CN105472034B - The load-balancing method of mobility is reduced in a kind of multiserver DVE systems - Google Patents

The load-balancing method of mobility is reduced in a kind of multiserver DVE systems Download PDF

Info

Publication number
CN105472034B
CN105472034B CN201610000833.7A CN201610000833A CN105472034B CN 105472034 B CN105472034 B CN 105472034B CN 201610000833 A CN201610000833 A CN 201610000833A CN 105472034 B CN105472034 B CN 105472034B
Authority
CN
China
Prior art keywords
load
block
server
region
migration
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
Application number
CN201610000833.7A
Other languages
Chinese (zh)
Other versions
CN105472034A (en
Inventor
张伟
周恩强
董勇
陈志广
张钊宁
游鹏飞
李紫阳
Current Assignee (The listed assignees may be inaccurate. Google has not performed a legal analysis and makes no representation or warranty as to the accuracy of the list.)
National University of Defense Technology
Original Assignee
National University of Defense Technology
Priority date (The priority date 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 date listed.)
Filing date
Publication date
Application filed by National University of Defense Technology filed Critical National University of Defense Technology
Priority to CN201610000833.7A priority Critical patent/CN105472034B/en
Publication of CN105472034A publication Critical patent/CN105472034A/en
Application granted granted Critical
Publication of CN105472034B publication Critical patent/CN105472034B/en
Expired - Fee Related legal-status Critical Current
Anticipated expiration legal-status Critical

Links

Classifications

    • HELECTRICITY
    • H04ELECTRIC COMMUNICATION TECHNIQUE
    • H04LTRANSMISSION OF DIGITAL INFORMATION, e.g. TELEGRAPHIC COMMUNICATION
    • H04L67/00Network arrangements or protocols for supporting network services or applications
    • H04L67/01Protocols
    • H04L67/10Protocols in which an application is distributed across nodes in the network
    • H04L67/1001Protocols in which an application is distributed across nodes in the network for accessing one among a plurality of replicated servers

Landscapes

  • Engineering & Computer Science (AREA)
  • Computer Networks & Wireless Communication (AREA)
  • Signal Processing (AREA)
  • Information Transfer Between Computers (AREA)
  • Computer And Data Communications (AREA)

Abstract

The invention discloses the load-balancing method for reducing mobility in a kind of multiserver DVE systems, step includes:The two-dimensional space of virtual environment is divided into multiple regions with same size according to number of servers in advance, each region is made of the block of the square with same size;Obtained region will be divided and be individually assigned to different servers;During multiserver DVE system operations, each server checks own load, if load exceeds preset load threshold, then the alternative migration block of institute's distribution region is ranked up according to the migration block selection principle of curstomer-oriented end mobility, and is sent to the server for safeguarding adjacent area according to ranking results selection block.The present invention is fast with load balancing process speed, effect is good, can reduce client mobility, the operation expense for reducing system, the advantage that user's actual experience is good, applicability is good.

