vCluster Distributed Storage

Products

Home Product Video Cloud Storage vCluster Distributed Storage

vCluster Distributed Storage is a distributed storage product with the best performance and the most complete functions in China. It has more than 100 functions than some popular open-source distributed software and more than 30 optimized functions. The vCluster distributed storage series has greatly improved storage efficiency, as it adopts an advanced distributed architecture that allocates one task to multiple storage nodes for parallel processing. Its stackable and modular design has improved the robustness and scalability of storage, which is in line with the IT architecture development trends for enterprise storage from now to the future.

Product advantages

Extreme performance

Fastest speed: Single storage with a maximum bandwidth of up to 1GB/s which is three times that of similar products.
Highest concurrency: A single system can support tens of thousands of rendering servers to read and write concurrently
Most complete function: most complete storage functions, including parallel storage, automatic hierarchy, and second deletion of massive small files

Extreme security

Flexible: support a variety of data protection modes, including snapshot, remote disaster tolerance, copy, and erasure.
Reliable: support disk-level, node-level, cabinet-level, and data center-level redundancy.
Fast: online automatic reconstruction and multi-node reconstruction of lost data at the same time, which is 10 times faster than RAID and has no impact on the front-end businesses.

Extreme compatibility

Storage type: support three storage types, respectively block level, object level, and file level.
Multi-client: support a variety of operating systems, such as Windows and Linux.
Interface protocol: support a great variety of protocols, including NFS, CIFS, iSCSI, HDFS, FTP, HTTP, S3, Swift, and POSIX

Product features

Snapshot

It can generate snapshots in an instant to back up data for zero backup windows and ensure the business data security, thus minimizing the impact on business performance and the occupation of storage space. The generated snapshots can either be read by the host or used as data backup of certain time points.

Erasure

The erasure ratio can be automatically adjusted depending on the importance degree of the data, supporting 70% disk utilization and following flexible strategies, to improve the reliability of key data. The reconstruction of data does not require full participation and needs to rebuild the lost part only, so the reconstruction speed far surpasses the RAID mode and a high disk utilization ratio can be maintained to save investment.

Dynamic Storage Tiering

Physical nodes within the same file system are allowed to be divided into different node pools which are collections of nodes with the same characteristics. Users are allowed to define the value of data in a workflow based on a file pool policy. Dynamic Storage Tiering is triggered depending on the file pool policy to migrate data automatically, and it will not affect the businesses.

Full copy

It provides level 1 to 12 file copy protection and supports online modification of the copy number of files and simple block copy protection, so the impacts of data reconstruction on the whole cluster have been minimized. Also, it provides heterogeneous pool copy protection levels with storage, directory and file as base units as well as data grouping protection policies at hard disk level, node level and cabinet level, and dynamic data copy level adjustment.

Single name space

The storage space of all nodes is integrated into a real single file system, and through the global sharing of data, the efficient collaboration between different servers of the same application can be realized to improve the application efficiency. A unified storage resource pool is made available for the users for centralized management and distribution, to avoid the difficulties in maintenance of multiple sets of storage equipment.

Fault domain division

The hardware is divided into two or more logical areas, and the data copy and computing business, etc. are divided into different hardware domains according to the security isolation logics. Therefore, the system will still be able to prevent data loss and maintain businesses continuously available even in the accidental occurrence of a variety of force majeure circumstances.

Usage scenarios

Cartoon rendering

Operator

Satellite surveying and mapping

Energy exploration

Scientific research and education

Gene sequencing

Related cases

West China Fourth Hospital of Sichuan University

Director Pang of the Information Department, West China Fourth Hospital of Sichuan University: As an important business system platform, the intranet private cloud system bears more than 90% of the business systems in our hospital, among which the storage section is the top priority for data security in the whole system. The hybrid architecture storage of vClusters has fully met the demands of our hospital, in which the distributed storage based on X86 architecture can well meet the performance requirements of the private cloud platform and the ARM-architecture cold storage can perfectly back up the data of the private cloud. This solution can not only meet the online business system requirements but also the requirement of data backup with only one set of systems, which has well met the requirements of our hospital for newspaper publication, hospital informationization review and interconnection rating.

