Protection domain node separation

Separate storage node logic to increase security;
Maintain data integrity and storage services independently;
Separate data storage location from performance;

Ultra high performance

Distributed storage provides a differentiated data storage architecture, significantly improving performance

Availability

Built in multiple technologies ensure 99. 99999% availability

High scalability device utilization

Support seamless horizontal expansion and elastic scaling, and support automatic balancing of storage resources across devices/clusters

Reduce overall costs

Distributed storage sharing can ensure that all resources are used

Fault Domain - Enhance Cluster Reliability

Divide fault domains based on physical server racks;
Reduce the impact of collective offline failures on the entire cabinet;
Only one copy of the data is stored in each fault domain;

Storage Pool - Performance Tiering

Support partitioning different performance storage volumes;
Support full HDD, full SSD, and SSD+HDD hybrid deployment;
Meet the different performance requirements of different business systems;

Storage Pool - Cache Acceleration

Support using SSD and NVDIMM as cache to accelerate storage volume read and write performance;
Support using memory as cache to accelerate storage volume read performance;

Data mirroring redundancy

Each data actually stores two copies, achieving data redundancy;
Two copies of each data will not be stored on the same node or in the same fault domain;
Any node or disk device failure does not affect the stable operation of storage services;

Hot backup space - reinforced data reliability

Reserve hot spare space to ensure data reconstruction;
Recommend reserving a hot spare space capacity for one node;
Support limiting the maximum read/write speed and IOPS of the client;

Storage Volume QoS

Limit the maximum performance of branch line business systems and protect core business performance;
Support setting client QoS configurations for different storage volumes;
Support limiting the maximum read/write speed and IOPS of the client;

Writable snapshot

Support creating instant, writable snapshots for storage volumes;
Support defining automatic snapshot policies, supporting units of minutes, hours, days, weeks, and months;
Snapshot can be mounted as a storage volume to the client for data recovery or restoration;

Volume migration

As business pressure increases, storage can be migrated from low performance storage pools to high-performance storage pools;
Support the migration of storage volumes between different storage pools;
Support modifying storage volume types during migration (thin provisioning, thick provisioning)

Business integration virtualization scenario (computing+storage)

Customer Demand

High scalability: With rapid data growth, it is necessary to build an elastic growth storage platform
Data security and reliability: ensuring the security and reliability of PB level data storage
Fast storage: Improve the read and write performance of massive data, ensuring data timeliness
Reduce overall operation and maintenance costs: Deploy high reliability storage resource pools and computing resource pools simultaneously in the same physical server cluster.

Highlights of the plan

Elastic scalability: Supports horizontal expansion and PB level storage capacity expansion
High data reliability: supports multiple replicas and storage self balancing, achieving reliability of 6 9s for read and write. High performance: improves overall throughput performance through parallel node access
Rich node types: The storage system includes multiple models, and capacity and performance models can be provided to storage nodes or computing nodes accordingly

Large capacity cluster

storage scenario

Customer Demand

High performance and scalability: With rapid data growth, it is necessary to build an elastic growth storage platform that provides suitable performance storage resources according to different business systems
Data security and reliability: ensuring the security and reliability of PB level data storage
Fast storage: Improve the read and write performance of massive data, ensuring data timeliness
Reduce overall operational costs: achieve the optimal balance between performance, capacity, and cost

Highlights of the plan

High performance and elastic scalability: Supports horizontal expansion, PB level storage capacity expansion, and utilizes flash memory to ensure high-performance storage systems
High data reliability: supports multiple replicas and storage self balancing, achieving reliability of 6 9s for read and write. High performance: improves overall throughput performance through parallel node access
Reduce overall operational costs: achieve the optimal balance between performance, capacity, and cost

Cloud desktop application

scenarios

Customer Demand

Security of VDI data: VDI data is an important asset for industry users, and ensuring the secure and reliable storage of diverse data is a key requirement for VDI management
Fast data extraction: Ensure that long-term stored data is stored well, found, and retrieved. Enhance timeliness and improve overall VDI management capabilities
Reduce operation and maintenance costs: Through data lifecycle management, plan data classification reasonably

Highlights of the plan

High reliability: Supports multiple replicas and mirror hot backup space technology to ensure high data reliability and high data reading performance. By sharing the load of parallel nodes, the overall throughput performance is improved, ensuring fast extraction of archived media data
Rich types of storage nodes: Desktop resources and storage resources can be deployed in the same server cluster to achieve VDI data migration according to different popularity levels