CN102624911A - Cluster-based visible media storage system - Google Patents
Cluster-based visible media storage system Download PDFInfo
- Publication number
- CN102624911A CN102624911A CN2012100683254A CN201210068325A CN102624911A CN 102624911 A CN102624911 A CN 102624911A CN 2012100683254 A CN2012100683254 A CN 2012100683254A CN 201210068325 A CN201210068325 A CN 201210068325A CN 102624911 A CN102624911 A CN 102624911A
- Authority
- CN
- China
- Prior art keywords
- server
- backup
- data
- module
- cmss
- 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 discloses a cluster-based visible media storage system, which is divided into nine modules. A client comprises a user interface, an integrated file system (IFS) module, a client cache module and a network protocol module. A server comprises a member management module, a data backup module, a network protocol, a metadata manager and a server cache module. A highly-autonomous and self-maintaining cluster strategy is adopted for the system, so that the concurrency and extensibility of the system are improved; and in addition, the system has the characteristics of clear logical structure and capability of automatically balancing a load.
Description
Technical field
The present invention relates to the visual media technical field, relate in particular to a kind of visual media storage system based on cluster.
Background technology
Vision is the important means in the human cognitive world.The mankind receive information 83% from vision, the information of obtaining through vision is diversified.The visual media of being made up of these three kinds of media informations of image, video and digital geometry has become hot research in recent years.Visual media has unstructuredness, data volume is big, dimension is high and characteristics such as semantic diversity.The visual media information of magnanimity is Intelligent treatment mechanism for want of, has increased people's cognitive load, has reduced the efficient that people obtain knowledge, grasp rule and make a strategic decision.On the active computer architecture basics; Cognitive mechanism and artificial intelligence in conjunction with visual information; Seeking the new method of visual media Intelligent Information Processing, is the important directions that present academia makes great efforts, and people hope can bring whereby the breakthrough of the information processing technology.The fast development of information technology is that unprecedented condition has been created in deep development and extensive use, and the magnanimity growth of information is also handled information intelligent and proposed active demand.
Along with the explosive increase of multi-medium data on the network, a large amount of multi-medium datas produces all over the world and shares, thereby the demand of the extendible storage system of magnanimity is also increased fast.Existing memory system technologies receives unprecedented challenge.Because traditional system R (Relation Datatbase Management System; RDBMS) support target storage class; Simple object operating function weakness is difficult to competent also powerless for uncertain data to the management of semi-structured and non-structured large complicated data, far can not satisfy the demand of multi-medium data storage.Therefore on the relational data model basis to its expansion and improvement, form multiple multi-medium data memory model commonly used at present.
Exist two kinds of multi-medium data memory models commonly used at present: 1) OO relational data model.Adopt Object-oriented Technique and database technology to combine; And adopt Object-Oriented Programming Language (like C++, Java) to carry out the exploitation of DBAS on this basis, offer the function of database user with OO storage, management, retrieval.The positional information of storage multimedia object in relational database, promptly some additional descriptors such as the store path of keyword, text header, data and title as the index stores of media data in RDBMS.And actual multimedia object is stored in certain naming rule under the server designated directory of the separation outside the database, between the store path of object and keyword, sets up contact, through address link display media data.This mode with external file stores that multimedia object can not reduce performance of database but there are some shortcomings in this mode, leaves the outside of Database Systems in like multi-medium data, can't obtain the safeguard protection that database provides himself data.2) pure object-oriented database.Pure object-oriented database (Object-oriented Database; OODB) adopt brand-new data model and pattern; The database technology of doing object oriented programming model, method and mechanism and advanced person from the bottom organically combines, and utilizes the inquiry of ODQL language fulfillment database, and its function is on the basis of SQL function; Increased the mechanism of suitable information retrieval based on contents, like the information modeling ability of complicacy, the design of semantic pattern; The dynamic mode evolution of support target instance and class; Support the heavy construction transaction management; The version of object and configuration management; Inquiry of navigation formula and correlation inquiry visit; Support to be independent of in the distributed system storage of the large object of platform; Calculate complete database programming language.At present, there are ONTOS DB, Gemstone, Objectivity, ObjectStone, O2, Poet, Versant etc. in typical commercial OODB Object Oriented Data Base system.