West China Fourth Hospital of Sichuan University

Guangshui Traffic Police Bayonet

Director Yang of Suizhou Business Division of Zhongke: vClusters’ fully integrated distributed storage solution has been running for three years without any failure, undertaking the storage and forwarding of the data of violations of laws and regulations and throughput at the traffic police bayonets of Guangshui City. Based on the demands of continuous power supply and uninterrupted businesses, the solution has completed phase-2 capacity expansion this year, which will greatly reduce our company's input in operation and maintenance and system construction costs. vClusters’ service attitude and technical support are very satisfactory throughout the process, and we also look forward to further win-win cooperation with vClusters in more projects.

Guangshui Traffic Police Bayonet

Pingdingshan Government Services Cloud

Technical Director Wang of Product Technology Department, HIT Robot Group: Since the implementation of the Pingdingshan Internet + Government Service Project, vClusters has been providing solid support as a manufacturer of computing and data equipment with a good service attitude and excellent technical strength. Through perfect matching and deployment of various business requirements, it has completed deployment in one week and continuously guaranteed the business operation. By handling special malfunctions and incidents immediately, completing support business inspection efficiently, adapting and processing linkage faults, and optimizing and improving the configurations continuously according to the faults and demands, it has enabled the continuous and stable operation of business data and provided strong support for the Pingdingshan Internet + Government Service Project.

Pingdingshan Government Services Cloud

China Merchants Bank Chongqing Branch

Director Wang, Science and Technology Information Department, China Merchants Bank Chongqing Branch: Our bank’s double-record system has been using vClusters’ distributed storage for nearly two years. During this period, Virtual Clusters has provided excellent after-sales services and stable products, which has well supported our businesses.

China Merchants Bank Chongqing Branch

Rayvision Technology

General Manager Zou of Shenzhen Rayvision Technology Co., Ltd.: vClusters’ cluster storage solution has fulfilled our company's strict requirements for high concurrency and large capacity storage. Its leading performance, product stability, and perfect after-sales services can not only provide a solid guarantee for our businesses but also greatly reduce rendering and hosting costs.

Rayvision Technology

Product specifications

Product model	vCluster-2U12	vCluster-4U24	vCluster-4U36	vCluster-4U60	vCluster-4U90

Processor type	Intel multi-core processor	Intel multi-core processor	Intel multi-core processor	Intel multi-core processor	Intel multi-core processor
High-speed cache	32-512GB cache	64-1024GB cache	96-1024GB cache	192-512GB cache	256-512GB cache
Supported hard disk type	Enterprise SSD/SAS/SATA hard disk	Enterprise SSD/SAS/SATA hard disk	Enterprise SSD/SAS/SATA hard disk	Enterprise SSD/SAS/SATA hard disk	Enterprise SSD/SAS/SATA hard disk
Network data interface	21GbE (may be upgraded to 41GbE )	41GbE (may be upgraded to 210GbE)	2*10GbE	4*10GbE	4*10GbE

