CN102195890A - Internet application dispatching method based on cloud computing - Google Patents
Internet application dispatching method based on cloud computing Download PDFInfo
- Publication number
- CN102195890A CN102195890A CN2011101497075A CN201110149707A CN102195890A CN 102195890 A CN102195890 A CN 102195890A CN 2011101497075 A CN2011101497075 A CN 2011101497075A CN 201110149707 A CN201110149707 A CN 201110149707A CN 102195890 A CN102195890 A CN 102195890A
- Authority
- CN
- China
- Prior art keywords
- application
- load
- server
- new
- article
- 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.)
- Granted
Links
Images
Abstract
The invention provides an internet application dispatching method based on cloud computing, and the method comprises the following steps: 1) a dispatcher which is installed at the front end of application servers monitors the configuration information of the application servers, the application requirements on each server, the user request number of the previous moment and the current moment and the operation information of all examples in each application server; 2) when the change of the application is monitored, the server load with the changed application is adjusted by utilizing the load descending of the application, an application logging-out system, an application adding system and the load ascending of the application through a packing algorithm, and a transponder modifies the load distribution among the examples of each application, thereby reducing started new application examples on a new server; and 3) the dispatcher outputs the application example which needs to be closed, the application example which needs to be started again and the server on which the application examples are started. In the method provided by the invention, the service number of resources is dynamically adjusted for users according to the dynamic requirements of the user application; the new application examples are avoided from being started to the greatest extent; and the cost is low.
Description
Technical field
The invention belongs to the Computer Applied Technology field, relate to a kind of cloud computing resource regulating method, especially a kind of internet, applications dispatching method based on cloud computing.
Background technology
The cloud computing pattern of rising in recent years has become the hot issue of industrial quarters and academia's concern.Traditional computation schema is server computational resource, the storage resources that each user safeguards oneself.And cloud computing is a brand-new application model, and cloud computing manufacturer provides online service to the client, makes the computational resource centralized management, and the user obtains the cloud computing resource as required.Cloud computing service quality is current important one of study a question.Under the pattern of cloud computing, how cloud computing manufacturer becomes a problem that presses for solution for the user provides a stable service environment.
Cloud computing has further reduced the expense that maintenance server brought to the computational resource centralized management.Investigation shows, the no longer own service data of more and more enterprises center, and the APD of oneself is deployed on the cloud computing platform is by the cloud computing provider resource of distributing according to need.Computer cluster uses for reasonable resources and effective scheduling has all proposed very high requirement, because the types of applications program has a great difference to the demand of computational resource, even same class is applied in the resource requirement of different time and also can changes a lot, this just requires all kinds of resources such as a tactful fast and efficiently next CPU to the calculating cluster, internal memory, network to carry out whole optimization.But actually, the software and hardware resources between each node is often variant, comprises type, version, patch, function library, applied environment of operating system etc.This makes the application program of operation on it be difficult in migration freely between each node.Virtual technology provides on physical machine abstract.It has masked the difference of bottom software and hardware effectively and provides the running environment of a unanimity, i.e. " virtual platform " for the application program on it.
A flexibility that outstanding advantage is a resource of cloud computing service: the user can adjust the resource that oneself uses according to demand at any time, carries out purchasing and safeguarding of resource and need not drop into a large amount of capitals as in the past.The for example EC2 of Amazon service allows the client to determine to rent the number of virtual machine according to demand.Yet in these services, concrete number of resources but still need be determined by the user voluntarily.If the cloud service provider can dynamically adjust the usage quantity of resource according to the dynamic need that the user uses for the user, very big benefit is all arranged to serving provider and user.
Summary of the invention
Deficiency in view of prior art exists the invention provides a kind of cloud computing environment internet, applications dispatching method, can dynamically adjust the operation of application example on which server according to the load variations that the user uses.
