CN105282230B - A kind of hard real time computer cluster dynamic scheduling system - Google Patents
A kind of hard real time computer cluster dynamic scheduling system Download PDFInfo
- Publication number
- CN105282230B CN105282230B CN201510579543.8A CN201510579543A CN105282230B CN 105282230 B CN105282230 B CN 105282230B CN 201510579543 A CN201510579543 A CN 201510579543A CN 105282230 B CN105282230 B CN 105282230B
- Authority
- CN
- China
- Prior art keywords
- layer
- management
- server
- running
- cluster
- 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.)
- Expired - Fee Related
Links
Classifications
-
- H—ELECTRICITY
- H04—ELECTRIC COMMUNICATION TECHNIQUE
- H04L—TRANSMISSION OF DIGITAL INFORMATION, e.g. TELEGRAPHIC COMMUNICATION
- H04L67/00—Network arrangements or protocols for supporting network services or applications
- H04L67/01—Protocols
- H04L67/10—Protocols in which an application is distributed across nodes in the network
Abstract
A kind of hard real time computer cluster dynamic scheduling system is claimed in the present invention; mainly solve the high reliability of the redundancy and operation designed in information processing application system; solve the Single Point of Faliure easily occurred in software and hardware architecture structure design, safeguards system stable operation.Main contents：1. hard real time computer cluster dynamic dispatching software and hardware modelling.2. Dynamic Scheduling Strategy designs, mainly include：(1) the soft duplexing design of dynamic, the soft duplex of dynamic is to realize most basic, the crucial unit of real-time computer cluster dynamic dispatching, is the basic guarantee of information processing system computer cluster operation stability；(2) operation dynamic allocation scheme modelling, " operation " are exactly the application software operated in server cluster, and the equalization that operation distributes in server cluster is to improve the key of server operational efficiency and stability；(3) design on the soft duplexing switchover policy of the hard real-time Main Basiss of computer cluster dynamic dispatching is ensured.
Description
Technical field
The present invention designs large and medium-sized hard real time, the data processing system of high reliability that space launch application field is related to
System, is specifically a kind of hard real time computer cluster dynamic scheduling system.
Background technology
In real-time, reliability and the high information processing application system of security requirement, at least need to meet three key elements
Ensure：First, the design of whole system can embody higher redundant ability, any one section in information processing system
Point is both needed to ensure without any Single Point of Faliure, if an information processing system since the defects of design, produces Single Point of Faliure, to whole
A system be all mortality and be likely at any time since the defects of Single Point of Faliure has influenced the normal operation of whole system；
Second, the operational efficiency of lifting whole system and emergent level, whether efficient to the operation of one large and medium-sized information system and emergent water
The flat target for whether embodying " timely, accurate ", key are which type of model is system establish and use which type of technology
Route；Third, automation and the management level of whole information processing application system are improved, using by distributing monitoring means transition
To centralized monitoring, three-dimensional ground collection system running state information, warning information.It is combined using automatic and human assistance
Mode, is greatly improved the automatization level of information processing system.For above application demand, a kind of hard real time computer is devised
Cluster Dynamic Scheduling Strategy, sets from software and hardware architecture structural model, the soft duplexing design of dynamic, operation dynamic allocation scheme respectively
Count and the resource overhead of " lightweight " ensure that hard real-time, security and high reliability of system etc. in duplexing handoff procedure
Feature.