Number and specification of disks	12 SSD/SAS/SAT Node A (disks)	24 SSD/SAS/SAT Node A (disks)	36 SSD/SAS/SAT Node A (disks)	60 SSD/SAS/SAT Node A (disks)	90 SSD/SAS/SAT Node A (disks)
Maximum storage capacity			72PB
Total number of files			> 10 billion
Maximum file supported			32TB
Access protocol			iSCSI, NFS, CIFS, FTP, HTTP, POSIX, S3, and SWIFT
Copy number			2-12
Erasure support			N（support up to 18 nodes）+M (1, 2, 3, and 4 nodes)
Maximum volume number			1024
Maximum snapshot number			1024
Single name space			A unified storage access directory and space
Load balancing			Storage load balancing and access load balancing
Online expansion			Support online horizontal/vertical storage node expansion
Automatic recovery			Support the automatic migration and recovery of failure nodes and hard disk data without manual intervention
Intelligent monitoring			Support the monitoring of a variety of detailed running status, including overall storage status, storage node status, hard disk status, and network status
Management interface			CLI and Web-based graphical user interface
Alarm function			Email, log, and SNMP
Thin provisioning			Support excessive capacity assignment technology
Quota management			Support directory - and user-based file number and storage capacity quota technology
Automatic grading			Optional; support media of different performance to create different storage pools, and periodic statistics of IO to achieve dynamic file migration
WORM			Optional; support the mode of single writer multiple readers to protect files from tampering
Snapshot			Support directory and LUN timed snapshot, manual snapshot, and LUN readable and writeable snapshot
Remote copy			Optional; support directory-based synchronous and LUN asynchronous remote copy
NDMP			Optional; support mainstream backup software and tape libraries, such as Veritas NBU and CommVault
Fault domain			Support fault domain division based on node, cabinet, and machine room
NFS interface refactoring			Support high-performance NFS access interfaces
Windows client			Support Windows 2008/2012 clients to provide higher read/write performance or scalability in a concurrent manner
Anti-virus			Optional; support real-time scanning of the file sharing directory with Rising AntiVirus
Storage virtualization			Support heterogeneous management and virtualization of storage in FC SAN and IP SAN mode
Volume QoS			Support online change, QoS access and data recovery restriction of volumes and QoS management; allow configuration of policies, such as fault migration or business access priority and fault migration rate, during the disk or node failure, and allow designation of storage pool and IOPS and bandwidth upper limit setting for volumes.
Lun multipath			Support more than 2 MPIO paths to provide business access security
Volume management			Support the creation, deletion, and mounting as well as the online capacity expansion and copy of volumes
Lun security			Support iSCSI CHAP authentication and authorization to prevent unauthorized access and provide data access isolation function
High-performance media support			Support different high-performance media, such as SSD and PCI-E card, to create a variety of storage pools based on main memory or cache
Unified storage			Provide capabilities of access to files, blocks and objects and support S3 and Swift object access; fully compatible with files, blocks and objects, such as HDFS and OpenStack.
High access availability			High-availability access functions available for file, block and object access interfaces to achieve uninterrupted business access.
Cache management			Caching capabilities, such as caching read and caching write, available to increase the data hit ratio and improve the overall performance of storage; compatible with the main stream SSD, PCI-E, and NVME disks.
Disk management			Support S.M.A.R.T detection and slow disk detection; can find bad disks and defective tracks effectively and quickly and give fault warning in advance; support disk roaming, hot plug, disk localization and other functions
SSD management			Life management functions available for the mainstream SSDs, PCI-E cards and NVME disks, etc. by monitoring the total amount of data written and the continuous running time
SNMP support			System log, SNMP and SMS available; allow configuration of CPU utility, used space, read/write bandwidth, IOPS, latency, concurrency, and critical services at the system, node, and user levels and user definition of alarm type, alarm level, and recovery notification; can dock the centralized network management platform and synchronize and report the alarm information timely through SNMP V2/V3 to achieve unified operation and maintenance management and centralized alarm view, performance management and configuration operations.
Storage upgrade			Online upgrade storage system and repair patches available; can realize automatic distribution of upgrade packages and automatic deployment grading through the built-in software of the system; will not cause data loss or affect the business system access during the upgrade.
Virtualization support			Support virtualization technologies, such as KVM, VWware, Xen, and HyperV, etc.; standard blocks and file copy interfaces available for virtualization, and original block devices available for virtual machines; compatible with OpenStack Kilo version and the later versions up to the latest version; Cinder Driver and Cinder-Volume standard interfaces available.
VAAI support			Support the VMware VAAI function, which can greatly reduce the occupation of virtualized host resources and the use of network bandwidth, including Full Copy, Block Zero, and Hardware Assisted Locking (ATS)
Object storage quota management			Object storage and access buckets and user quota management available, including the number of objects and the capacity size.
Object storage permission management			Access functions, such as sub-user (SWIFT), permission and key, as well as bucket access permission management.
Object storage multi-site access			Support asynchronous data replication among multiple cluster systems to achieve multi-address disaster tolerant access to objects and buckets.
Multiple storage pools for object storage			Allow the creation of a new storage pool during the scaling of a storage bucket to enable the bucket to use the old and new storage buckets simultaneously, so that the data can be rebalanced during the capacity expansion
Object storage bucket access log			Statistical functions of data usage space, upload and download traffic, upload and download times and so on available at system level, user level and bucket level, and bucket download access logs and access permissions control available.

Apply for trial quickly within just 10 seconds.

Apply for Trial