To achieve these goals, the technical solution used in the present invention is summarized as follows:
A kind of internet, applications dispatching method based on cloud computing comprises step:
1) be installed on the configuration information of the scheduler monitoring application server of application server front end, demands of applications on each server, and last one constantly and the user of current time ask number; Each uses the operation information of all examples;
2) change when monitoring to use, the load that comprises application descends, uses the load that logs off, uses adding system, application and rises, by bin packing algorithm the server load that application changes is adjusted, and change load Distribution between each example of each application by transponder, reduce on new server and open new application example;
3) scheduler exports the application example that need cut out, the application example that needs new startup, and starts on which server.
Described step 1) application server is the isomorphism application server.
If described step 1) application server is an isomery application server then according to the composition isomorphism application server of isomorphic relations with identical configuration.
Described step 2) bin packing algorithm is:
Server is considered as chest, and the capacity of cpu resource is considered as size of a case;
Application is considered as class article, and the article of different application correspondence belong to different classifications;
The number of the corresponding article of request sum of each application;
The classification restriction c of a chest is taken as physical machine memory source sum divided by the maximum memory resource requirement of using;
The classification of article arbitrarily is divided into several set, and the classification number that guarantees to contain except set at the most is less than the c, other set all comprises a lucky c classification, handles each independent set respectively.
Described step 2) when monitoring application load the variation that descends takes place, collect the tabulation that is applied to of load decline, judge whether and under the situation that does not start new application example, to adjust load, if can not start new application example, directly adjustment respectively is applied in the load allocating on the server, if must start new application example, from former load less than server on one of random choose use the load that the load of moving a unit is filled up to have descended.
Described step 2) when monitoring application load the variation of rising takes place, collect the tabulation that is applied to of load rising, judge whether and under the situation that does not start new application example, to adjust load, if can adjust the load allocating that respectively is applied on the server, if cannot, with load allocating to former load less than server on.
Described step 2) classification of using arbitrarily is divided into several set, and the applicating category number that guarantees to contain except set at the most is less than the c, other set all comprises a lucky c applicating category, when the new application that monitors adding, judge whether the new application that adds is enough to fill up these set, be enough to fill up, use greedy method to form new set the new application of remainder, and satisfy applicating category number that maximum set contain character less than c.
Compared with prior art, the technique effect that has of method of the present invention has:
The dynamic need of using according to the user is dynamically adjusted the usage quantity of resource for the user, avoid starting new application example as far as possible, and cost is little, can keep higher request to satisfy rate when high capacity, calculates the saving energy and can carry out green when low load.
Description of drawings
Fig. 1 is a load ascent algorithm flow chart of the present invention;
Fig. 2 is a load descent algorithm flow chart of the present invention.
Fig. 3 handles the flow chart that load is risen for the inventive method;
Fig. 4 handles the flow chart that load descends for the inventive method;
Fig. 5 handles the new flow chart of using that adds for the inventive method;
Fig. 6 is the flow chart of the inventive method;
Fig. 7 a is the relation between approximation ratio R and the c variable;
Fig. 7 b is the relation between approximation ratio R and the v variable.
Embodiment
Below in conjunction with the drawings and specific embodiments method of the present invention is done detailed description.
1) scheduling problem that server under the cloud environment is used is carried out modeling
Tradition bin packing (bin-packing) is a np hard problem that obtains broad research, its require with a series of sizes (0,1] pack into the size of minimum number of article in the scope is in 1 the chest.This problem has various modification, as online bin packing, requires to carry out minimum chest number vanning under the situation of not knowing article sequence on the horizon; The bin packing of semi-on-line (semi-online) on online basis, when allowing each new article to arrive, is adjusted the position of original article; The multidimensional bin packing, article and size of a case expand to the vector of a multidimensional; Allow bin packing that article leave or the like in addition midway.