The content of the invention
In view of the deficiencies of the prior art, it is proposed that the higher strong reality of a kind of hard real-time, high reliability, security requirement
When computer cluster dynamic scheduling system.Technical scheme is as follows：A kind of hard real time computer cluster dynamic dispatching system
System, it includes：Hardware architecture model and software architecture model；Wherein hardware architecture model includes three layers of hardware
System model and three-layer network model, the software architecture model include management and running operation control terminal layer, management and running layer
And scheduling broker layer；
The hardware architecture model：For building the hardware support platform of dynamic dispatching, including three layers of hardware systems
Model and three-layer network model, the first layer of three layers of hardware systems model is management and running control terminal layer：For to computer
Management server layer carries out terminal control；The second layer is management and running server layer：For the clothes initiated computer cluster layer
Business is managed scheduling；Third layer is computer cluster layer, including some servers；
The three-layer network model includes management and running terminal access layer, management and running convergence-level and scheduling broker core
Layer, wherein scheduling broker core layer is connected with the computer cluster layer of three layers of hardware systems model by interchanger, the tune
Degree management server layer is connected by interchanger with management and running convergence-level, and the management and running terminal access layer passes through exchange
Machine is connected with management and running control terminal layer；
The software architecture model includes three layers of software model：Management and running operation control terminal layer, management and running layer and
Scheduling broker layer；Wherein management and running operation control terminal：It is mainly responsible for the application to service, the startup of service, software and hardware resources
System monitoring, soft duplexing switching at runtime, the manipulation of operation, that is, process, the control of operation flow, letter are run in server cluster
The emergency measure of breath system and the release of service；Management and running layer：It is mainly responsible for the state inspection that works online to server cluster
Survey, to the soft duplexing distribution of dynamic between server cluster, to operating in the operation in server cluster i.e. course allocation strategy,
The synchronization of data and state between management and running server, operational order, server cluster software and hardware resources statistics and time card
Forwarding；Scheduling broker layer is mainly responsible for the execution to being ordered between the operation i.e. process run in server cluster and returns card
Interaction, collects cluster internal system hardware and software resource operating status, collects turn of job run state, network registry and message
Hair, the information exchange between operation.
Further, the management and running control terminal layer is made of two management and running control terminals, and two scheduling
It is hot standby each other to manage control terminal；The management and running server layer is by two stylobates in the scheduling being managed to server cluster
Management server forms, and two-server is hot standby each other.
Further, the scheduling broker core layer is two with the number of units that computer cluster layer is connected by interchanger
Platform, and Hot Spare each other；The interchanger that the management and running server layer is connected by interchanger with management and running convergence-level
Number of units is two and Hot Spare, the management and running terminal access layer pass through interchanger and management and running control terminal layer phase each other
The interchanger number of units of connection is two and Hot Spare each other.
Further, the separate mesh clamping where every server in some servers of the computer cluster layer
Enter an interchanger, independent networking forms Small-scale LAN.
Further, pass through between the management and running operation control terminal layer and management and running layer of the software architecture model
Transmission Control Protocol is point-to-point to communicate, and is communicated between management and running layer and scheduling broker by udp protocol Any-Source Multicast.
Further, in the server cluster of the computer cluster layer, it is intended to for identical business game, that is, process
Ask and be present in using the double soft engineering methods of dynamic, i.e., identical business game, that is, process in host and slave two-server, Ren Heyi
When the software and hardware of platform server breaks down, the mode that system can perform automatic and artificial two kinds of combinations carries out duplexing switching,.
Further, the double soft engineering methods of the dynamic are specially：
It is assumed that some information processing application system server cluster has m platform line servers, operation in the cluster has n
Operation, the principle of distribution is that requirement each operation must be run in two-server, if cluster server number is odd number,
So m platforms server exists with unit state, as the redundancy idle server of system, if n<=m/2, at least requires every
A operation can independently be run in two servers, if n>M/2 is low for D grades security level according to the safe class of operation
Operation be combined and operate in together in the server of Hot Spare each other, security level for A grades, B grades, C grades and to soft
The high operation of hardware resource requirements is individually operated in the server of Hot Spare each other.
Advantages of the present invention and have the beneficial effect that：
Hardware structure model is designed as three layers of service device model, and network structure model is also designed as three layer model.It is each
Layer operation control terminal, management and running server, interchanger are all Hot Spares, and network interface card access switch mode uses the virtual skill of double netcard
Art.Each system node embodies Redundancy Design without Single Point of Faliure, each node.