The shortcoming of prior art:
1, OO relevant database exists that concurrency is low, network-adaptive property is poor; The visual media of magnanimity can not be accomplished good support; Multi-medium data leaves the outside of Database Systems in, can't obtain the safeguard protection that database provides himself data, and the data poor recovery; Anti-burst scarce capacity lacks load balance ability in distributed system.
2, the pure described OODB of object oriented programming model has clear in structure, efficient advantages of higher; But it is big that the method is implemented difficulty, and the data model of lack of uniform and solid theoretical foundation are compatible and portable poor; Incompatible with RDBMS; Do not support SQL statement, versatility is relatively poor, is difficult to accomplish general at present.
Summary of the invention
The objective of the invention is in order to overcome the defective of prior art, a kind of visual media storage system based on cluster is provided, improved the concurrency and the extensibility of system through this system, it is clear to have logical construction simultaneously, to load can automatic equalization characteristics.
A kind of visual media storage system based on cluster is characterized in that this system is made up of two kinds of clustered nodes, memory node and multimedia server node; The memory device of memory node tissue distribution formula provides an extendible storage volume to the user, is stored in the The data hierarchical directory structure tissue of this volume; Multimedia server visits CMSS through the CMSS Client Agent; The software module of this system is divided into 9 parts; Be that client comprises user interface, IFS module, client-cache module and network protocol module, server comprises member management module, data backup module, procotol, metadata management and server cache module; Wherein, the procotol of client and server has realized the communication function of client with server, is based on TCP/IP; Subscriber Interface Module SIM is for the routine interface through user's attitude, so that application program is utilized the characteristic of CMSS better; The CMSS client realizes that on windows platform the IFS interface that the IFS module utilizes Windows to provide provides the file access interface of a standard of application program, thereby the assurance application program is used native system pellucidly; Client-cache and server cache module are formed the two-level cache structure together; The metadata management module has realized the metadata management function of CMSS, and the member management module realizes the member management function of CMSS; The data backup module has realized data backup and the migration between the different disk server, thus the high reliability and the load balance of the system of realization.
Preferably; CMSS system user request processing flow is: the user sends out request through Client Agent to the logic basis data server; The storage server information that logic basis data server return data is deposited; Client Agent is sent out request from the corresponding storage server of trend more then, and last storage server returns the data of user's request.
Preferably, in CMSS, all storage servers regularly send a broadcasting heartbeat packet, and the member management device comes the activity of constructing Storage Server List through monitoring these heartbeat packet; If the member management device can not be received the heartbeat packet of a storage server in the regular hour, the member management device will be thought this fails, and it is deleted from the server list of oneself; Heartbeat packet also comprises information such as server metadata and storage resources operating position.
Preferably, have 3 kinds of incidents can cause the renewal of server list:
1) periodic refresh incident: each its metadata information of storage server periodic refresh; Because a memory node possibly change the information such as metadata of oneself; And other memory nodes possibly preserved is old information, therefore needs regular update, adopts 3 minutes update cycle;
2) memory node adds incoming event: when adding new memory node; New memory node can be broadcasted one and added incoming event; Other memory node is received this incident; Can upgrade own server list information, and own nodal information is fed back to new node, new node is also tabulated according to the memory node of feedback information generation oneself;
3) memory node disconnected event: find a node failure when member's manager, can broadcast a disconnected event, other memory nodes are received this incident, automatically the information of deletion failure node in the memory node tabulation.
Preferably, a backup manager BM (backup manage) is arranged on server, it accepts user's backup request, and according to the loading condition of whole system, creates the backup tasks processing threads in due course, starts backup procedure; Backup source BS (backup source) is the optical disc server that needs backup, and backup purpose BT (backup target) is the optical disc server that will backup to data, and it will provide enough spaces to deposit backed up data.
