CN103442036A - System integrating design development, post production and data storage and based on cloud platform - Google Patents
System integrating design development, post production and data storage and based on cloud platform Download PDFInfo
- Publication number
- CN103442036A CN103442036A CN2013103450778A CN201310345077A CN103442036A CN 103442036 A CN103442036 A CN 103442036A CN 2013103450778 A CN2013103450778 A CN 2013103450778A CN 201310345077 A CN201310345077 A CN 201310345077A CN 103442036 A CN103442036 A CN 103442036A
- Authority
- CN
- China
- Prior art keywords
- management
- cloud
- platform
- data
- dcache
- Prior art date
- Legal status (The legal status is an assumption and is not a legal conclusion. Google has not performed a legal analysis and makes no representation as to the accuracy of the status listed.)
- Pending
Links
Images
Landscapes
- Information Retrieval, Db Structures And Fs Structures Therefor (AREA)
Abstract
The invention provides a system integrating design development, post production and data storage and based on a cloud platform. The system integrating design development, post production and data storage and based on the cloud platform comprises an LAND-STORE parallel storage system, a COMS, a GAMS, a DIM infrastructure cloud management platform and a Golden Farm cluster rendering management platform. The system integrating design development, post production and data storage and based on the cloud platform provides software and hardware resources applied to creation and post production of a product or a service for a user and an enterprise in a high-performance computing service cloud mode, the earlier-stage creation process and the post production process of the product or the service of the user or the enterprise are combined, and the problem that clout computing and the cloud storage technology cannot be effectively combined in the prior art is solved.
Description
Technical field
The present invention relates to computer realm, particularly a kind of based on the designing and developing of cloud platform, post-production and data storage integral system.
Utilize the High-Performance Computing Cluster computing technique of independent research, the High-Performance Computing Cluster Rendering, the cloud service administrative skill, the high performance parallel memory technology changes method and the technology of the business model of traditional work creation and design, the Resource Calculation ability of utilizing the web network to provide by above-mentioned technology by works or the service from the design process, make render process, parallel storing process is unified in the cloud service end and realizes.
Background technology
At present, along with the development of computer realm, for example high-performance calculation has appearred successively, grid computing, and cloud computing, play up the cloud platform, the technology in the forward positions such as cloud storage platform.
High-performance calculation (HPC) refers to usually use computing system and the environment of several the computers (as single computational resource operation) of organizing in a lot of processors (as the part of individual machine) or a certain cluster, permitted eurypalynous HPC system, its scope is from the large-scale cluster of standard computer, to highly special-purpose hardware.The HPC system of great majority based on cluster used the high performance network interconnection, such as those network interconnection from InfiniBand or Myrinet.Basic network topology and tissue can be used a simple bus topology, in performance in very high environment, mesh network system provides shorter incubation period between main frame, so can improve overall network performance and transmission rate (as the external IBM of HPC hardware supplier, HP, Dell, Sun, domestic HPC hardware supplier dawn, tide, association, HPC software service provider: IBM, parallel science and technology, connection section, Beijing is super imperial, blue extra large red Xiang etc.).
Grid computing is by the up resources (cpu cycle and disk storage) that utilizes a large amount of heterogeneous computers, a virtual computer cluster using it in being embedded in distributed telecommunications infrastructure, provide a model for solving large-scale computational problem.The focus of grid computing is placed on the ability of calculating across management domain of supporting, this distinguishes it and traditional computer cluster or traditional Distributed Calculation.
The cloud computing cloud is by making Computation distribution on a large amount of distributed computers, but not, in local computer or remote server, the operation of enterprise data center will be more similar to the Internet.This makes the enterprise can be by resource switch to the application of needs, according to demand access computer and storage system.The representative of Typical Foreign has: Amazon, Google, Microsoft; Domestic Typical Representative has: Ali's cloud, grand cloud
Pattern and the conventional cloud computing of playing up the cloud platform are similar, soon the 3D program is placed in long-range server and plays up, user terminal is by Web software and by the high-speed Internet access access resources, instruction is sent from user terminal, server is carried out corresponding rendering task according to instruction, and the rendering result picture is transferred back in user terminal and is shown.As external RenderCore, ReBus Farm, Render Rocket, domestic " Milky Way-cruel card " supercomputing CG plays up cloud platform, Csrender, RenderBus.