Software model is divided according to three-decker, and each layer is without Single Point of Faliure, and the division of labor is clear and definite between layers, scheduling
Communicate between management operation control terminal layer and management and running layer by the way that Transmission Control Protocol is point-to-point, management and running layer and scheduling broker
Between communicated by udp protocol Any-Source Multicast.Add between layers information return card mechanism, ensure that order with
The safety and reliability of message transmissions.
In server cluster, it is required to be present in two services of host and slave for identical business game (process)
In device, when the software and hardware of any server breaks down, system can perform automatic and artificial two kinds of modes combined into
Row duplex switching, ensure that the reliability of system.
Allocation strategy of the operation (process) in server cluster can ensure that each operation (process) can operate in
Any two each other in the server of Hot Spare, ensure that the stability and reliability of job run.
The present invention is suitable for real-time, reliability, security and the data more demanding to system redundancy designed capacity
Processing system.Its effect is illustrated by prototype of space launching site data processing centre system Construction, server cluster shares
10 medium-sized data processing servers, operation share 5 data processing configuration items and an agent component composition, each number of units evidence
The software of the static deployment of processing server is just the same, and whether the operation of software and basic of distribution server are presence conduct
Foundation, is allocated in the cluster formed for 5 operations in 1 to 10 server.Can be any group per two-server
Into duplex, strengthen the real-time and reliability of its system.
Brief description of the drawings
Fig. 1 is that the present invention provides preferred embodiment real-time server colony dispatching strategy hardware architecture illustraton of model；
Fig. 2 real-time server colony dispatching tactical software illustratons of model；
The operation allocation strategy flow chart of Fig. 3 real-time server colony dispatching strategies；
The soft duplexing switching flow figure of Fig. 4 real-time server colony dispatching strategies.
Embodiment
Below in conjunction with attached drawing, the invention will be further described：
The software and hardware architecture structural model that the purpose of the present invention is be related to computer cluster Dynamic Scheduling Strategy, dynamic
The technical problem such as soft duplex, the guarantee of operation dynamically distributes model and real-time is described.To hard real-time, high reliability,
The higher information system of security requirement, there is reference and reference significance very much.Related to especially for as space launch application field
And arrive large and medium-sized hard real time, the data handling system of high reliability, more meet current demand, its main contents：
1st, real-time computer cluster dynamic dispatching software and hardware modelling
(1) hardware architecture model
1. three layers of hardware systems model
Three layers of hardware systems model：First layer is made of two management and running control terminals, and two operation control terminals are warm each other
It is standby；The second layer is made of two stylobates in the management and running server being managed to server cluster, and two-server is warm each other
It is standby；Third layer is computer cluster, is made of more platform servers.(Fig. 1)
2. the network model of Three-tider architecture
Network model：Three-layer network system model, is respectively management and running terminal access layer, management and running convergence-level and tune
Spend Broker Core layer.The access of two management and running terminals is by the access-layer switch of two Hot Spares each other, two management and running
By the convergence layer switch of two Hot Spares each other, each server in computer cluster accessed by two for server access
The core layer switch of Hot Spare each other.In addition each server in server cluster is clamped into a friendship with separate mesh
Change planes, independent networking forms LAN, is mainly used for duplexing heartbeat message and exchanges use.Wherein, connect per station terminal with server
It is all to use double netcard binding technology to enter interchanger.
3. the effect of the method
Hardware structure model is designed as three layers of service device model, and network structure model is also designed as three layer model.It is each
Layer operation control terminal, management and running server, interchanger are all Hot Spares, and network interface card access switch mode uses the virtual skill of double netcard
Art.Each system node embodies Redundancy Design without Single Point of Faliure, each node.
(2) software architecture model
1. three layers of software model
Three layers of software model：Management and running operation control terminal layer, management and running layer and scheduling broker layer.(Fig. 2)
Management and running operation control terminal is mainly responsible for the application of service, the startup of service, and the system monitoring of software and hardware resources is soft
The switching at runtime of duplex, the manipulation of server cluster operation operation (process), the control of operation flow, information processing system are answered
The functions such as anxious and service release.