Preferably; In order to realize backing up 3 sides' collaborative work; At first need define a NBP (network backup protocol) agreement, it has stipulated each step of whole backup procedure, and has defined communication interface and the reciprocal process between them between BM, BS and the BT; The backup flow process of data is following 8 steps:
Step1: the backup tasks processing threads sends backup command to source command process thread;
Step2: source command process thread will be ordered and inserted command queue, and on response agent, wait for the result of source worker thread;
Step3: command queue reads corresponding order to the source worker thread from the source, and transmits this order to the purpose processing threads;
Step4: after the purpose processing threads receives order, handle accordingly, and return result and give the source worker thread;
Step5: the source worker thread is received result insertion response queue;
Step6: source command process thread reads result from response queue, and result is sent to the backup tasks processing threads;
Step7: the source worker thread arrives destination with transfer of data continuously;
Step8: in the process of transfer of data, the current state that source command process thread in time backs up to the report of backup tasks processing threads.
The beneficial effect that technical scheme of the present invention is brought:
1) the CMSS system is based on common PC and can be configured to a kind of cluster distributed network data base, need not purchase special sophisticated equipment and mainframe computer;
2) the CMSS system adopts to the multimedia application design, and the cluster strategy of have high degree of autonomy, safeguarding has certainly improved the concurrency and the extensibility of system, and it is clear to have logical construction simultaneously, to load can automatic equalization characteristics.
Description of drawings
In order to be illustrated more clearly in the embodiment of the invention or technical scheme of the prior art; To do to introduce simply to the accompanying drawing of required use in embodiment or the description of the Prior Art below; Obviously, the accompanying drawing in describing below only is some embodiments of the present invention, for those of ordinary skills; Under the prerequisite of not paying creative work, can also obtain other accompanying drawing according to these accompanying drawings.
Fig. 1 is the system construction drawing of CMSS among the present invention;
Fig. 2 is the system module structure chart of CMSS among the present invention;
Fig. 3 is name space directory tree instance graph among the present invention;
Fig. 4 is a data backup agreement flow chart among the present invention.
Embodiment
To combine the accompanying drawing in the embodiment of the invention below, the technical scheme in the embodiment of the invention is carried out clear, intactly description, obviously, described embodiment only is the present invention's part embodiment, rather than whole embodiment.Based on the embodiment among the present invention, those of ordinary skills are not making the every other embodiment that is obtained under the creative work prerequisite, all belong to the scope of the present invention's protection.
The object of the invention provides a High-Performance Computing Cluster multimedia storage system based on the PC cluster, and the cluster strategy that this system adopts high degree of autonomy, safeguards has certainly improved the concurrency and the extensibility of system; Adopt logic basis data management technique, server data redundancy technique, improve the Information Security of system and the load balancing between the settlement server.
The characteristics of multimedia application are that read request is far away more than writing request; Thereby to the storage data modification seldom; So cluster multimedia storage system (cluster multimedia storage system of the present invention's proposition; CMSS) adopt saveset group structure based on PC, the memory device of each node administration oneself, organizing in concert is an extendible high performance storage volume.Fig. 1 has provided the system construction drawing of CMSS.CMSS is made up of two kinds of clustered nodes; Memory node and multimedia server node.The memory device of memory node tissue distribution formula provides an extendible storage volume to the user; Be stored in the The data hierarchical directory structure tissue of this volume.Multimedia server visits the CMSS system through the CMSS Client Agent; Change and traditional all issue the management server mode to all users' requests; But let the user ask to be distributed to each server; Strengthen the degree of parallelism of each server so greatly, realized load balance, thereby improved the performance of whole cluster.
The CMSS system adopts high distributed strategy autonomous, that safeguard certainly, and each memory node can be safeguarded the metadata and the data itself of oneself storage resources and file independently, and file service can independently be provided, thereby improves the extensibility of system.Adopt data to place strategy, safeguard the load balance of each memory node with the technology that the data migration combines.
The CMSS system adopts intelligent data to place strategy; Operating position through data; Analyze correlation of data, employing will have the data of correlation to be placed on same memory node as far as possible, thereby reduce the data dependence between the memory node; Thereby the inefficacy of any node is reduced to minimum to the influence of other nodes, improves the availability of system.
CMSS system user request processing flow is shown in dotted line among Fig. 1; The user can send out request (step (1)) to the logic basis data server through Client Agent; The storage server information that logic basis data server return data is deposited (step (2)); Client Agent is sent out request (step (3)) from the corresponding storage server of trend more then, and last storage server returns the data (step (4)) of user's request.