The cloud storage is in cloud computing (cloud computing) conceptive extension and a derivative development new concept out.Cloud computing is the development of distributed treatment (Distributed Computing), parallel processing (Parallei Computing) and grid computing (Grid Computing), be to see through network huge computing program is split into to numerous less subprogram automatically, then transfer to bulky systems that the multi-section server forms and analyze as calculated and afterwards result is returned to the user.By cloud computing technology, Internet Service Provider can be within the several seconds, processes number in hundred million information necessarily even, reach and " supercomputer " same powerful network service.As external Amazon Cloud Drive, the storage of Dropbox cloud, the cloud storage SkyDrive of Microsoft, apple cloud storage iCloud etc., domestic Baidu's cloud storage, Huawei's cloud storage, Sina, Kingsoft cloud storage etc.
But, international all mainly the utilization with domestic enterprise is that above-mentioned a kind of or two kinds of technology walk abreast into product at present, as domestic Ali's cloud is used, cloud computing, Ali's cloud computing product or the service of cloud memory technology formation oneself, due to the restriction of technology barriers and maturity, still do not have enterprise effectively above-mentioned technology to be integrated at present, and provide product or the service of an integral body, thereby provide the software and hardware resources of creation and the post-production of product or service for user and enterprise.
Summary of the invention
(1) technical problem that will solve
The technical problem to be solved in the present invention is: the pattern that adopts the high performance computing service cloud provides the software and hardware resources of creation and the post-production of product or service for user and enterprise, by the product of user or enterprise or service from early stage production process to post production process, be combined together, improve in prior art cloud computing and cloud memory technology and fail the effectively problem of combination.
(2) technical scheme
The invention provides a kind ofly based on the designing and developing of cloud platform, post-production and data storage integral system, wherein comprise: LAND-STORE parallel memory system, COMS cluster Optimal Management System, GAMS grid application management system, DIM architecture cloud management platform and Golden Farm Cluster Rendering management platform.
Described LAND-STORE parallel memory system is for native system provides a concentrated data space, can meet utilize the art of this patent to realize the cloud platform in the different application field to input and output (I/O) performance requirement of high bandwidth, low delay.
Described COMS cluster Optimal Management System provides the high-performance calculation platform management software of an optimization for native system, realized the unified management to the cluster many levels, in order stable carry out of computational tasks on HPCC such as can automatically guarantee to play up.
Described DIM architecture cloud management system has realized the fast transferring of automation deployment, monitoring in real time and P2P, P2V and the V2V of native system IT base frame, support the fault of " N-to-1 " to shift, effectively improved the resource utilization of platform hardware equipment.Architecture cloud management platform is to select the major technique basis of server configuration in user's the art of this patent.
Described GAMS grid application management system provides a convenience, transparent message transmission channels for native system, it is that single-sign-on, cross-region and striding equipment are carried out and played up or the middleware basis of other calculation tasks, make the art of this patent not only can fully use local abundant computational resource, but also can effectively utilize the different resource that is dispersed in other position.
Described Golden Farm Cluster Rendering management platform is a professional application on native system, is the combination of animation rendering and HPCC system.Rendering platform software has been contained the distribution of system monitoring management, rendering task and has been played up the various functions subsystem such as service management, can effectively solve and play up on a large scale length consuming time, play up the problems such as resource management complexity.
Native system adopts computational resource virtual management technology, has realized unified management and the scheduling of all hardware resource, has removed the binding relationship between HPCC systems soft ware and hardware device, has built dynamic HPCC system.
Native system adopts cloud plateform system administrative skill: tentatively completed the unified management of playing up on a large scale user account number on service cloud platform based on digital certificate, realized single-sign-on; The user right completed based on the role is controlled, and has realized that the hierarchical access of hardware infrastructure and professional application is controlled.This technology can be very fast complete from playing up the cloud platform to the design process, make render process, the transplanting on data parallel storing process platform.
Native system adopts the professional application administrative skill: realized the professional application administrative mechanism of " hot plug " formula, and intelligentized DDM, guarantee that various computational tasks adopts the data of transmission cost minimum to be calculated all the time.
Native system adopts the combination technology of high-efficiency computer and professional Parallel application: present technique XML-based Web Services has developed a kind of Parallel application gateway that is applicable to HPCC, has realized the smooth connection of parallel computing trunking system and professional application software.