Description

The load-balancing method of mobility is reduced in a kind of multiserver DVE systems
Technical field
The present invention relates to the load-balancing techniques in MultiServer Structure DVE systems, and in particular to a kind of multiserver DVE The load-balancing method of mobility is reduced in system.
Background technology
Distributed virtual environment (Distributed Virtual Environment, abbreviation DVE) system is a kind of Real-time network interactive environment based on virtual reality technology and network technology, it is by being distributed between the calculate node of various regions Being distributed in more calculate nodes for task is completed in communication, the locally or globally data information of shared virtual environment, collaboration, gives user A kind of whole, the true, Virtual Space that can immerse is provided.Distributed virtual environment technology is adapted to due to its favorable expandability The features such as property is wide, has obtained extensively in Military Simulation, network amusement, long-distance education, engineering design and e-commerce etc. General application.In recent years, along with the drastically expansion of Distributed Virtual Environment System application scale, the real-time of user's interaction is wanted Ask continuous improvement, the scalability problem getting worse of system traditional to load energy based on the DVE system architectures of single server Power, response time etc. are difficult to meet the functional requirement of system.Therefore, the multiserver based on distributed server architecture DVE systems have been increasingly becoming the research hotspot of academia and industrial quarters.
As shown in Figure 1, in existing certain multiserver DVE systems, entire virtual environment system by high speed by being interconnected The multiple servers that network is connected with each other(Server#1~Server#4)It is common to safeguard.When system scale increases, Ke Yitong It crosses and increases number of servers to improve the integrity service ability of system, therefore its autgmentability will be substantially better than single server architecture DVE systems.However, relative to single server architecture, MultiServer Structure DVE systems will provide reliable and stable service, need Solve series of technical.Wherein, the problem of load balancing between each server is to influence system availability and user The key problem of experience, due to the mobility of each client controlled entity in virtual environment system, each server is actually responsible for Client terminal quantity can constantly change, therefore its loading condition also can constantly change.When the load of certain server When more than its load capacity upper limit, the response demand for the client being responsible for will be unable to be met, and exist to influence user Interactive experience in system.Therefore, how the load balance between the runtime effectively realizes multiserver, just become influence One key element of multiserver virtual environment system availability.
Current existing load-balancing method is primary concern is that the efficiency of balancing procedure and the validity of equilibrium result, i.e., How in a relatively short period of time, the load balance between different server is realized.However in fact, load balancing process can be to void The space in near-ring border divides and brings variation, to be had an impact to the migration situation of the clients entities in virtual environment.If Virtual Space is divided into multiple scrappy fritters, then clients entities just have prodigious probability from a server migration to Another server, and the on the one hand transition process of client can bring prodigious overhead, on the other hand also can to The actual experience at family affects.Therefore, how to reduce the mobility of client as far as possible during load balancing, just at In order to which DVE systems need the critical issue solved under MultiServer Structure.
Invention content
The technical problem to be solved in the present invention:For the above problem of the prior art, a kind of load balancing process speed is provided Degree is fast, effect is good, can reduce that client mobility, the operation expense for reducing system, user's actual experience is good, applicability is good The load-balancing method of mobility is reduced in multiserver DVE systems.
In order to solve the above-mentioned technical problem, the technical solution adopted by the present invention is:
The load-balancing method of mobility is reduced in a kind of multiserver DVE systems, step includes:
1)Before multiserver DVE system operations, the two-dimensional space of virtual environment is divided according to number of servers in advance For multiple regions with same size, each region is made of the block of the square with same size, the pros The block of shape carries out the least unit of two-dimensional space migration between different server during load balancing;Division is obtained Region is individually assigned to different servers, and each server is responsible for safeguarding that institute's distribution region and the client run thereon are real Body;
2)During multiserver DVE system operations, each server periodically checks the loading condition of itself, if Load exceeds preset load threshold, then according to the migration block selection principle of curstomer-oriented end mobility to institute's distribution region Alternative migration block is ranked up, and the server for safeguarding adjacent area is sent to according to ranking results selection block.
Preferably, the step 2)Detailed step include:
2.1)During multiserver DVE system operations, each current server periodically checks the load feelings of oneself Condition, judges whether the load of itself exceeds preset load threshold, is jumped if the load of itself is beyond preset load threshold Turn to execute step 2.2), otherwise redirected in next cycle and execute step 2.1);
2.2)Current server to safeguarded in the two-dimensional space of virtual environment adjacent area server send load hair Request is sent, load is received and sends the server asked and given according to itself current loading condition and feed back, if being currently in low-load Situation, which is then fed back, to be received, and otherwise feedback can not receive;
2.3)Current server receives the feedback from other servers, can not be received currently if other servers are fed back The load of server, then the load balancing process suspension in this period, redirects in next period and executes step 2.1), otherwise current to take Business device, which selects alternatively to migrate block from the region of self maintained, is ready for load migration;