The software configuration of CMSS system is as shown in Figure 2; Divide according to function; CMSS can be divided into 9 modules; Client comprises user interface, IFS module, client-cache (Cache) module and network protocol module, and server comprises member management module, data backup module, procotol, metadata management and server buffer (Cache) module.Wherein, the procotol of client and server has realized the communication function of client with server, is based on TCP/IP.Subscriber Interface Module SIM is for the routine interface through user's attitude, so that application program is utilized the characteristic of CMSS better.The CMSS system client realizes that on windows platform the IFS interface that the IFS module utilizes Windows to provide provides the file access interface of a standard of application program, thereby the assurance application program is used native system pellucidly.Client Cache and server Cache module are formed two-stage Cache structure together, are in order to improve the performance of system.The metadata management module has realized the metadata management function of CMSS system, and the member management module realizes the member management function of CMSS; The data backup module has realized data backup and the migration between the different disk server, thus the high reliability and the load balance of the system of realization.
Traditional SAN storage system has good expandability; But its meta data server is its bottleneck; To the characteristics of multi-medium data visit, the CMSS system adopts the logic basis data server, and (logical metadata server LMS) realizes the metadata management of whole system; Just adopt high autonomous cluster strategy from maintenance, each memory node can be safeguarded storage resources and metadata and the data itself of oneself independently; And stores service can independently be provided, meta data server is only managed the metadata between the different storage servers, thereby has alleviated the burden of meta data server, improves the performance of system.Adopt data to place strategy, safeguard the load balance of each storage server with the technology that the data migration combines.
LMS elects (for example, adopting the minimum server in IP address to serve as) by each server, and the name space metadata of its management gathers generation by each server.In case LMS lost efficacy, system will produce new LMS automatically, does not therefore have single point failure.LMS only preserve the first class catalogue of namespace and with parent directory not at the catalogue or the fileinfo of same server, the LMS global dictionary information of preserving is seldom and static relatively like this.In order to improve performance; LMS issues all memory nodes and user node with it after collecting and accomplishing all global dictionary information, and the user node Agent has this global dictionary information; The memory node at file that just can directly obtain serving and catalogue place need not visited LMS.
Data on each memory node exist with the form of a plurality of independently subtrees, and each memory node is formed a smaller catalogue listing with these subtree root node information.Each memory node is submitted to LMS with its catalogue listing, by the LMS unified management.Generate global dictionary tree, just a name space directory tree by LMS.Fig. 3 is an instance of name space directory tree, and the node that rectangle frame is represented is stored on the storage server 1, and the node that diamond is represented is stored on the storage server 2, and the node that the round rectangle frame table shows is stored on the storage server 3.
In CMSS; Storage Server List is as a soft state (with the Sorrento system similarity); The change that can conform automatically (like the inefficacy or the increase of node); System is based upon on the basis of mutual autonomy, and each node detects and the strategy of adjustment oneself, the resource virtualizing of cooperation completion system, data location and placement etc. together automatically each other.
In CMSS, all storage servers regularly send a broadcasting heartbeat packet, and the member management device comes the activity of constructing Storage Server List through monitoring these heartbeat packet.If the member management device can not be received the heartbeat packet of a storage server in the regular hour, the member management device will be thought this fails, and it is deleted from the server list of oneself.Heartbeat packet also comprises information such as server metadata and storage resources operating position.Have 3 kinds of incidents can cause the renewal of server list:
1) periodic refresh incident.Each its metadata information of storage server periodic refresh; Because a memory node possibly change the information such as metadata of oneself, be old information and other memory nodes possibly preserved, therefore need regular update; At our test environment, we adopt 3 minutes update cycle.
2) memory node adds incoming event.When adding new memory node; New memory node can be broadcasted one and added incoming event, and other memory node is received this incident, can upgrade the server list information of oneself; And oneself nodal information fed back to new node, new node also generates own memory node according to feedback information and tabulates.
3) memory node disconnected event.Find a node failure when member's manager, can broadcast a disconnected event, other memory nodes are received this incident, automatically the information of deletion failure node in the memory node tabulation.