Non-structured data file management technology is measured in the native system employing: combine with the parallel storage of LAND-STORE, realized that the efficient administrative skill of the magnanimity destructuring being played up to data has met the demand of operation to high I/O performance of playing up on the high-performance computer.This technology by building efficient distributed cache system between computing node, realized safety, reliably, data transport service fast, avoid the concurrent reading and writing operation of a large amount of computing nodes to the centralised storage system, thereby effectively improve the overall performance of rendering task.The magnanimity destructuring is played up data management technique and is made and play up the performance bottleneck that data I/O no longer becomes system platform, guarantees that magnanimity rendering result data can export fast.
Native system adopts the multistage distribution scheduling administrative skill of concurrent job: adopt the computing node organizational form of hierarchy type to replace single management and dispatching node, solved the system bottleneck problem that in extensive high-efficiency computer, the single management scheduling node brings.Adopt the computing node organizational form of hierarchy type to replace single management and dispatching node, tentatively solved the system bottleneck problem that in extensive HPCC, the single management scheduling node brings.The multistage distribution scheduling management of parallel rendering operation makes system platform according to the loading condition of hardware device and the renderer type of operation appointment, automatically to select and to build optimum node tree and to be played up.
As preferably, in the Rendering application, a kind of novel high-performance distributed caching system (DCache) of native system employing provides high performance data buffer storage service for playing up operation when moving.The DCache system is comprised of one group of cache node (CacheNode), between node, can connect by ICP/IP protocol.A DCache service processes of each CacheNode operation provides the data buffer storage service, when service processes starts, can as buffer memory, use according to parameter application part Installed System Memory, and user data can be stored in wherein.The DCache access layer is deployed on each client that need to use DCache, thereby application program realizes the zero access of data alternately by calling it and each DCache service processes.
DCache adopts the preferential storage mode of internal memory, each node CacheNode of DCache does not communicate by letter and shares information mutually, Distributed Storage is that the distributed algorithm in client program calls DChache access layer is realized, so just do not need meta data server to preserve global data, do not have Single Point of Faliure.Obtain the preservation position of data between each client by the set of synchrodata key assignments, realize the access of data.When client needs save data, need to set a key assignments for each piece of data, then according to the data distributed algorithm, calculate the sequence number of cache node, and data are saved on corresponding node.When other clients need reading out data, can calculate according to key assignments the sequence number of cache node with same algorithm, and corresponding node reading out data.Whether DCache system simple in structure, build easily, as long as the service processes that node starts DCache just can join existing caching system, utilize the node listing of the distributed algorithm of this system and customer end adopted and client storage relevant.The cache node of DCache system and client communication are not used the complicated forms such as XML, and use the simple agreement based on line of text.Therefore, by Telnet also can be on DCache save data, obtain data.Do like this and can reduce communication overhead, the access speed of buffer memory is provided.
As preferably, the DCache system can adopt the remainder Distribution Algorithm, according to the remainder of cache node quantity, is distributed.Algorithm is the integer cryptographic Hash of calculated data key assignments at first, then, divided by cache node quantity, according to remainder, selects cache node.The calculating of remainder Distribution Algorithm is simple, and the dispersiveness of data is also quite outstanding.But its limitation is that, when adding or removing node, the cost that buffer memory is recombinated is quite huge.After adding cache node, remainder will produce great change, so just can't obtain cache node identical when preserving, thereby affects the hit rate of buffer memory, is unfavorable for having the expansion of DCache system.
As preferably, the DCache system can adopt the consistency Distribution Algorithm.At first obtain the cryptographic Hash of each cache node of DCache system, and it is configured on 0～232 circle.Then use the same method and calculate the cryptographic Hash of storage data key assignments, and be mapped on circle.Then from data-mapping to position start to search clockwise, data are saved on first node found.Still can not find server if surpass 232, will be saved on the First cache node, when increasing cache node, the remainder Distribution Algorithm has in state from the graph adds a cache node.The remainder distributed algorithm is because great variety can occur the node of preserving key, and affect the hit rate of buffer memory, but utilizes in the consistency Distribution Algorithm, and only on garden, the key on the conterclockwise First node of inverse position of increase node can be affected.Therefore, the consistency Distribution Algorithm has suppressed the redistribution of key to greatest extent, and the buffer memory while reducing the node increase and decrease redistributes, thereby has improved the performance of total system.