2.4)Current server according to the migration block selection principle of curstomer-oriented end mobility to alternatively migrate block into Row sequence;
2.5)Current server selects block to be sent to feedback according to ranking results can receive the server implementation of load Migration;
2.6)Current server judges whether the loading condition of itself still exceeds preset load threshold, if still super Go out preset load threshold, then redirects and execute step 2.4);Otherwise transition process terminates, and execution step is redirected in next cycle 2.1).
Preferably, the step 2.4)Detailed step include:
2.4.1)It is ranked up by the sequence of boundary line adjusted value from low to high to alternatively migrating block, the boundary line tune Whole value subtracts by an alternative transition process of migration block region boundary line quantity of being moved out and moves into region boundary line quantity and obtain The result arrived;
2.4.2)It sorts from high in the end according to rectangular adjusted value for the identical alternative migration block of boundary line adjusted value, institute It states rectangular adjusted value and is moved out by an alternative transition process of migration block and the rectangular index in region divided by move into the rectangular finger in region Number obtain as a result, the rectangular index be the alternative migration block distance areas central point in horizontal and vertical directions The maximum value of distance divided by the evolution value of region area, when the evolution value of region area is that the region is fitted to virtual square The length of side.
The load-balancing method tool that mobility is reduced in multiserver DVE systems of the present invention has the advantage that:Face of the present invention The critical issue problem of load balancing that system availability and user's actual experience are influenced into MultiServer Structure DVE systems, While keeping the advantage that other load-balancing methods are efficient, equilibrium result is good, equilibrium result is further contemplated to virtual The influence that space divides optimizes load balancing process by introducing the decision principles such as boundary line rule and rectangular rule, Allow the result of load balancing to reduce the probability that client migrates between the multiple servers of virtual environment system, reduces The operation expense of system, improves the actual experience of user, to improve the availability of system on the whole.In addition, this method Expense is small in the process of running, and is that operation is distributed on multiple servers, "bottleneck" problem is not present, for having on a large scale Have more preferable compared with the Distributed Virtual Environment System applicability of multiserver, has that load balancing process speed is fast, effect is good, it can Reduce client mobility, the operation expense for reducing system, the advantage that user's actual experience is good, applicability is good.
Description of the drawings
Fig. 1 is the structural schematic diagram of existing multiserver DVE systems.
Fig. 2 is the basic procedure schematic diagram of present invention method.
Fig. 3 is step 2 of the embodiment of the present invention)Basic procedure schematic diagram.
Specific implementation mode
As shown in Fig. 2, the step of reducing the load-balancing method of mobility in the present embodiment multiserver DVE systems, wraps It includes:
1)Before multiserver DVE system operations, the two-dimensional space of virtual environment is divided according to number of servers in advance For multiple regions with same size, each region is made of the block of the square with same size, square Block carries out the least unit of two-dimensional space migration between different server during load balancing;The region that division is obtained It is individually assigned to different servers, each server is responsible for the clients entities safeguarded institute's distribution region and run thereon;
2)During multiserver DVE system operations, each server periodically checks the loading condition of itself, if Load exceeds preset load threshold, then according to the migration block selection principle of curstomer-oriented end mobility to institute's distribution region Alternative migration block is ranked up, and the server for safeguarding adjacent area is sent to according to ranking results selection block, to avoid The generation of overload situations.
As shown in figure 3, step 2)Detailed step include:
2.1)During multiserver DVE system operations, each current server periodically checks the load feelings of oneself Condition, judges whether the load of itself exceeds preset load threshold, is jumped if the load of itself is beyond preset load threshold Turn to execute step 2.2), otherwise redirected in next cycle and execute step 2.1);
2.2)Current server to safeguarded in the two-dimensional space of virtual environment adjacent area server send load hair Request is sent, load is received and sends the server asked and given according to itself current loading condition and feed back, if being currently in low-load Situation, which is then fed back, to be received, and otherwise feedback can not receive;
2.3)Current server receives the feedback from other servers, can not be received currently if other servers are fed back The load of server, then the load balancing process suspension in this period, redirects in next period and executes step 2.1), otherwise current to take Business device, which selects alternatively to migrate block from the region of self maintained, is ready for load migration;
2.4)Current server according to the migration block selection principle of curstomer-oriented end mobility to alternatively migrate block into Row sequence;
2.5)Current server selects block to be sent to feedback according to ranking results can receive the server implementation of load Migration;
2.6)Current server judges whether the loading condition of itself still exceeds preset load threshold, if still super Go out preset load threshold, then redirects and execute step 2.4);Otherwise transition process terminates, and execution step is redirected in next cycle 2.1).
In the present embodiment, step 2.4)Detailed step include:
2.4.1)It is ranked up by the sequence of boundary line adjusted value from low to high to alternatively migrating block, boundary line adjusted value It is subtracted for the moved out region boundary line quantity of alternative migration transition process of block and moves into what region boundary line quantity obtained As a result;