Management and running layer is mainly responsible for the detection for the state that works online to server cluster, to dynamic between server cluster
Soft duplexing distribution, the allocation strategy to operating in operation in server cluster (process) operation, between management and running server
The synchronization of data and state, the statistics of all software and hardware resources and returns the functions such as the forwarding of card in operational order, server cluster.
Scheduling broker layer is mainly responsible for the execution to being ordered between the operation (process) that is run in server cluster and returns card
Return, collect server cluster inside system hardware and software resource operating status, collect operation (process) operating status, network
Registration and the forwarding of message, the information exchange (shared drive, pipeline, signal, message etc.) between operation (process).
(3) the characteristics of model
Software model is divided according to three-decker, and each layer is without Single Point of Faliure, and the division of labor is clear and definite between layers, scheduling
Communicate between management operation control terminal layer and management and running layer by the way that Transmission Control Protocol is point-to-point, management and running layer and scheduling broker
Between communicated by udp protocol Any-Source Multicast.Add between layers information return card mechanism, ensure that order with
The safety and reliability of message transmissions.
2nd, Dynamic Scheduling Strategy designs
(1) the soft duplex implementation of dynamic
Dynamically soft duplex is to realize the most basic unit of real-time computer cluster dynamic dispatching, is information processing system statistics
The basic guarantee of calculation machine cluster operation stability.
1. realization principle
Dynamically soft duplex forms duplex mainly for being randomly assigned two-server in server cluster.Server set
In the process of running, every server can be obtained with states such as " host, standby host, unit and off lines " group.Host and standby host it
Between status information by local area network carry out periodically interact heartbeat message.Switching between standby machine mainly passes through system
Two kinds of forms of automatic switchover and manual switching.
(2) the dynamic allocation scheme modelling of operation
" operation " is exactly the application software of operation in the server, and the equalization that operation distributes in server cluster is to carry
The efficiency of high server operation and the key of stability.
1. realization principle
It is assumed that some information processing application system server cluster has m platform line servers, operation in the cluster has n
Operation (process), the principle of distribution are that requirement each operation (process) must be run (if cluster service in two-server
Device number is odd number, then m platforms server exists with unit state, the redundancy idle server as system).If n<=
M/2, at least requires each operation independently to be run in two servers.If n>M/2, can be according to operation (process) scale
Size, requires software and hardware resources small operation to be combined the server for operating in Hot Spare each other together business scale
In, business scale is big and require high operation (process) isolated operation in the server of Hot Spare each other to software and hardware resources
In.
(3) computer cluster dynamic dispatching Real-time Ability problem
The guarantee of real-time is mainly reflected in the innovation on duplexing switchover policy, and dynamic soft duplexing switching time is in 100us
Within, major-minor machine state can be obtained by the change of state.For based on tcp stream protocols, management and running operation control terminal layer,
The content transmitting-receivings such as order, message, time card between management and running layer, Agent layer use data-driven strategy, and time delay is no more than
2ms.These parameters ensure that the hard real-time of dynamic dispatching.
(4) the characteristics of Dynamic Scheduling Strategy
In server cluster, it is required to be present in two services of host and slave for identical business game (process)
In device, when the software and hardware of any server breaks down, system can perform automatic and artificial two kinds of modes combined into
Row duplex switching, ensure that the reliability of system.
Allocation strategy of the operation (process) in server cluster can ensure that each operation (process) can operate in
Any two each other in the server of Hot Spare, ensure that the stability and reliability of job run.
To hard real time, the higher large and medium-sized information processing system of reliability requirement, according to Fig. 1 model integrated hardware system
System.Fig. 2 is designed according to management and running operation control terminal layer, three layers of software architecture of management and running layer and Agent layer.In server
Equilibrium assignment is carried out to some operations according to Fig. 3 to the distribution of operation in cluster.The real-time and reliability of information system is according to Fig. 4
Carry out soft duplex switching, it is only necessary to which the overhead of lightweight carries out Real-time Ability.