In order to improve the reliability of whole C MSS system, also designed the data backup module in the CMSS system, it realizes the data backup between the different storage servers.It also realizes the data migration between the storage server simultaneously, and the storage server data that load is heavy are moved on the low weight storage server of load, thereby guarantees the load balance of whole C MSS system.Whole standby system is divided into 3 parts, and is as shown in Figure 4.A backup manager BM (backup manage) is arranged on server, and it accepts user's backup request, and according to the loading condition of whole system, creates the backup tasks processing threads in due course, starts backup procedure.Backup source BS (backup source) is the optical disc server that needs backup, and backup purpose BT (backup target) is the optical disc server that will backup to data, and it will provide enough spaces to deposit backed up data.
In order to realize backing up 3 sides' collaborative work; At first need define a NBP (network backup protocol) agreement; It has stipulated each step of whole backup procedure, and has defined communication interface and the reciprocal process between them between BM, BS and the BT.Through introducing whole backup procedure the NBP agreement is described below.
As shown in Figure 4, the backup flow process of data can be divided following 8 steps:
Step1: the backup tasks processing threads sends backup command to source command process thread;
Step2: source command process thread will be ordered and inserted command queue, and on response agent, wait for the result of source worker thread;
Step3: command queue reads corresponding order to the source worker thread from the source, and transmits this order to the purpose processing threads;
Step4: after the purpose processing threads receives order, handle accordingly, and return result and give the source worker thread;
Step5: the source worker thread is received result insertion response queue;
Step6: source command process thread reads result from response queue, and result is sent to the backup tasks processing threads;
Step7: the source worker thread arrives destination with transfer of data continuously;
Step8: in the process of transfer of data, the current state that source command process thread in time backs up to the report of backup tasks processing threads.
Generally speaking, the CMSS system mainly designs to multimedia application, adopts high cluster strategy autonomous, that safeguard certainly, has improved the concurrency and the extensibility of system; Adopt the logic basis data management, both realized single name space, avoided single point failure again.For load balance problem, the CMSS system has solved static load balancing through data migration between the server, guarantees the data volume balance on each server as far as possible.
The beneficial effect that technical scheme of the present invention is brought
1) the CMSS system is based on common PC and can be configured to a kind of cluster distributed network data base, need not purchase special sophisticated equipment and mainframe computer;
2) the CMSS system adopts to the multimedia application design, and the cluster strategy of have high degree of autonomy, safeguarding has certainly improved the concurrency and the extensibility of system, and it is clear to have logical construction simultaneously, to load can automatic equalization characteristics.
In the design of load balancing, what CMSS of the present invention system adopted is the static load balance policy, can guarantee the load balancing of system preferably.But gone out this kind method, can also adopt other dynamic load balancing to survey slightly, the load balancingization that will more effectively make whole system.But the dynamic equalization strategy also means more complicated design.
More than to a kind of visual media storage system that the embodiment of the invention provided based on cluster; Carried out detailed introduction; Used concrete example among this paper principle of the present invention and execution mode are set forth, the explanation of above embodiment just is used for helping to understand method of the present invention and core concept thereof; Simultaneously, for one of ordinary skill in the art, according to thought of the present invention, the part that on embodiment and range of application, all can change, in sum, this description should not be construed as limitation of the present invention.
Claims (6)
1. the visual media storage system based on cluster is characterized in that, this system is made up of two kinds of clustered nodes, memory node and multimedia server node; The memory device of memory node tissue distribution formula provides an extendible storage volume to the user, is stored in the The data hierarchical directory structure tissue of this volume; Multimedia server visits CMSS through the CMSS Client Agent; The software module of this system is divided into 9 parts; Be that client comprises user interface, IFS module, client-cache module and network protocol module, server comprises member management module, data backup module, procotol, metadata management and server cache module; Wherein, the procotol of client and server has realized the communication function of client with server, is based on TCP/IP; Subscriber Interface Module SIM is for the routine interface through user's attitude, so that application program is utilized the characteristic of CMSS better; The CMSS client realizes that on windows platform the IFS interface that the IFS module utilizes Windows to provide provides the file access interface of a standard of application program, thereby the assurance application program is used native system pellucidly; Client-cache and server cache module are formed the two-level cache structure together; The metadata management module has realized the metadata management function of CMSS, and the member management module realizes the member management function of CMSS; The data backup module has realized data backup and the migration between the different disk server, thus the high reliability and the load balance of the system of realization.