2.4.2)It sorts from high in the end according to rectangular adjusted value for the identical alternative migration block of boundary line adjusted value, side Shape adjusted value is moved out by an alternative transition process of migration block and the rectangular index in region divided by moves into the rectangular index in region and obtain It is arriving as a result, rectangular index be this it is alternative migration block distance areas central point horizontal and vertical directions distance most Big value divided by the evolution value of region area, the evolution value of region area are the length of side when region is fitted to virtual square.
For the clients entities run in virtual environment, if it has passed through the area that a server is safeguarded The cross-server migration of client will occur for the boundary in domain, i.e., the client is converted to from the maintenance area of a server The maintenance area of another server.This process will produce larger overhead, and influence the user experience of user, because We need to reduce the probability of this transition process as far as possible for this.In fact, the probability and multiple servers of client migration Boundary length between maintenance area is closely bound up, and the boundary length between region is longer, and the probability of client migration is bigger. Therefore, it to reduce the probability of client migration, needs to reduce the boundary length between multiple regions as far as possible.For this purpose, this hair Bright embodiment designs a kind of migration block selection principle based on client mobility, and it includes following two specific rules:
I)Boundary line rule, referring to step 2.4.1), in particular to alternatively migrating block by boundary line adjusted value by low It is ranked up to high sequence.For a specific migration block, its transition process can only influence server of moving out And the border condition of server is moved into, the border condition without influencing other servers.Specifically, in this time transition process In, migration block and move into server institute maintenance area boundary line will disappear, and with the friendship for server institute maintenance area of moving out Boundary line will become new boundary.Therefore this time transition process, caused boundary line length varying value are exactly the block and move The quantity for going out the boundary line of server subtracts and moves into the quantity of the boundary line of server, and we term it the boundaries of this block Line adjusted value, i.e. boundary line adjusted value=region boundary line quantity of moving out-move into region boundary line quantity.Boundary line adjusted value is When positive value, illustrate that this time migration increases the entire length of boundary line, and when being negative value, illustrate this time to migrate and reduce side The entire length in boundary line.For example, if only 1 side connects between some migration block and server maintenance area of moving out, and There are 3 sides to connect with moving between server maintenance area, then its boundary line adjusted value is -2, illustrates that this time migration will Reduce whole boundary 2 units of line length.As it can be seen that the lower boundary line adjusted value the better, therefore we are to all potential migrations It when block sorts, needs to be ranked up from low to high according to boundary line adjusted value, migrates the low area of boundary line adjusted value as far as possible Block.
II)Rectangular rule, referring to step 2.4.2), in particular to for the identical alternative migration block of boundary line adjusted value It sorts from high in the end according to rectangular adjusted value.The basic ideas of rectangular rule are in transition process, and will move out service as far as possible The region that device is safeguarded and the region for moving into server maintenance are fitted to square.This is because coming for all sliceable figures It says, square perimeter area ratio is 4/ length of side, has lower perimeter area ratio compared with other most of shapes, i.e., same The maintenance area of size possesses shorter boundary line.For example, for a square area, if to be divided into four parts, It is divided into twice that total boundary line that the division methods of 4 small square obtain is the length of side, and if be divided into from top to bottom 4 rectangles then total boundary length just becomes for three times of the length of side, here it is due to square rectangle more unequal than length and width With lower perimeter area ratio.In fact, for square, round and regular hexagon has smaller perimeter area Than, but circle is not sliceable figure, the boundary line of each region can not share, and regular hexagon is small for base unit It is difficult to completely splice for square, is more suitable for the Virtual Space using regular hexagon as base unit.For this purpose, the present invention retouches It states with the rectangular the rule of fitting for target, concrete methods of realizing is equally the rectangular adjusted value for calculating the block, and computational methods are It calculates separately some migration block and server maintenance area and moves into the rectangular index of server maintenance area moving out, the two Quotient is square adjusted value, i.e., rectangular adjusted value, and=the rectangular index in region of moving out/moves into the rectangular index in region.The meter of rectangular index Calculation method is maximum value divided by region area of the block distance areas central point in the distance of horizontal and vertical directions(I.e. Total number of blocks)Evolution value, the evolution value of wherein region area represents the side when region is fitted to virtual square It is long, therefore rectangular index represents the block and is fitted the influence degree being square to region in the zone, it is bigger to illustrate to influence more Greatly.Therefore rectangular adjusted value is bigger, and representing the block influences the side of moving out bigger, and smaller on the side's of moving into influence, therefore we are more Tend to migrate this kind of block.
The above is only a preferred embodiment of the present invention, protection scope of the present invention is not limited merely to above-mentioned implementation Example, all technical solutions belonged under thinking of the present invention all belong to the scope of protection of the present invention.It should be pointed out that for the art Those of ordinary skill for, several improvements and modifications without departing from the principles of the present invention, these improvements and modifications It should be regarded as protection scope of the present invention.