Foundation Fig. 1, Fig. 2, Fig. 3, Fig. 4 progress hardware environment integrates and the design of software architecture, in server cluster
In realize to the allocation strategy of operation and the soft duplexing realization of dynamic.The school that the system satisfactorily completes measuring and controlling equipment flies, CZ-
7 rehearsal, CZ-5 rehearsal, three phase of moon exploration program return flight test mission, Fengyun II, Big Dipper special project test satellite etc. are multinomial heavy
Big task, security, stability, the high reliability of system are examined.
The above embodiment is interpreted as being merely to illustrate the present invention rather than limits the scope of the invention.
After the content for having read the record of the present invention, technical staff can make various changes or modifications the present invention, these equivalent changes
Change and modification equally falls into the scope of the claims in the present invention.
Claims (7)
- A kind of 1. hard real time computer cluster dynamic scheduling system, it is characterised in that including：Hardware architecture model and software Architectural model；Wherein hardware architecture model includes three layers of hardware systems model and three-layer network model, the software Architectural model includes management and running operation control terminal layer, management and running layer and scheduling broker layer；The hardware architecture model：For building the hardware support platform of dynamic dispatching, including three layers of hardware systems model With three-layer network model, the first layer of three layers of hardware systems model is management and running control terminal layer：For to computer management Server layer carries out terminal control；The second layer is management and running server layer：For to the service that computer cluster layer is initiated into Row management and dispatching；Third layer is computer cluster layer, including some servers；The three-layer network model includes management and running terminal access layer, management and running convergence-level and scheduling broker core layer, its Middle scheduling broker core layer is connected with the computer cluster layer of three layers of hardware systems model by interchanger, the management and running Server layer is connected by interchanger with management and running convergence-level, and the management and running terminal access layer is by interchanger with adjusting Degree management control terminal layer is connected；The software architecture model includes three layers of software model：Management and running operation control terminal layer, management and running layer and scheduling Agent layer；Wherein management and running operation control terminal：It is mainly responsible for the application to service, the startup of service, the system of software and hardware resources Monitor, soft duplexing switching at runtime, the manipulation of operation, that is, process, the control of operation flow, information system are run in server cluster The emergency measure of system and the release of service；Management and running layer：The state-detection that works online to server cluster is mainly responsible for, it is right The soft duplexing distribution of dynamic between server cluster, to operating in the operation in server cluster i.e. course allocation strategy, scheduling The synchronization of data and state between management server, operational order, server cluster software and hardware resources statistics and turn for returning card Hair；Scheduling broker layer is mainly responsible for the execution to being ordered between the operation i.e. process run in server cluster and returns the friendship of card Mutually, collection cluster internal system hardware and software resource operating status, collection job run state, the forwarding of network registry and message, Information exchange between operation.
- A kind of 2. hard real time computer cluster dynamic scheduling system according to claim 1, it is characterised in that the scheduling Management control terminal layer is made of two management and running control terminals, and two management and running control terminals are hot standby each other；It is described Management and running server layer is made of two stylobates in the management and running server being managed to server cluster, and two services Device is hot standby each other.
- A kind of 3. hard real time computer cluster dynamic scheduling system according to claim 1, it is characterised in that the scheduling Broker Core layer is two with the number of units that computer cluster layer is connected by interchanger by interchanger, and Hot Spare each other； The interchanger number of units that the management and running server layer is connected with management and running convergence-level by interchanger is for two and each other Hot Spare, the interchanger number of units that the management and running terminal access layer is connected by interchanger with management and running control terminal layer For two and each other Hot Spare.
- A kind of 4. hard real time computer cluster dynamic scheduling system according to claim 1, it is characterised in that the calculating The separate mesh where every server in some servers of machine cluster layer is clamped to be formed into an interchanger, independent networking Small-scale LAN.