2. the system of claim 1; It is characterized in that; CMSS system user request processing flow is: the user sends out request through Client Agent to the logic basis data server; The storage server information that logic basis data server return data is deposited, Client Agent is sent out request from the corresponding storage server of trend more then, and last storage server returns the data of user's request.
3. the system of claim 1 is characterized in that, in CMSS, all storage servers regularly send a broadcasting heartbeat packet, and the member management device comes the activity of constructing Storage Server List through monitoring these heartbeat packet; If the member management device can not be received the heartbeat packet of a storage server in the regular hour, the member management device will be thought this fails, and it is deleted from the server list of oneself; Heartbeat packet also comprises information such as server metadata and storage resources operating position.
4. like claim 1 or 3 described systems, it is characterized in that having 3 kinds of incidents can cause the renewal of server list:
1) periodic refresh incident: each its metadata information of storage server periodic refresh; Because a memory node possibly change the information such as metadata of oneself; And other memory nodes possibly preserved is old information, therefore needs regular update, adopts 3 minutes update cycle;
2) memory node adds incoming event: when adding new memory node; New memory node can be broadcasted one and added incoming event; Other memory node is received this incident; Can upgrade own server list information, and own nodal information is fed back to new node, new node is also tabulated according to the memory node of feedback information generation oneself;
3) memory node disconnected event: find a node failure when member's manager, can broadcast a disconnected event, other memory nodes are received this incident, automatically the information of deletion failure node in the memory node tabulation.
5. the system of claim 1 is characterized in that, a backup manager BM (backup manage) is arranged on server; It accepts user's backup request; And, create the backup tasks processing threads in due course according to the loading condition of whole system, start backup procedure; Backup source BS (backup source) is the optical disc server that needs backup, and backup purpose BT (backup target) is the optical disc server that will backup to data, and it will provide enough spaces to deposit backed up data.
6. like claim 1 or 5 described systems; It is characterized in that; In order to realize backing up 3 sides' collaborative work; At first need define a NBP (network backup protocol) agreement, it has stipulated each step of whole backup procedure, and has defined communication interface and the reciprocal process between them between BM, BS and the BT; The backup flow process of data is following 8 steps:
Step1: the backup tasks processing threads sends backup command to source command process thread;
Step2: source command process thread will be ordered and inserted command queue, and on response agent, wait for the result of source worker thread;
Step3: command queue reads corresponding order to the source worker thread from the source, and transmits this order to the purpose processing threads;
Step4: after the purpose processing threads receives order, handle accordingly, and return result and give the source worker thread;
Step5: the source worker thread is received result insertion response queue;
Step6: source command process thread reads result from response queue, and result is sent to the backup tasks processing threads;
Step7: the source worker thread arrives destination with transfer of data continuously;
Step8: in the process of transfer of data, the current state that source command process thread in time backs up to the report of backup tasks processing threads.
Priority Applications (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN2012100683254A CN102624911A (en) | 2012-03-14 | 2012-03-14 | Cluster-based visible media storage system |
Applications Claiming Priority (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN2012100683254A CN102624911A (en) | 2012-03-14 | 2012-03-14 | Cluster-based visible media storage system |