Another mutation of bin packing is Class Constrained Bin Packing (CCBP), in this problem, chest has fixing big or small v, the article size all is a unit 1, each article has and belongs to a class, and a chest can only be put into the article of c kind class at the most, and the target of problem is that a series for articles is put into the minimized number chest.The internet, applications scheduling problem that this problem and the inventive method are considered is very close, is isomorphisms if the inventive method is supposed all servers, and only considers the internal memory used and the load of CPU two dimensions, can with its abstract be the mutation of a CCBP problem.Concrete for the present invention modeling process is:
Server is abstracted into chest, and the capacity of cpu resource is a size of a case;
Application is considered as class article, and the article of different application correspondence belong to different classifications;
The request sum of each application is regarded the number that this uses corresponding article as after being scaled the cpu resource demand;
The classification restriction c of a chest is taken as physical machine memory source sum divided by the maximum memory resource requirement of using.
Like this, when each scheduler was called, the increase of application request number or minimizing can be regarded the arrival of corresponding such article as and leave away.So,, just can apply in the scheduler, and satisfy the demand of maximization application request service rate and green calculating simultaneously as long as one of design can be handled the algorithm of the newly-increased and CCBP that leaves away of article.
The present invention arbitrarily is divided into several set with the classification of article, and the classification number that guarantees to contain except set at the most is less than the c, and other set all comprises a lucky c classification.In the article vanning process afterwards, all consider the set that each is independent respectively, promptly be independent of each other between each set.Because the classification number is no more than c in the single set, so can case: open the another one chest again after a chest filled with a simple greedy algorithm.
What algorithm mainly will be considered is the situation that newly-increased article or article are left away.The problem of a key is: put into article and correspond to actual environment in different chests, two kinds of situations are arranged: the one, in a new server, open a new application example; The 2nd, if moved the example of this application in this server, as long as that simply heighten the proportioning of load in the transponder.In the both of these case, the former cost is very big, and the latter is then very little by contrast, so the inventive method will be managed in the process of each vanning, reduces the appearance of first kind of situation by the method for adjusting load allocating as far as possible.
As shown in Figure 1, be the schematic diagram of algorithm process process when certain application load rises.The internet system of cloud environment has three servers, and three application (corresponding three kinds of different classes of article) when the application load of classification c1 representative rises, can abstractly require to case for the article that a new c1 classification is arranged arrive.If use simple greedy algorithm, these article will be placed to last not to be had in the full chest.Corresponding to actual conditions like this is exactly the example that does not have to start on the full server this application at this, and cost is bigger.But examine, if adjust article by the step shown in the figure: the article that move a c2 classification in the b2 set are in the unfilled set, the article that move a c3 classification in the b1 set are in the b2 set, at last newly arrived c1 classification article are placed in the b1 set, correspond in the actual conditions operation that does not just have new startup application example so, all are adjusted the load Distribution that all only needs by transponder is changed between each example of each application and get final product.
For the situation that load descends, processing method is similar (as Fig. 2).The application load of c1 representative has descended leaving of these article of classification corresponding.Has only a state in order to keep less than chest, mobile c3 classification article are in the b1 set in the b2 set, mobile c2 classification article are in the b2 set in the unfilled set, by adjustment shown in Figure 2, system reaches once more and has only a state less than chest, and is not activated new application example.
Algorithm Analysis
The good and bad degree of CCBP algorithm A can recently be weighed with approximate, and approximation ratio is defined as follows:
Then have:
Wherein, σ represents any one list entries, and A (σ) expression algorithm A imports required chest number hereto, and OPT (σ) represents that this imports minimum how many chests of needs, the i.e. optimal solution of problem.
The inventive method note m is the applicating category sum in the system, and c and v are foregoing kind restriction and chest size, l
1For applicating category is the total number of items of i, and order:
Then have:
Consider to get integral symbol when OPT (σ) can ignore when being tending towards infinity, the approximation ratio that can obtain this algorithm is:
So smaller approximation ratio corresponds in the actual conditions, and representative is when system load is low, and system can satisfy all demands with few server of trying one's best, thereby has saved the energy, reaches green computation purpose; And too high when system load, in the time of can't satisfying all demands, can maximize the rate that satisfies of request again as much as possible.