- A kind of 5. hard real time computer cluster dynamic scheduling system according to claim 1, it is characterised in that the software Communicate, adjust by the way that Transmission Control Protocol is point-to-point between the management and running operation control terminal layer of architectural model and management and running layer Communicated between degree management level and scheduling broker by udp protocol Any-Source Multicast.
- A kind of 6. hard real time computer cluster dynamic scheduling system according to claim 1, it is characterised in that the calculating In the server cluster of machine cluster layer, be required to for identical business game, that is, process using the double soft engineering methods of dynamic, i.e., it is identical Business game, that is, process be present in host and slave two-server, the software and hardware of any server breaks down When, the mode that system can perform automatic and artificial two kinds of combinations carries out duplexing switching.
- A kind of 7. hard real time computer cluster dynamic scheduling system according to claim 6, it is characterised in that the dynamic Double soft engineering methods are specially：It is assumed that some information processing application system server cluster has m platform line servers, operation in the cluster has n work Industry, the principle of distribution is that requirement each operation must be run in two-server, if cluster server number is odd number, that M platforms server exists with unit state, as the redundancy idle server of system, if n<=m/2, at least requires each Operation can independently be run in two servers, if n>M/2, according to the safe class of operation, business scale to software and hardware The small operation of resource requirement is combined to be operated in the server of Hot Spare each other together, business scale greatly and to soft or hard The high operation of part resource requirement is individually operated in the server of Hot Spare each other.
Priority Applications (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN201510579543.8A CN105282230B (en) | 2015-09-11 | 2015-09-11 | A kind of hard real time computer cluster dynamic scheduling system |
Applications Claiming Priority (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN201510579543.8A CN105282230B (en) | 2015-09-11 | 2015-09-11 | A kind of hard real time computer cluster dynamic scheduling system |
Publications (2)
| Publication Number | Publication Date |
|---|---|
| CN105282230A CN105282230A (en) | 2016-01-27 |
| CN105282230B true CN105282230B (en) | 2018-05-04 |
Family
ID=55150535
Family Applications (1)
| Application Number | Title | Priority Date | Filing Date |
|---|---|---|---|
| CN201510579543.8A Expired - Fee Related CN105282230B (en) | 2015-09-11 | 2015-09-11 | A kind of hard real time computer cluster dynamic scheduling system |
Country Status (1)
| Country | Link |
|---|---|
| CN (1) | CN105282230B (en) |
Families Citing this family (2)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN110062032A (en) * | 2019-04-01 | 2019-07-26 | 北京航天发射技术研究所 | A kind of computer network clustered control platform application reconfiguration system |
| CN113220448B (en) * | 2021-04-15 | 2023-02-17 | 广州广哈通信股份有限公司 | Dual-computer system and control method thereof |
Citations (6)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN101694631A (en) * | 2009-09-30 | 2010-04-14 | 曙光信息产业(北京)有限公司 | Real-time operation dispatching system and method thereof |
| WO2010085298A1 (en) * | 2009-01-21 | 2010-07-29 | Raytheon Company | Communication scheduling of network nodes using a cluster coefficient |
| CN101938504A (en) * | 2009-06-30 | 2011-01-05 | 深圳市融创天下科技发展有限公司 | Cluster server intelligent dispatching method and system |
| CN102083007A (en) * | 2011-01-25 | 2011-06-01 | 上海交通大学 | GSM-R (Global System for Mobile Communications-Railway) cluster scheduling service analysis system and implementing method thereof |
| CN102457814A (en) * | 2010-10-22 | 2012-05-16 | 中兴通讯股份有限公司 | Cluster dispatching method and system |
| CN102629906A (en) * | 2012-03-30 | 2012-08-08 | 浪潮电子信息产业股份有限公司 | Design method for improving cluster business availability by using cluster management node as two computers |