Publications (1)
| Publication Number | Publication Date |
|---|---|
| CN102624911A true CN102624911A (en) | 2012-08-01 |
Family
ID=46564593
Family Applications (1)
| Application Number | Title | Priority Date | Filing Date |
|---|---|---|---|
| CN2012100683254A Pending CN102624911A (en) | 2012-03-14 | 2012-03-14 | Cluster-based visible media storage system |
Country Status (1)
| Country | Link |
|---|---|
| CN (1) | CN102624911A (en) |
Cited By (11)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN102855294A (en) * | 2012-08-13 | 2013-01-02 | 北京联创信安科技有限公司 | Intelligent hash data layout method, cluster storage system and method thereof |
| CN104580536A (en) * | 2015-02-06 | 2015-04-29 | 浪潮电子信息产业股份有限公司 | Metadata cluster load balance realization method |
| CN104683374A (en) * | 2013-11-28 | 2015-06-03 | 北京华云网际科技有限公司 | Cache update method for distributed storage and distributed storage system |
| CN105812425A (en) * | 2014-12-30 | 2016-07-27 | Tcl集团股份有限公司 | Method for establishing cross-room server cluster and node servers |
| CN106096038A (en) * | 2016-06-28 | 2016-11-09 | 浪潮软件股份有限公司 | A kind of distributed database architecture under cloud computing framework |
| CN107894874A (en) * | 2017-11-17 | 2018-04-10 | 北京联想超融合科技有限公司 | Data read-write control method, terminal and system based on super fusion storage system |
| CN109344011A (en) * | 2018-10-22 | 2019-02-15 | 郑州云海信息技术有限公司 | A kind of data back up method and device |
| CN110611699A (en) * | 2019-08-09 | 2019-12-24 | 南京泛函智能技术研究院有限公司 | Wireless self-organizing data storage method and data storage system |
| CN110690986A (en) * | 2018-07-04 | 2020-01-14 | 富士通株式会社 | Storage system, storage control method, and storage control apparatus |
| CN111488321A (en) * | 2020-03-05 | 2020-08-04 | 北京联创信安科技股份有限公司 | Management system for storage volume |
| CN112181950A (en) * | 2020-10-19 | 2021-01-05 | 北京米连科技有限公司 | Method for constructing distributed object database |
Citations (5)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN1601485A (en) * | 2004-10-15 | 2005-03-30 | 清华大学 | Network optical disc database |
| CN1801923A (en) * | 2005-01-05 | 2006-07-12 | 中央电视台 | Graded memory management system |
| US20070256081A1 (en) * | 2006-04-28 | 2007-11-01 | Michael Comer | System and method for management of jobs in a cluster environment |
| CN101299695A (en) * | 2007-04-30 | 2008-11-05 | 北京亚细亚智业科技有限公司 | Cluster type multilevel storage management system and method based on Ethernet |
| CN101252603B (en) * | 2008-04-11 | 2011-03-30 | 清华大学 | Cluster distributed type lock management method based on storage area network SAN |
-
2012
- 2012-03-14 CN CN2012100683254A patent/CN102624911A/en active Pending
Patent Citations (5)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN1601485A (en) * | 2004-10-15 | 2005-03-30 | 清华大学 | Network optical disc database |
| CN1801923A (en) * | 2005-01-05 | 2006-07-12 | 中央电视台 | Graded memory management system |
| US20070256081A1 (en) * | 2006-04-28 | 2007-11-01 | Michael Comer | System and method for management of jobs in a cluster environment |
| CN101299695A (en) * | 2007-04-30 | 2008-11-05 | 北京亚细亚智业科技有限公司 | Cluster type multilevel storage management system and method based on Ethernet |
| CN101252603B (en) * | 2008-04-11 | 2011-03-30 | 清华大学 | Cluster distributed type lock management method based on storage area network SAN |
Non-Patent Citations (1)
| Title |
|---|
| 万继光、谢长生: "一种集群多媒体存储***的研究与设计", 《计算机研究与发展》 * |
Cited By (18)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN102855294A (en) * | 2012-08-13 | 2013-01-02 | 北京联创信安科技有限公司 | Intelligent hash data layout method, cluster storage system and method thereof |
| CN102855294B (en) * | 2012-08-13 | 2016-12-21 | 北京联创信安科技股份有限公司 | Intelligent hash data layout method, cluster storage system and method thereof |
| CN104683374B (en) * | 2013-11-28 | 2019-03-08 | 北京华云网际科技有限公司 | A kind of buffering updating method of distributed storage |
| CN104683374A (en) * | 2013-11-28 | 2015-06-03 | 北京华云网际科技有限公司 | Cache update method for distributed storage and distributed storage system |
| CN105812425A (en) * | 2014-12-30 | 2016-07-27 | Tcl集团股份有限公司 | Method for establishing cross-room server cluster and node servers |