Claims (1)

1. reducing the load-balancing method of mobility in a kind of multiserver DVE systems, it is characterised in that step includes:
1)Before multiserver DVE system operations, the two-dimensional space of virtual environment is divided into according to number of servers in advance more A region with same size, the multiserver DVE systems refer to multi-server Distributed virtual environment system, Mei Gequ Domain is all made of the block of the square with same size, and the block of the square is different services during load balancing The least unit of two-dimensional space migration is carried out between device;Obtained region will be divided and be individually assigned to different servers, each Server is responsible for the clients entities safeguarded institute's distribution region and run thereon;
2)During multiserver DVE system operations, each server periodically checks the loading condition of itself, if load Beyond preset load threshold, then according to the migration block selection principle of curstomer-oriented end mobility to the alternative of institute's distribution region Migration block is ranked up, and the server for safeguarding adjacent area is sent to according to ranking results selection block;
The step 2)Detailed step include:
2.1)During multiserver DVE system operations, each server periodically checks the loading condition of oneself, judges certainly Whether the load of body exceeds preset load threshold, and execution step is redirected if the load of itself is beyond preset load threshold 2.2), otherwise redirected in next cycle and execute step 2.1);
2.2)Current server to safeguarded in the two-dimensional space of virtual environment adjacent area server send load send ask It asks, receives load and send the server asked and given according to itself current loading condition and feed back, if being currently in low-load situation Then feedback can receive, and otherwise feedback can not receive;
2.3)Current server receives the feedback from other servers, if other servers, which are fed back, can not receive current service The load of device, then the load balancing process suspension in this period, redirects in next period and executes step 2.1), otherwise current server Alternative migration block is selected to be ready for load migration from the region of self maintained;
2.4)Current server is arranged according to the migration block selection principle of curstomer-oriented end mobility alternatively migrating block Sequence;
2.5)Current server selects block to be sent to feedback according to ranking results can receive the server implementation migration of load;
2.6)Current server judges whether the loading condition of itself still exceeds preset load threshold, if still exceeded pre- If load threshold, then redirect execute step 2.4);Otherwise transition process terminates, and execution step is redirected in next cycle 2.1);
The step 2.4)Detailed step include:
2.4.1)It is ranked up by the sequence of boundary line adjusted value from low to high to alternatively migrating block, the boundary line adjusted value It is subtracted for the moved out region boundary line quantity of alternative migration transition process of block and moves into what region boundary line quantity obtained As a result;
2.4.2)It sorts from high in the end according to rectangular adjusted value for the identical alternative migration block of boundary line adjusted value, the side Shape adjusted value is moved out by an alternative transition process of migration block and the rectangular index in region divided by moves into the rectangular index in region and obtain It is arriving as a result, the rectangular index be this it is alternative migration block distance areas central point horizontal and vertical directions distance Maximum value divided by region area evolution value, the evolution value of region area is the side when region is fitted to virtual square It is long.
CN201610000833.7A 2016-01-04 2016-01-04 The load-balancing method of mobility is reduced in a kind of multiserver DVE systems Expired - Fee Related CN105472034B (en)

Priority Applications (1)

Application Number Priority Date Filing Date Title
CN201610000833.7A CN105472034B (en) 2016-01-04 2016-01-04 The load-balancing method of mobility is reduced in a kind of multiserver DVE systems

Applications Claiming Priority (1)

Application Number Priority Date Filing Date Title
CN201610000833.7A CN105472034B (en) 2016-01-04 2016-01-04 The load-balancing method of mobility is reduced in a kind of multiserver DVE systems

