybrid deployment and unified management of x86 and ARM servers
- Unified management of heterogeneous resource pools, simplifying O&M management and improving resource
utilization
- P2V/V2V/P2P/V2P migration in multiple scenarios, reducing service interruption
- AI memory fault self-healing technology, minimizing server downtime
- AI drive fault prediction, providing more secure data
- vGPU VM HA, enabling automatic service recovery in case of hardware or software failure
- Boundless live migration, supporting VM live migration under SR-IOV and USB mounting scenarios
- 24/7 automatic cloud inspection, comprehensively safeguarding customer services
- Physical server monitoring and turning on indicators of faulty drives
- Wizard-based installation, GUI-based configuration, and one-click system initialization realizing one-click deployment
- Automatic device discovery during system installation, initialization, and capacity expansion
- Unified management including resource management, performance monitoring, alarm management, operation log
management, permission management, hardware management, health checks, and log collection
- One-click O&M including one-click capacity expansion, log collection, and health checks
- Built-in OS image repository supporting one upload, available for all clusters

System Reliability

- Device-level reliability: high-reliability design for components and 30% lower failure rate than the industry average
- DC-level reliability: 1 TB drive data reconstruction in 10 minutes, leading the industry
- Cross-DC reliability: cross-site DR and backup with RPO close to 0
- Redundant PSUs, fan modules, networks, and nodes, ensuring system availability