| CN105812425B (en) * | 2014-12-30 | 2019-04-19 | Tcl集团股份有限公司 | A kind of method and node server of the establishment across server in machine room cluster |
| CN104580536A (en) * | 2015-02-06 | 2015-04-29 | 浪潮电子信息产业股份有限公司 | Metadata cluster load balance realization method |
| CN106096038A (en) * | 2016-06-28 | 2016-11-09 | 浪潮软件股份有限公司 | A kind of distributed database architecture under cloud computing framework |
| CN107894874A (en) * | 2017-11-17 | 2018-04-10 | 北京联想超融合科技有限公司 | Data read-write control method, terminal and system based on super fusion storage system |
| CN107894874B (en) * | 2017-11-17 | 2020-08-21 | 北京联想超融合科技有限公司 | Data read-write control method, terminal and system based on super-fusion storage system |
| CN110690986A (en) * | 2018-07-04 | 2020-01-14 | 富士通株式会社 | Storage system, storage control method, and storage control apparatus |
| CN109344011A (en) * | 2018-10-22 | 2019-02-15 | 郑州云海信息技术有限公司 | A kind of data back up method and device |
| CN109344011B (en) * | 2018-10-22 | 2022-03-08 | 郑州云海信息技术有限公司 | Data backup method and device |
| CN110611699A (en) * | 2019-08-09 | 2019-12-24 | 南京泛函智能技术研究院有限公司 | Wireless self-organizing data storage method and data storage system |
| CN110611699B (en) * | 2019-08-09 | 2022-06-24 | 南京泛函智能技术研究院有限公司 | Wireless self-organizing data storage method and data storage system |
| CN111488321A (en) * | 2020-03-05 | 2020-08-04 | 北京联创信安科技股份有限公司 | Management system for storage volume |
| CN112181950A (en) * | 2020-10-19 | 2021-01-05 | 北京米连科技有限公司 | Method for constructing distributed object database |
| CN112181950B (en) * | 2020-10-19 | 2024-03-26 | 北京米连科技有限公司 | Construction method of distributed object database |
Similar Documents
| Publication | Publication Date | Title |
|---|---|---|
| CN102624911A (en) | Cluster-based visible media storage system | |
| US11030185B2 (en) | Schema-agnostic indexing of distributed databases | |
| CN102541990B (en) | Database redistribution method and system utilizing virtual partitions | |
| US11068501B2 (en) | Single phase transaction commits for distributed database transactions | |
| US8862540B2 (en) | Replica placement strategy for distributed data persistence | |
| JP6009097B2 (en) | Separation of content and metadata in a distributed object storage ecosystem | |
| Zhao et al. | Distributed data provenance for large-scale data-intensive computing | |
| CN102855239B (en) | A kind of distributed geographical file system | |
| CN105190623A (en) | Log record management | |
| US10572466B2 (en) | Multiple collections of user-defined metadata for self-describing objects | |
| JP2012504262A (en) | Distributed cache placement | |
| CN102394923A (en) | Cloud system platform based on n*n display structure | |
| CN102708158B (en) | PostgreSQL (postgres structured query language) cloud storage filing and scheduling system | |
| CN104142957A (en) | Method and system for regional medical treatment-orientated data sharing | |
| CN103458044A (en) | Metadata sharing management method for multi-storage clusters under wide area network environment | |
| CN103488546A (en) | Backup and recovery method capable of supporting online concurrency of multi-level data and database | |
| Xiong et al. | Data vitalization: a new paradigm for large-scale dataset analysis | |
| Vogt et al. | Polypheny-DB: towards a distributed and self-adaptive polystore | |
| WO2020019313A1 (en) | Graph data updating method, system, computer readable storage medium, and device | |
| CN109783018A (en) | A kind of method and device of data storage | |
| CN102937964A (en) | Intelligent data service method based on distributed system | |
| CN102012946A (en) | High-efficiency safety monitoring video/image data storage method | |
| US20160275085A1 (en) | Methods for facilitating a nosql database with integrated management and devices thereof | |
| CN101763390A (en) | Database storing system and method based on Berkeley DB | |
| Agrawal et al. | Mobile data sync in a blink |
Legal Events
| Date | Code | Title | Description |
|---|---|---|---|
| C06 | Publication | ||
| PB01 | Publication | ||
| C10 | Entry into substantive examination | ||
| SE01 | Entry into force of request for substantive examination | ||
| C12 | Rejection of a patent application after its publication | ||
| RJ01 | Rejection of invention patent application after publication |
Application publication date: 20120801 |