Except the situation that consideration application load above-mentioned changes, dispatching method of the present invention also will be considered withdrawing from of new adding of using and original application.As shown in Figure 3, the total flow process of the inventive method is as follows:
1) be installed on the configuration information of the scheduler monitoring application server of application server front end, demands of applications on each server, and last one constantly and the user of current time ask number; Each uses the operation information of all examples;
2) change when monitoring to use, the load that comprises application descends, uses the load that logs off, uses adding system, application and rises, change the load Distribution between each example of each application by transponder, reduce on new server and open new application example;
3) scheduler exports the application example that need cut out, the application example that needs new startup, and starts on which server.
Handle application load with the inventive method and rise to example, flow process begins as shown in Figure 3, is installed on the configuration information of the scheduler monitoring application server of application server front end, demands of applications on each server, and the user of a last moment and current time asks number; Each uses the operation information of all examples; The variation of rising takes place when monitoring application load, collect the tabulation that is applied to of load rising, by judging whether tabulation is that sky judges whether application load rises, if tabulation is for empty, take out the first application of tabulation, make that delta is this application load rising number,, change over to and judge whether tabulation is empty if delta 0 directly shifts out this application from tabulation; If delta is not 0, judging whether can be according to method adjustment shown in Figure 1, promptly under the situation that does not start new application example, adjust load, if can respectively be applied in the load allocating on the server by the strategy adjustment of Fig. 1, if cannot, from the load allocating of a newly-increased unit to former load less than server on (needing to start new example).
Handle application load with the inventive method and drop to example, flow process begins as shown in Figure 4, is installed on the configuration information of the scheduler monitoring application server of application server front end, demands of applications on each server, and the user of a last moment and current time asks number; Each uses the operation information of all examples; The variation that descends takes place when monitoring application load, collect the tabulation that is applied to of load decline, by judging whether tabulation is that sky judges whether application load descends, if tabulation is for empty, take out the first application of tabulation, make that delta is this application load decline number,, change over to and judge whether tabulation is empty if delta 0 directly shifts out this application from tabulation; If delta is not 0, judging whether can be according to method adjustment shown in Figure 2, promptly under the situation that does not start new application example, adjust load, if can respectively be applied in the load allocating on the server by the strategy adjustment of Fig. 2, if cannot, from former load less than server on one of random choose use the load of moving a unit fill up descended load.
Handle new adding with the inventive method and be applied as example, flow chart as shown in Figure 5; Beginning, the scheduler that is installed on the application server front end is monitored the configuration information of application server, demands of applications on each server, and the user of a last moment and current time asks number; Each uses the operation information of all examples; When having monitored initiate application, with previous step produce less than the set of c kind applicating category according to the descending of applicating category number, order according to set will newly be used the set that is assigned to correspondence with greedy method, judge new adding uses whether be enough to fill up these set, if newly use abundant, be enough to fill up these set, use greedy method to form new set the new application of remainder, if new the application is not enough to fill up these set, with this moment less than the set of c kind applicating category once more according to the descending of applicating category number, judge whether to satisfy maximum set less than character, if do not satisfy maximum set less than character, fill up the maximum set of application with using minimum set, thereby form full set or empty set.
Supposed that in the inventive method server is an isomorphism, this is very typical in bin packing.But in the large-scale data center, the configuration of server is diversified, only need putting together of identical configuration be got final product with method processing of the present invention according to isomorphic relations with classification server.
In addition, the place that can optimize in addition of the inventive method.In system applicating category and number of sets for a long time, each the set in load do not allow to ignore less than the vacant resource sum of server.So, when satisfying rate in the system and cannot reach 100%, can be in the last independence of considering to break between set of algorithm, temporarily with the load allocating of some application to the set that does not belong to it in, from but satisfy the rate rising, system resource utilization is more abundant.