As preferably, the DCache system also provides standard interface to allow the user to realize the algorithm of oneself.
(3) invention beneficial effect
The pattern that the present invention adopts the high performance computing service cloud provides the software and hardware resources of creation and the post-production of product or service for user and enterprise, by the product of user or enterprise or the service from early stage production process to post production process, be combined together, production process by the cloud service pattern by product or service achievement, manufacturing process integrates, adopt cloth formula cache systems simultaneously, realized safety, reliably, data transport service fast, avoid the concurrent reading and writing operation of a large amount of computing nodes to the centralised storage system, can fully meet the I/O performance requirement of the application access mass small documents on high-efficiency computer.
For the Rendering on extensive high-efficiency computer, apply, native system adopts a kind of novel high-performance distributed caching system (DCache), for distributed system, it is the key of decision systems autgmentability, performance and stability that metadata is processed, the DCache system has thoroughly been abandoned Metadata Service, use distributed algorithm to replace concentrating or distributed Metadata Service in the traditional distributed file system, carry out the data location according to the data key assignments, thereby essence has solved this difficulty of metadata.
The accompanying drawing explanation
Below in conjunction with drawings and Examples, the invention will be further described:
Fig. 1 is the architecture that the present invention is based on the designing and developing of cloud platform, post-production and data storage integral system.
Fig. 2 is the DCache system architecture diagram.
Fig. 3 is for passing through DCache system save data schematic diagram
Fig. 4 is for passing through DCache system reading out data schematic diagram
Embodiment
The cloud platform that utilizes the art of this patent to realize, can adopt Distributed Implementation, overall planning, need to be grasped therein outside the function configuration method that windows server2012 operating system and (SuSE) Linux OS are detailed, in addition in conjunction with the HPCC management software, the render farm management software, architecture cloud management system, parallel memory file system, by the enforcement of said system, carry out consolidation and innovation by CG field production process in early stage and anaphase effect making.
What current internet cloud service pattern generally adopted is the cloud host technology, as Ali's cloud, grand cloud only provides the cloud main frame, the single technology such as cloud storage, the art of this patent not merely provides the cloud main frame, and the cloud storing technology also will provide authoring software, it is render process that the follow-up effect manufacturing process of creation product will be provided in addition, by the art of this patent, the creation of product and manufacturing process is gathered closely.
In the art of this patent manufacturing process, Rendering carries out deep optimization: after one is played up job initiation and completes the structure of playing up the computing node tree, the root node of tree is responsible for starting the DCache server processes of all nodes, logically form two DCache systems, the cache node of first systematic comprises root node and all intermediate nodes; The cache node of second system comprises all leaf nodes.
Then, the file data that root node needs this operation from concentrated parallel file system is cached to first systematic, the path of its File and filename are as the key assignments of file data, finally the key assignments set of file is issued to all leaf nodes, leaf node after receiving the key assignments set from first systematic reading out data being kept on local disk, for leaf node when carrying out rendering task.
After leaf node completes playing up of a frame, directly this document data are outputed on the machine in the sub-buffer memory that belongs to second system, then frame number is reported to root node.When whole, play up after operation completes, root node has just been preserved the corresponding look-up table of each frame and leaf node, and then root node, using this look-up table as distributed algorithm, reads frame data and it is preserved back to centralised storage from second system.
In whole render process, utilize the DCache system, rendering node has been avoided concurrent reading and writing centralised storage system, and most file operations are all carried out on DCache system cache and local disk, improve the efficiency of file read-write, thereby accelerated to play up the execution of operation.
Above-mentioned example is only explanation technical conceive of the present invention and characteristics, and its purpose is to allow the person skilled in the art can understand content of the present invention and implement according to this, can not limit the scope of the invention with this.All equivalent transformations that Spirit Essence is done according to the present invention or modification, within all should being encompassed in protection scope of the present invention.