Publications (2)

Publication Number Publication Date
CN105472034A CN105472034A (en) 2016-04-06
CN105472034B true CN105472034B (en) 2018-09-28

Family

ID=55609296

Family Applications (1)

Application Number Title Priority Date Filing Date
CN201610000833.7A Expired - Fee Related CN105472034B (en) 2016-01-04 2016-01-04 The load-balancing method of mobility is reduced in a kind of multiserver DVE systems

Country Status (1)

Country Link
CN (1) CN105472034B (en)

Families Citing this family (1)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
CN112882714B (en) * 2021-02-04 2024-01-30 中国工商银行股份有限公司 SonarQube server cluster capacity expansion method and device

Citations (3)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
CN101005514A (en) * 2006-12-27 2007-07-25 北京航空航天大学 Multiple server organizing method for network game
CN101009648A (en) * 2006-12-27 2007-08-01 北京航空航天大学 Multi-server hierarchical alterative load balance method
CN101355518A (en) * 2008-09-08 2009-01-28 北京航空航天大学 Method for regulating multi-server dynamic area based on path combination

Patent Citations (3)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
CN101005514A (en) * 2006-12-27 2007-07-25 北京航空航天大学 Multiple server organizing method for network game
CN101009648A (en) * 2006-12-27 2007-08-01 北京航空航天大学 Multi-server hierarchical alterative load balance method
CN101355518A (en) * 2008-09-08 2009-01-28 北京航空航天大学 Method for regulating multi-server dynamic area based on path combination

Non-Patent Citations (3)

* Cited by examiner, † Cited by third party
Title
A load balancing algorithm for a distributed multimedia game server architecture;Dugki Min等;《IEEE》;20020806;全文 *
MSDVE:一个支持动态区域调整的分布式虚拟现实系统;许林等;《2008系统仿真技术及其应用学术会议论文集》;20080731;全文 *
一种基于生成树的多服务器动态区域调整算法;许林等;《系统仿真学报》;20090131;全文 *

Also Published As

Publication number Publication date
CN105472034A (en) 2016-04-06

Similar Documents

Publication Publication Date Title
Wang et al. Energy-aware dynamic virtual machine consolidation for cloud datacenters
CN104850450B (en) A kind of load-balancing method and system towards mixed cloud application
CN102567080B (en) Virtual machine position selection system facing load balance in cloud computation environment
Beniwal et al. A comparative study of static and dynamic load balancing algorithms
Sreenivas et al. Load balancing techniques: Major challenge in Cloud Computing-a systematic review
CN103441918A (en) Self-organizing cluster server system and self-organizing method thereof
Kapgate Efficient service broker algorithm for data center selection in cloud computing
Srivastava et al. CGP: Cluster-based gossip protocol for dynamic resource environment in cloud
Goyal A study of load balancing in cloud computing using soft computing techniques
CN103918239A (en) Load balancing method, device, system and computer readable medium
Li et al. Deployment of edge servers in 5G cellular networks
CN113645146A (en) New stream density-based load balancing method and system for software defined network controller
Chang et al. Adaptive replication for mobile edge computing
Khodar et al. New scheduling approach for virtual machine resources in cloud computing based on genetic algorithm
Duolikun et al. A monotonically increasing (MI) algorithm to estimate energy consumption and execution time of processes on a server
CN105472034B (en) The load-balancing method of mobility is reduced in a kind of multiserver DVE systems
Shabeera et al. Curtailing job completion time in MapReduce clouds through improved Virtual Machine allocation
Lin et al. A workload-driven approach to dynamic data balancing in MongoDB
Xue et al. Communication-aware virtual machine migration in cloud data centres
Swarnakar et al. A novel improved hybrid model for load balancing in cloud environment
Goyal et al. Adaptive and dynamic load balancing methodologies for distributed environment: a review
Khanli et al. A new hybrid load balancing algorithm in grid computing systems
Marinescu et al. Distributed hierarchical control versus an economic model for cloud resource management
Jayasinghe et al. Task assignment in multiple server farms using preemptive migration and flow control
Bragard et al. Global dynamic load-balancing for decentralised distributed simulation

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
CF01 Termination of patent right due to non-payment of annual fee

Granted publication date: 20180928

Termination date: 20220104