Whole scheduler is a plug-in unit as software Usher, realize with Python, as the plug-in unit on the Virtual Machine Manager software Usher, down can be as long as the code of this scheduler is put into usher/plugins so that the cluster virtual machine of Usher management is subjected to the scheduling of scheduler distributes.Various parameters are arranged in the scheduler, use percentage etc. as decision-making time of algorithm, kind restriction c, the abstract capacity v of server, real resource when server is considered as full load, need be by attempting and adjusting according to some experiences.Relation between approximation ratio R and c, the several variablees of v, L can be referring to the reference of Fig. 7 a as the parameter c setting; The reference that Fig. 7 b is provided with as parameter v.R is more little for approximation ratio, and algorithm is accurate more.
Claims (7)
1. internet, applications dispatching method based on cloud computing comprises step:
1) be installed on the configuration information of the scheduler monitoring application server of application server front end, demands of applications on each server, and last one constantly and the user of current time ask number; Each uses the operation information of all examples;
2) change when monitoring to use, the load that comprises application descends, uses the load that logs off, uses adding system, application and rises, by bin packing algorithm the server load that application changes is adjusted, and change load Distribution between each example of each application by transponder, reduce on new server and open new application example;
3) scheduler exports the application example that need cut out, the application example that needs new startup, and starts on which server.
2. the method for claim 1 is characterized in that, described step 1) application server is the isomorphism application server.
3. the method for claim 1 is characterized in that, if described step 1) application server is an isomery application server then according to the composition isomorphism application server of isomorphic relations with identical configuration.
4. the method for claim 1 is characterized in that, described step 2) bin packing algorithm is:
Server is considered as chest, and the capacity of cpu resource is considered as size of a case;
Application is considered as class article, and the article of different application correspondence belong to different classifications;
The number of the corresponding article of request sum of each application;
The classification restriction c of a chest is taken as physical machine memory source sum divided by the maximum memory resource requirement of using;
The classification of article arbitrarily is divided into several set, and the classification number that guarantees to contain except set at the most is less than the c, other set all comprises a lucky c classification, handles each independent set respectively.
5. the method for claim 1, it is characterized in that, described step 2) when monitoring application load the variation that descends takes place, collect the tabulation that is applied to of load decline, judge whether and under the situation that does not start new application example, to adjust load, if can not start new application example, directly adjustment respectively is applied in the load allocating on the server, if must start new application example, from former load less than server on one of random choose use the load that the load of moving a unit is filled up to have descended.
6. the method for claim 1, it is characterized in that, described step 2) when monitoring application load the variation of rising takes place, collect the tabulation that is applied to of load rising, judge whether and under the situation that does not start new application example, to adjust load, if can adjust the load allocating that respectively is applied on the server, if cannot, with load allocating to former load less than server on.
7. the method for claim 1, it is characterized in that, described step 2) classification of using arbitrarily is divided into several set, and the applicating category number that guarantees to contain except set at the most is less than the c, other set all comprises a lucky c applicating category, when the new application that monitors adding, judge whether the new application that adds is enough to fill up these set, be enough to fill up, use greedy method to form new set the new application of remainder, and satisfy applicating category number that maximum set contain character less than c.
