HF3000G7

1. The industry-leading Active-Active storage architecture is adopted to achieve load balancing between controllers, eliminate controller performance bottleneck, and improve system performance. The storage system supports the configuration of volumes with or without ownership. The volume without ownership is combined with self-developed multi-path software to achieve load balancing of multiple controllers and better system performance.

2. The system adopts full redundancy architecture and modular design, the design of storage reduces cable connections,with all core components being fully redundant to eliminate single point of failure and achieve high availability of the storage system.The controller, hard drive, BBU battery, power module, and host interface card of the storage system all support Hot-Swap .

3. The system supports 2U12 and 2U25 specification controller enclosures, flexible combination of SAS SSD, SAS HDD and other hard disks, to provide higher performance and larger capacity data storage experience. The system Supports 2U25, 3U48 SAS protocol disk enclosures and 2U25 NVMe protocol disk enclosure, which can be cascaded vertically in multiple levels as required to flexibly expand storage space and reduce investment costs. The system supports the simultaneous use of  IP-SAN, FC-SAN, RoCE and other protocols. A set of storage supports multiple application platforms, and the system architecture is simpler to work.

1. Based on a new hardware platform design, the processor, motherboard, media, IO cards, etc. all support PCIE 4.0 protocol. The CPU is equipped with an NTB module to increase the image bandwidth between controllers to 128GB, providing higher data read and write capabilities

2. Two controllers support 64-512GB physical cache, combined with iTurbo7.0's intelligent IO sensing algorithm and intelligent cache partitioning function, to improve cache hit rate, accelerate hotspot data read and write capabilities, reduce IO bottlenecks, and enhance system throughput and concurrent processing capabilities

3. Supports intelligent cache acceleration: Places frequently used data on SSDs for fast access. Supports intelligent QoS, enabling upper/lower limit and burst control for IOPS, bandwidth, and latency, rationally allocating storage resources and ensuring quality of service for critical businesses.

4. Employs multi-path intelligent distribution design: Reduces data forwarding between controllers and shortens IO transmission paths, improving processing efficiency. RoCE interconnection between controllers utilizes the RDMA communication mechanism to reduce latency for inter-controller message or data transmission. Uses a resource pool lock-free design to reduce lock conflicts during resource requests across multiple cores, improving CPU core operating efficiency and performance.

1. The system can be configured with a 1+1 redundant platinum power module, supporting two power supply modes: current sharing redundancy and primary/backup redundancy, which are more energy-efficient. Paired with a 15.2WH large capacity BBU battery, it supports two consecutive storage power failure protections, ensuring that cache data can be temporarily stored on the system disk in case of power failure, ensuring that cache data is not lost. It supports quick startup within 10 minutes after restoring power supply after one power failure, and supports no system restart within 2 seconds of power outage

2. Intelligent distributed RAID technology, redundant data blocks can be quickly restored, data blocks and hot backup blocks are distributed on all hard disks. After a bad disk appears, all hard disks participate in reconstruction, and the data reconstruction time is shortened with reducing the risk of RAID failure and enhancing  system  reliability.

3. The system supports intelligent local replication, including intelligent cloning, SAN and NAS intelligent snapshots, and intelligent backup, which can meet the needs of application testing, development, analysis, backup, and effectively save storage space and reduce investment costs

4. The system supports intelligent disaster recovery replication, providing three disaster recovery replication technologies: synchronous, asynchronous, and asynchronous cycles. It can optimize IP links by adjusting the number of windows and virtual links through AI algorithms to improve performance under extreme unstable links, meet customers' different RTO and RPO requirements, and protect business continuity

5. The system supports intelligent data destruction, achieving one click intelligent destruction of data through DoD standard mode and custom mode, ensuring the security of user confidential data

6. The system supports intelligent data encryption, using AES algorithm and SM4 algorithm, and encrypting data through external/internal encryption tubes

1. infinistor® Intelligent Management Platform: The infinistor® management platform is intelligent management software for data centers, unifying the management of various devices like storage, switches, hosts, and virtual machines. Using big data and AI technologies, it builds an intelligent O&M management platform, supporting automated storage platform deployment, asset status monitoring, asset capacity forecasting, storage performance optimization, remote asset inspection, fault diagnosis, storage disk prediction, asset management, storage "3-data-center" disaster recovery, etc., enabling intelligent operations. Provides scheduled intensive snapshot protection, supports multi-copy replication, easily building cross-device disaster recovery, adding an extra layer of data security.

2. Intelligent Heterogeneous Virtualization: Leverages heterogeneous virtualization functionality to take over 95% of storage models in the industry, eliminating platform differences and enabling storage reuse. Existing storage systems can utilize this system's advanced features like intelligent online compression, intelligent thin provisioning, intelligent online deduplication, etc., improving capacity utilization in the storage pool. Enhances data mobility, eliminates data silos, and enables unified resource management, significantly reducing management complexity and lowering customer TCO.

3. Intelligent Thin Provisioning: Configures full-capacity license thin provisioning, resources are allocated on-demand with thin provisioning. Adjustable thin provisioning granularity (32k, 64k, 128k, 256k), greatly improving storage resource utilization. Single volume supports up to 512TB;future expansion does not require additional licenses, reducing customer TCO.

4. Intelligent Online Compression: Achieves data deduplication and compression based on CPU-built-in compression engines. Each controller board has 8 QAT compression engines. The compression process does not occupy CPU resources and increases throughput by over 30%, reducing costs by 50%. Data compression ratio can be as high as 5:1, significantly improving resource utilization efficiency and markedly lowering storage costs.

5. Intelligent Online Deduplication: Significant duplicate data often exists in business data. The intelligent online deduplication feature identifies duplicate data and performs deduplication as data is written to disk. Duplicate user data is stored only once on the disk, improving disk space utilization and effectively reducing user costs.

6. Intelligent Volume Mirroring: Creates a mirror for volumes mapped to hosts, keeping two identical copies of data in two different storage pools. If any copy is damaged, it provides uninterrupted application IO access.

7. Intelligent Multi-Tenancy: Isolates applications and services between different tenants through virtualization, database isolation, and other technologies, protecting their privacy and security.

8. Data Consistency: End-to-end support from host port to disk for hardware-based, industry-standard T10-PI (Protection Information) data consistency checking, comprehensively ensuring data consistency and avoiding data loss caused by media corruption.