Family Cites Families (1)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| US20070233835A1 (en) * | 2006-03-31 | 2007-10-04 | Nandakishore Kushalnagar | Methodology for scheduling data transfers from nodes using path information |
-
2015
- 2015-09-11 CN CN201510579543.8A patent/CN105282230B/en not_active Expired - Fee Related
Patent Citations (6)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| WO2010085298A1 (en) * | 2009-01-21 | 2010-07-29 | Raytheon Company | Communication scheduling of network nodes using a cluster coefficient |
| CN101938504A (en) * | 2009-06-30 | 2011-01-05 | 深圳市融创天下科技发展有限公司 | Cluster server intelligent dispatching method and system |
| CN101694631A (en) * | 2009-09-30 | 2010-04-14 | 曙光信息产业(北京)有限公司 | Real-time operation dispatching system and method thereof |
| CN102457814A (en) * | 2010-10-22 | 2012-05-16 | 中兴通讯股份有限公司 | Cluster dispatching method and system |
| CN102083007A (en) * | 2011-01-25 | 2011-06-01 | 上海交通大学 | GSM-R (Global System for Mobile Communications-Railway) cluster scheduling service analysis system and implementing method thereof |
| CN102629906A (en) * | 2012-03-30 | 2012-08-08 | 浪潮电子信息产业股份有限公司 | Design method for improving cluster business availability by using cluster management node as two computers |
Also Published As
| Publication number | Publication date |
|---|---|
| CN105282230A (en) | 2016-01-27 |
Similar Documents
| Publication | Publication Date | Title |
|---|---|---|
| Xu et al. | Survivable virtual infrastructure mapping in virtualized data centers | |
| CN107707393B (en) | Multi-active system based on Openstack O version characteristics | |
| CN106790436B (en) | Traffic system monitoring method based on cloud architecture and control center cloud server | |
| CN102164056B (en) | Stacked link aggregation fault detection method and stacked devices | |
| CN101266633B (en) | Seamless super large scale dummy game world platform | |
| CN104980361B (en) | A kind of load-balancing method, apparatus and system | |
| CN104301391B (en) | Multi-area optical network data center resource virtualizes mapping method | |
| Xu et al. | Enhancing survivability in virtualized data centers: A service-aware approach | |
| CN107276662A (en) | A kind of software definition Information Network multi-controller dynamic deployment method | |
| CN101883029A (en) | Application implantation method and system in the cloud | |
| CN102486739A (en) | Method and system for distributing data in high-performance computer cluster | |
| CN103444256A (en) | Self-organization of a satellite grid | |
| CN109005108A (en) | A kind of spatial network route planning method and system | |
| CN105282230B (en) | A kind of hard real time computer cluster dynamic scheduling system | |
| CN106603696A (en) | High-availability system based on hyper-converged infrastructure | |
| CN105959347A (en) | All-digital collaborative simulation system communication method based on P2P network | |
| CN101702678B (en) | Star-shaped stack system and method thereof | |
| CN103106126A (en) | High-availability computer system based on virtualization | |
| CN114337783B (en) | Space distributed edge computing device and business processing method | |
| CN104123183B (en) | Cluster job scheduling method and apparatus | |
| CN106230622A (en) | A kind of cluster implementation method and device | |
| CN106961700B (en) | Wireless communication method for dynamic remote fault-tolerant reconstruction of cluster avionics system computing resources | |
| Tang et al. | Slicing-based software-defined mobile edge computing in the air | |
| CN106250048B (en) | Manage the method and device of storage array | |
| CN107682410A (en) | A kind of control method and device of distributed SDN controllers cluster |
Legal Events
| Date | Code | Title | Description |
|---|---|---|---|
| C06 | Publication | ||
| PB01 | Publication | ||
| C10 | Entry into substantive examination | ||
| SE01 | Entry into force of request for substantive examination | ||
| GR01 | Patent grant | ||
| GR01 | Patent grant | ||
| CF01 | Termination of patent right due to non-payment of annual fee |
Granted publication date: 20180504 Termination date: 20180911 |
|
| CF01 | Termination of patent right due to non-payment of annual fee |