Priority Applications (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN201110149707.5A CN102195890B (en) | 2011-06-03 | 2011-06-03 | Internet application dispatching method based on cloud computing |
Applications Claiming Priority (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN201110149707.5A CN102195890B (en) | 2011-06-03 | 2011-06-03 | Internet application dispatching method based on cloud computing |
Publications (2)
| Publication Number | Publication Date |
|---|---|
| CN102195890A true CN102195890A (en) | 2011-09-21 |
| CN102195890B CN102195890B (en) | 2014-04-16 |
Family
ID=44603298
Family Applications (1)
| Application Number | Title | Priority Date | Filing Date |
|---|---|---|---|
| CN201110149707.5A Expired - Fee Related CN102195890B (en) | 2011-06-03 | 2011-06-03 | Internet application dispatching method based on cloud computing |
Country Status (1)
| Country | Link |
|---|---|
| CN (1) | CN102195890B (en) |
Cited By (8)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN102567080A (en) * | 2012-01-04 | 2012-07-11 | 北京航空航天大学 | Virtual machine position selection system facing load balance in cloud computation environment |
| CN103197952A (en) * | 2012-01-09 | 2013-07-10 | 华为技术有限公司 | Management system and method aiming at maintenance and deployment of application system based on cloud infrastructure |
| CN105052080A (en) * | 2013-02-18 | 2015-11-11 | 泰科来股份有限公司 | Methods, systems, and computer readable media for providing a thinking diameter network architecture |
| CN105052074A (en) * | 2013-02-18 | 2015-11-11 | 泰科来股份有限公司 | Methods, systems, and computer readable media for providing a virtualized diameter network architecture and for routing traffic to dynamically instantiated diameter resource instances |
| CN106888237A (en) * | 2015-12-15 | 2017-06-23 | 中国移动通信集团公司 | A kind of data dispatching method and system |
| CN110855646A (en) * | 2019-10-31 | 2020-02-28 | 苏州经贸职业技术学院 | Approach loading system for network computer and application method |
| CN112956172A (en) * | 2018-10-31 | 2021-06-11 | 思科技术公司 | Transaction-based event tracking mechanism |
| US11388082B2 (en) | 2013-11-27 | 2022-07-12 | Oracle International Corporation | Methods, systems, and computer readable media for diameter routing using software defined network (SDN) functionality |
Citations (2)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN101719081A (en) * | 2009-12-01 | 2010-06-02 | 北京大学 | Method for scheduling virtual machines |
| CN101938416A (en) * | 2010-09-01 | 2011-01-05 | 华南理工大学 | Cloud computing resource scheduling method based on dynamic reconfiguration virtual resources |
-
2011
- 2011-06-03 CN CN201110149707.5A patent/CN102195890B/en not_active Expired - Fee Related
Patent Citations (2)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN101719081A (en) * | 2009-12-01 | 2010-06-02 | 北京大学 | Method for scheduling virtual machines |
| CN101938416A (en) * | 2010-09-01 | 2011-01-05 | 华南理工大学 | Cloud computing resource scheduling method based on dynamic reconfiguration virtual resources |
Cited By (16)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN102567080B (en) * | 2012-01-04 | 2015-03-04 | 北京航空航天大学 | Virtual machine position selection system facing load balance in cloud computation environment |
| CN102567080A (en) * | 2012-01-04 | 2012-07-11 | 北京航空航天大学 | Virtual machine position selection system facing load balance in cloud computation environment |
| US9519469B2 (en) | 2012-01-09 | 2016-12-13 | Huawei Technologies Co., Ltd. | Cloud infrastructure-based management system and method for maintenance and deployment of application system |
| CN103197952A (en) * | 2012-01-09 | 2013-07-10 | 华为技术有限公司 | Management system and method aiming at maintenance and deployment of application system based on cloud infrastructure |
| WO2013104217A1 (en) * | 2012-01-09 | 2013-07-18 | 华为技术有限公司 | Cloud infrastructure based management system and method for performing maintenance and deployment for application system |
| CN105052074B (en) * | 2013-02-18 | 2019-03-19 | 泰科来股份有限公司 | For providing method, system and the computer-readable medium of virtualization diameter network framework and the diameter resource instances for portfolio to be routed to D ynamic instantiation |
| CN105052074A (en) * | 2013-02-18 | 2015-11-11 | 泰科来股份有限公司 | Methods, systems, and computer readable media for providing a virtualized diameter network architecture and for routing traffic to dynamically instantiated diameter resource instances |
| CN105052080A (en) * | 2013-02-18 | 2015-11-11 | 泰科来股份有限公司 | Methods, systems, and computer readable media for providing a thinking diameter network architecture |