Claims (10)
1. one kind based on the designing and developing of cloud platform, post-production and data storage integral system, comprising: LAND-STORE parallel memory system, COMS cluster Optimal Management System, GAMS grid application management system, DIM architecture cloud management platform and Golden Farm Cluster Rendering management platform;
Described LAND-STORE parallel memory system is for native system provides a concentrated data space, can meet utilize the art of this patent to realize the cloud platform in the different application field to input and output (I/O) performance requirement of high bandwidth, low delay;
Described COMS cluster Optimal Management System provides the high-performance calculation platform management software of an optimization for native system, realized the unified management to the cluster many levels, in order stable carry out of computational tasks on HPCC such as can automatically guarantee to play up;
Described DIM architecture cloud management system has realized the fast transferring of automation deployment, monitoring in real time and P2P, P2V and the V2V of native system IT base frame, support the fault of " N-to-1 " to shift, effectively improved the resource utilization of platform hardware equipment.Architecture cloud management platform is to select the major technique basis of server configuration in user's the art of this patent;
Described GAMS grid application management system provides a convenience, transparent message transmission channels for native system, it is that single-sign-on, cross-region and striding equipment are carried out and played up or the middleware basis of other calculation tasks, make the art of this patent not only can fully use local abundant computational resource, but also can effectively utilize the different resource that is dispersed in other position;
Described Golden Farm Cluster Rendering management platform is a professional application on native system, is the combination of animation rendering and HPCC system.Rendering platform software has been contained the distribution of system monitoring management, rendering task and has been played up the various functions subsystem such as service management, can effectively solve and play up on a large scale length consuming time, play up the problems such as resource management complexity.
2. integral system according to claim 1, it is characterized in that adopting computational resource virtual management technology, unified management and the scheduling of all hardware resource have been realized, remove the binding relationship between HPCC systems soft ware and hardware device, built dynamic HPCC system.
3. according to integral system claimed in claim 1, it is characterized in that native system adopts cloud plateform system administrative skill, tentatively completed the unified management of playing up on a large scale user account number on service cloud platform based on digital certificate, realized single-sign-on; The user right completed based on the role is controlled, and has realized that the hierarchical access of hardware infrastructure and professional application is controlled.This technology can be very fast complete from playing up the cloud platform to the design process, make render process, the transplanting on data parallel storing process platform.
4. according to integral system claimed in claim 1, it is characterized in that native system adopts the professional application administrative skill: the professional application administrative mechanism that has realized " hot plug " formula, and intelligentized DDM, guarantee that various computational tasks adopts the data of transmission cost minimum to be calculated all the time.
5. according to integral system claimed in claim 1, it is characterized in that native system adopts the combination technology of high-efficiency computer and professional Parallel application, XML-based Web Services has developed a kind of Parallel application gateway that is applicable to HPCC, has realized the smooth connection of parallel computing trunking system and professional application software.
6. according to integral system claimed in claim 1, it is characterized in that the native system employing measures non-structured data file management technology: combine with the parallel storage of LAND-STORE, realized that the efficient administrative skill of the magnanimity destructuring being played up to data has met the demand of operation to high I/O performance of playing up on the high-performance computer.
7. according to integral system claimed in claim 1, it is characterized in that native system adopts the multistage distribution scheduling administrative skill of concurrent job: adopt the computing node organizational form of hierarchy type to replace single management and dispatching node, solved the system bottleneck problem that in extensive high-efficiency computer, the single management scheduling node brings; The multistage distribution scheduling management of parallel rendering operation makes system platform according to the loading condition of hardware device and the renderer type of operation appointment, automatically to select and to build optimum node tree and to be played up.
8. according to integral system claimed in claim 1, it is characterized in that in the Rendering application, native system adopts distributed cache system (DCache), provide high performance data buffer storage service for playing up operation when moving, be that the DCache system is comprised of one group of cache node (CacheNode), can connect by ICP/IP protocol between node.A DCache service processes of each CacheNode operation provides the data buffer storage service, when service processes starts, can as buffer memory, use according to parameter application part Installed System Memory, and user data can be stored in wherein.The DCache access layer is deployed on each client that need to use DCache, thereby application program realizes data alternately by calling it and each DCache service processes.
9. according to integral system claimed in claim 8, it is characterized in that described DCache system adopts remainder Distribution Algorithm or consistency Distribution Algorithm.
10. according to integral system claimed in claim 8, it is characterized in that the DCache system provides standard interface to allow the user to only have the interpolation algorithm.
