The distributed storage system of Inspur provides the best choice with high cost performance, high scalability and high reliability for business decision makers.AS13000 G5provides the unified software definition solution for the file storage, object storage and block storage based on the concept of software definition and hardware reconfiguration ,to build a mass storage system with the capacity expansion on demand, performance supply on demand and service definition on demand.AS13000G5 reaches the highest level of same grade products in the industry in terms of the function, reliability and usability by means of its professional storage operating system, excellent hardware platform (AS13000-M12 (2U12 disk)), AS13000-M36 (4U36 disk), AS13000-M25 (2U25 disk) ,AS13000-E36(4U36 ,one-way) ,AS13000-M72(4U72)and AS13000-Rack (42U whole cabinet)),abundant host interfaces (1GE, 10GE,25GE,40GE and InfiniBand 56GE/100GE), strong storage protocol supporting (NFS, CIFS, Swift/S3 ,HDFS and iSCSI) and intelligent and intuitive management software.
Furthermore, the system supports up to 5,120 nodes extension, EB level mass storage space, tens of millions of IOPS and 100GB bandwidth and reliable automatic data recovery mechanism, to meet the storage requirements for the data center cloud and enterprise digital definition on-demand, and it is widely used in weather forecasting, security monitoring, radio and television media, finance, telecommunication, education, scientific research and biological medicine industries, we had 239 nodes with 70PB deployed in a single cluster at the customer site.
●Quota management: The quota is the limit to the available space allocated to the storage system.
●Global unified namespace: The global namespace provides a valid global unique directory tree for all clients of the file system.
●File self-provisioning: The file self-provisioning is also referred to as the Thin Provision technology. Its core idea is to cheat the operating system, and make use of the characteristics that the host will not check all spaces byte by byte, and the storage space allocated to the host will not instantly be written fully, to turn the traditional full supply into the supply on demand.
●WORM: support the WORM function setting for the CIFS share directory, and this WORM function is used for the CIFS client only.
●Access audit: Perform the access audit function under the NFS and CIFS access protocol. The access audit is developed on basis of high security requirements and features the high reliability.
●Multimodal classification: Make use of the solid state disk (SSD) and the plain disk to build the different data access patterns-oriented data storage layer, to implement the definable read/write acceleration for different applications flexibly.
●Cache QoS: Implement the file level read/write cache control, to ensure the performance of the single video, implement the constant and controllable video application performance with low delay, and ensure the quality of video application service effectively.
●Accidental deletion protection of recycle bin: prevent the accidental deletion and recover the data. The recycle bin can store the files deleted by users temporarily, to prevent them from accidental deletion by users, and guarantee the user data security and the ease of use.
●Online increase storage capacity by adding drives to existing nodes or adding additional nodes to existing cluster,online remove existing node from a cluster without impacting storage service.
|Type of nodes||AS13000-M12||AS13000-M36||AS13000-M25||AS13000-E36||AS13000-M72||AS13000-Rack|
|Independent R&D||Distributed storage system with independent intellectual property rights|
|Network||External shared network: 1Gb/10Gb/25Gb/40Gb Ethernet or 40Gb/56Gb/100Gb InfiniBand|
|Internal switched network: 1Gb/10Gb/25Gb/40Gb Ethernet or 40Gb/56Gb/100Gb InfiniBand|
|System architecture||Fully symmetric distributed architecture|
|Scalability||Node: 3~5120 nodes,Capacity:36TB~5160PB|
|Data redundancy protection||Erasure coding(N+1,N+2,N+3,N+4), Copy(2-8 copies)|
|Data recovery||Data recovery rate up to 2TB/h|
|The larger the system, the faster the reconstruction|
|Height||2U||4U||2U||4U||4U||42U(32U of storage space)|
|Size||W(width)447mm; H(height)87mm; D(depth)779.8mm||W(width)482.4mm; H(height)174.4mm; D(depth)818.7mm||W(width)447mm; H(height)87mm; D(depth)779.8mm||W(width)664mm; H(height)173mm; D(depth)733mm||W(width)448mm; H(height)175mm; D(depth)898mm||42U in total: size of 1U 1200mm(L)×600mm(W)×46.5mm(H)|