| CN105052080B (en) * | 2013-02-18 | 2019-05-03 | 泰科来股份有限公司 | For providing method, system and the computer-readable medium of thinking diameter network framework |
| US11388082B2 (en) | 2013-11-27 | 2022-07-12 | Oracle International Corporation | Methods, systems, and computer readable media for diameter routing using software defined network (SDN) functionality |
| CN106888237A (en) * | 2015-12-15 | 2017-06-23 | 中国移动通信集团公司 | A kind of data dispatching method and system |
| CN106888237B (en) * | 2015-12-15 | 2020-01-07 | 中国移动通信集团公司 | Data scheduling method and system |
| CN112956172A (en) * | 2018-10-31 | 2021-06-11 | 思科技术公司 | Transaction-based event tracking mechanism |
| CN112956172B (en) * | 2018-10-31 | 2024-02-23 | 思科技术公司 | Transaction-based event tracking mechanism |
| CN110855646A (en) * | 2019-10-31 | 2020-02-28 | 苏州经贸职业技术学院 | Approach loading system for network computer and application method |
| CN110855646B (en) * | 2019-10-31 | 2021-08-06 | 苏州经贸职业技术学院 | Approach loading system for network computer and application method |
Also Published As
| Publication number | Publication date |
|---|---|
| CN102195890B (en) | 2014-04-16 |
Similar Documents
| Publication | Publication Date | Title |
|---|---|---|
| CN102195890B (en) | Internet application dispatching method based on cloud computing | |
| CN103118124B (en) | A kind of cloud computing load balancing method based on the many agencies of layering | |
| Zhu et al. | Study on cloud computing resource scheduling strategy based on the ant colony optimization algorithm | |
| CN101938416B (en) | Cloud computing resource scheduling method based on dynamic reconfiguration virtual resources | |
| CN106534318B (en) | A kind of OpenStack cloud platform resource dynamic scheduling system and method based on flow compatibility | |
| CN103106222B (en) | The processing method and processing device of intervalometer | |
| CN102981890A (en) | Computing task and virtual machine deploying method within a virtual data center | |
| CN109005057A (en) | A kind of computational resource allocation based on contract matching theory and task unload scheme | |
| CN110231976B (en) | Load prediction-based edge computing platform container deployment method and system | |
| CN104283946A (en) | Resource self-adaptive adjusting system and method of multiple virtual machines under single physical machine | |
| CN104902001B (en) | Web request load-balancing method based on operating system virtualization | |
| CN103605578B (en) | Load equilibration scheduling method based on virtual machine (vm) migration | |
| CN106598733A (en) | Three-dimensional virtual resource scheduling method of cloud computing energy consumption key | |
| CN103713956A (en) | Method for intelligent weighing load balance in cloud computing virtualized management environment | |
| CN104317646B (en) | Based on cloud data center dispatching method of virtual machine under OpenFlow frameworks | |
| CN103077082A (en) | Method and system for distributing data center load and saving energy during virtual machine migration | |
| CN103488539A (en) | Data center energy saving method based on central processing unit (CPU) dynamic frequency modulation technology | |
| CN104023042B (en) | Cloud platform resource scheduling method | |
| CN105426241A (en) | Cloud computing data center based unified resource scheduling energy-saving method | |
| CN104735095A (en) | Method and device for job scheduling of cloud computing platform | |
| CN105446816A (en) | Heterogeneous platform oriented energy consumption optimization scheduling method | |
| CN104536803A (en) | Virtual machine scheduling method based on combination optimization | |
| CN104765642A (en) | Virtual machine deploying method and system based on dynamic prediction model in cloud environment | |
| CN106714323A (en) | Priority distinguishing and dispatching method of heterogeneous flow in 4G network | |
| CN103856355A (en) | Energy consumption perceptive virtual network mapping transportation model and algorithm |
Legal Events
| Date | Code | Title | Description |
|---|---|---|---|
| C06 | Publication | ||
| PB01 | Publication | ||
| C10 | Entry into substantive examination | ||
| SE01 | Entry into force of request for substantive examination | ||
| C14 | Grant of patent or utility model | ||
| GR01 | Patent grant | ||
| CF01 | Termination of patent right due to non-payment of annual fee |
Granted publication date: 20140416 Termination date: 20170603 |
|
| CF01 | Termination of patent right due to non-payment of annual fee |