Priority Applications (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN2013103450778A CN103442036A (en) | 2013-08-09 | 2013-08-09 | System integrating design development, post production and data storage and based on cloud platform |
Applications Claiming Priority (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN2013103450778A CN103442036A (en) | 2013-08-09 | 2013-08-09 | System integrating design development, post production and data storage and based on cloud platform |
Publications (1)
| Publication Number | Publication Date |
|---|---|
| CN103442036A true CN103442036A (en) | 2013-12-11 |
Family
ID=49695701
Family Applications (1)
| Application Number | Title | Priority Date | Filing Date |
|---|---|---|---|
| CN2013103450778A Pending CN103442036A (en) | 2013-08-09 | 2013-08-09 | System integrating design development, post production and data storage and based on cloud platform |
Country Status (1)
| Country | Link |
|---|---|
| CN (1) | CN103442036A (en) |
Cited By (12)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN105407159A (en) * | 2015-11-25 | 2016-03-16 | 北京华油信通科技有限公司 | Logistics transportation position service system |
| CN105635296A (en) * | 2016-01-09 | 2016-06-01 | 青岛大学 | Cloud startup system for computer network |
| CN106250562A (en) * | 2016-08-24 | 2016-12-21 | 苏州蓝海彤翔系统科技有限公司 | Processing data information system |
| CN106776783A (en) * | 2016-11-24 | 2017-05-31 | 福建亿榕信息技术有限公司 | Unstructured data memory management method, server and system |
| CN106899602A (en) * | 2017-03-13 | 2017-06-27 | 广州五舟科技股份有限公司 | distributed computing platform and its file management method |
| CN107196803A (en) * | 2017-05-31 | 2017-09-22 | 中国人民解放军信息工程大学 | The dynamic generation and maintaining method of isomery cloud main frame |
| CN107423321A (en) * | 2017-03-31 | 2017-12-01 | 上海斐讯数据通信技术有限公司 | It is applicable the method and its device of high-volume small documents cloud storage |
| CN107992392A (en) * | 2017-11-21 | 2018-05-04 | 国家超级计算深圳中心(深圳云计算中心) | A kind of automatic monitoring repair system and method for cloud rendering system |
| CN109815016A (en) * | 2019-01-25 | 2019-05-28 | 北京仿真中心 | A kind of High-Performance Computing Cluster method of servicing |
| CN110932914A (en) * | 2019-12-16 | 2020-03-27 | 深信服科技股份有限公司 | Deployment method, deployment device, hybrid cloud system architecture and computer storage medium |
| CN112328318A (en) * | 2020-09-27 | 2021-02-05 | 北京华胜天成科技股份有限公司 | Method and device for automatic planning of proprietary cloud platform and storage medium |
| CN115118779A (en) * | 2022-06-24 | 2022-09-27 | 济南浪潮数据技术有限公司 | Method, system, device and medium for building cluster based on centralized storage |
-
2013
- 2013-08-09 CN CN2013103450778A patent/CN103442036A/en active Pending
Cited By (20)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN105407159A (en) * | 2015-11-25 | 2016-03-16 | 北京华油信通科技有限公司 | Logistics transportation position service system |
| CN105407159B (en) * | 2015-11-25 | 2018-09-25 | 北京华油信通科技有限公司 | Logistics transportation position service system |
| CN105635296A (en) * | 2016-01-09 | 2016-06-01 | 青岛大学 | Cloud startup system for computer network |
| CN106250562B (en) * | 2016-08-24 | 2019-08-23 | 苏州蓝海彤翔系统科技有限公司 | Processing data information system |
| CN106250562A (en) * | 2016-08-24 | 2016-12-21 | 苏州蓝海彤翔系统科技有限公司 | Processing data information system |
| CN106776783A (en) * | 2016-11-24 | 2017-05-31 | 福建亿榕信息技术有限公司 | Unstructured data memory management method, server and system |
| CN106776783B (en) * | 2016-11-24 | 2019-10-01 | 福建亿榕信息技术有限公司 | Unstructured data memory management method and system |
| CN106899602A (en) * | 2017-03-13 | 2017-06-27 | 广州五舟科技股份有限公司 | distributed computing platform and its file management method |
| CN106899602B (en) * | 2017-03-13 | 2019-12-27 | 广州五舟科技股份有限公司 | Distributed computing platform and file management method thereof |
| CN107423321B (en) * | 2017-03-31 | 2020-12-01 | 北京亿智云科技有限公司 | Method and device suitable for cloud storage of large-batch small files |