|Power Supply||AC: 90 ~ 264V,or DC：200V-300V， 1 + 1 power supply redundancy||AC: 100~ 240V, DC:240V
1 + 1 power supply redundancy
|AC: 90 ~ 264V,or DC：200V-300V， 1 + 1 power supply redundancy||AC: 100~ 240V, DC:240V
1 + 1 power supply redundancy
|AC:100~240V,or 240V HVDC：240V， 1 + 1 power supply redundancy||5+5 power supply redundancy,2500W each power|
|CPU||2×Intel 4(5,6,8)XXXX||2×Intel 4(5,6,8)XXXX||2×Intel 4(5,6,8)XXXX||2×Intel 4(5,6,8)XXXX||2×Intel 4(5,6,8)XXXX||2×Intel 4(5,6,8)XXXX|
|Memory||32GB (recommended)||96GB (recommended)||96GB (recommended)||96GB (recommended)||128GB (recommended)||16Dimm DDR4, can be expanded to up to 512G|
|Power failure protection memory||16GB NVDIMM||16GB NVDIMM||16GB NVDIMM||16GB NVDIMM||16GB NVDIMM||16GB NVDIMM|
|Hard disk of the operation system||2 SAS/SATA 2.5” (RAID1)||2 SAS/SATA 2.5” (RAID1)||2 SAS/SATA 2.5” (RAID1)||2 SAS/SATA 2.5” (RAID1)||2 SAS/SATA 2.5” (RAID1)||2 SAS/SATA 3.5/2.5” (RAID1)|
|Supported hard disk types||SSD, SAS, SATA|
|Weight||Gross weight: 39 kg (full configuration of 12 disk positions). (Gross weight includes: Host + packing-case + rail + accessory box)||Gross weight: 67kg (full configuration of 36 disk positions). (Gross weight includes: Host + packing-case + rail + accessory box)||Gross weight: 36.5kg (full configuration of 25 disk positions). (Gross weight includes: Host + packing-case + rail + accessory box)||Gross weight: 63kg (full configuration of 36 disk positions). (Gross weight includes: Host + packing-case + rail + accessory box)||Gross weight: 120kg (full configuration of 72 disk positions). (Gross weight includes: Host + packing-case + rail + accessory box)||Nodes: 27.25Kg(including 3.5” hard disk 0.75*12)/21.25Kg(including 2.5” hard disk 0.25*12)|
|Cabinet: 280Kg+27.25Kg*number of nodes(12×3.5inch hard disk nodes)|
|Quota management||Support directory quota, user quota and user group quota|
|Multi-protocol sharing||File system supports multi-protocol (CIFS/NFS/FTP/HTTP) sharing|
|NAS protocol support||SMB 1.0/2.0/3.0, NFS V3/V4, FTP, HDFS, HTTP|
|NFS||Support automatic fail over and automatic recovery; can be combined with NIS domain and LDAP domain to achieve permission control|
|CIFS||Support automatic IP switch and automatic recovery when a failure happens (if the application supports retry mechanism); can be combined with AD domain to achieve permission control|
|NFS/CIFS HA||High availability of external protocol|
|Management of CIFS Shared Rights||Assign read and write permissions for CIFS sharing to local user/user group/domain users|
|Access audit||NFS/CIFS access audit supported|
|Dynamic storage tiering||Supported|
|Balanced client-connected load||Supported|
|Automatic capacity balancing||Supported|
|Asynchronous remote replication||Supported|
|Online capacity expansion of file system||Supported|
|Client API||Support Linux private client mounting|
|Supported domain||AD domain, NIS domain, LDAP domain|
|IPSAN||Block storage interface, snapshot, cloning|
|Amount of file system||Single namespace|
|File system size||Up to 1024PB|
|Maximum files in a single directory||20 million|
|Maximum file and object size||File:128TB,Object:48.8TB|
|Block size||Default block size is 4MB,can be adjusted to 1MB,2MB,4MB,8MB,16MB|
|Server SAN||Hyper-Converged deployment form|
|S3 object protocol||Supported|
|Swift object protocol||Supported|
|HDFS API||Hadoop big data analysis interface|
|RDMA||InfiniBand RDMA mechanism|
|Hard disk mis-unplug prevention||If the hard disk is mis-unplugged, plug it back in 5 minutes, then it can continue to be used as a data hard drive|
|User password management||Supported|
|Operation log management||Operation log can be queried and sorted in the interface|
|Log management||Log compression||Automatically compress and dump system logs|
|Alarm management||Alarm report||Supported|
|Management mode||Support graphic interface management|
|Management interface||Support CLI management mode|
|Alarm method||Support Alarm method like SNMP and Syslog|
|Compatibility list||For detailed information, please see the compatibility list|
|Operating temperature||5℃-35℃ (-60m~1800m); at an altitude of 1800m ~ 3000m, the temperature decreased by 0.6℃ for each 100m decreased|
|Storing temperature||-40℃ to +70℃|
|Highest operating altitude||-30.5m ~3000m;|
|Noise||Environmental temperature 25℃,< 72dBA|