| CN107423321A (en) * | 2017-03-31 | 2017-12-01 | 上海斐讯数据通信技术有限公司 | It is applicable the method and its device of high-volume small documents cloud storage |
| CN107196803B (en) * | 2017-05-31 | 2019-11-22 | 中国人民解放军信息工程大学 | The dynamic generation and maintaining method of isomery cloud host |
| CN107196803A (en) * | 2017-05-31 | 2017-09-22 | 中国人民解放军信息工程大学 | The dynamic generation and maintaining method of isomery cloud main frame |
| CN107992392A (en) * | 2017-11-21 | 2018-05-04 | 国家超级计算深圳中心(深圳云计算中心) | A kind of automatic monitoring repair system and method for cloud rendering system |
| CN109815016A (en) * | 2019-01-25 | 2019-05-28 | 北京仿真中心 | A kind of High-Performance Computing Cluster method of servicing |
| CN110932914A (en) * | 2019-12-16 | 2020-03-27 | 深信服科技股份有限公司 | Deployment method, deployment device, hybrid cloud system architecture and computer storage medium |
| CN110932914B (en) * | 2019-12-16 | 2023-04-07 | 深信服科技股份有限公司 | Deployment method, deployment device, hybrid cloud system architecture and computer storage medium |
| CN112328318A (en) * | 2020-09-27 | 2021-02-05 | 北京华胜天成科技股份有限公司 | Method and device for automatic planning of proprietary cloud platform and storage medium |
| CN112328318B (en) * | 2020-09-27 | 2024-05-14 | 北京华胜天成科技股份有限公司 | Method, device and storage medium for automatic planning of proprietary cloud platform |
| CN115118779A (en) * | 2022-06-24 | 2022-09-27 | 济南浪潮数据技术有限公司 | Method, system, device and medium for building cluster based on centralized storage |
Similar Documents
| Publication | Publication Date | Title |
|---|---|---|
| CN103442036A (en) | System integrating design development, post production and data storage and based on cloud platform | |
| CN109643312B (en) | Hosted query service | |
| CN108885582B (en) | Multi-tenant memory services for memory pool architecture | |
| Xu | Cloud-sea computing systems: Towards thousand-fold improvement in performance per watt for the coming zettabyte era | |
| JP2018088293A (en) | Database system providing single tenant environment and a plurality of tenant environments | |
| Adya et al. | Fast key-value stores: An idea whose time has come and gone | |
| CN106294870B (en) | Object-based distribution cloud storage method | |
| US10146814B1 (en) | Recommending provisioned throughput capacity for generating a secondary index for an online table | |
| US10102230B1 (en) | Rate-limiting secondary index creation for an online table | |
| US10158709B1 (en) | Identifying data store requests for asynchronous processing | |
| Mehdipour et al. | Energy-efficient big data analytics in datacenters | |
| US10135703B1 (en) | Generating creation performance metrics for a secondary index of a table | |
| CN103595799A (en) | Method for achieving distributed shared data bank | |
| Zhou et al. | Cost-aware partitioning for efficient large graph processing in geo-distributed datacenters | |
| US9185008B1 (en) | Operational reporting in a computing environment | |
| US9898614B1 (en) | Implicit prioritization to rate-limit secondary index creation for an online table | |
| Fazio et al. | How to enhance cloud architectures to enable cross-federation: Towards interoperable storage providers | |
| Sun et al. | Research on cloud computing application in the peer-to-peer based video-on-demand systems | |
| Ullah et al. | Cloud computing: a prologue | |
| Jiang et al. | Scalable video transcoding in public clouds | |
| Dongsheng et al. | Distributed cache memory data migration strategy based on cloud computing | |
| Rasool et al. | Replica placement in multi-tier data grid | |
| US20080235369A1 (en) | Distributing replication assignments among nodes | |
| Wang et al. | Research on the architecture of Open Education based on cloud computing | |
| Wei et al. | Towards a cloud storage data management model based on RNPT network |
Legal Events
| Date | Code | Title | Description |
|---|---|---|---|
| C06 | Publication | ||
| PB01 | Publication | ||
| WD01 | Invention patent application deemed withdrawn after publication |
Application publication date: 20131211 |
|
| WD01 | Invention patent application deemed withdrawn after publication |