CN104270322B - Internet of things oriented equipment accesses the self-adapted load balance scheduling mechanism of processing platform - Google Patents
Internet of things oriented equipment accesses the self-adapted load balance scheduling mechanism of processing platform Download PDFInfo
- Publication number
- CN104270322B CN104270322B CN201410597843.4A CN201410597843A CN104270322B CN 104270322 B CN104270322 B CN 104270322B CN 201410597843 A CN201410597843 A CN 201410597843A CN 104270322 B CN104270322 B CN 104270322B
- Authority
- CN
- China
- Prior art keywords
- node
- routing
- algorithm
- load
- internet
- 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.)
- Active
Links
Landscapes
- Data Exchanges In Wide-Area Networks (AREA)
Abstract
The present invention relates to technical field of the computer network, more particularly to the self-adapted load balance scheduling mechanism of internet of things oriented equipment access processing platform, the present invention proposes routing policy and the new dispatching algorithm frame of routing algorithm independent separate, routing policy is replaceable, evaluation index is expansible, the features such as routing algorithm is extendible, and provide the sudden and randomness that three kinds of routing policies tackle equipment connection request in true application environment, including single algorithm policy, combinational algorithm strategy and adaptation algorithm strategy.Using the present invention, internet of things equipment access system according to the distribution condition of the performance difference of each node in cluster and the effective adjustment equipment connection request of load state, can promote the overall performance of group system.Compared with prior art, the present invention has the advantages that novel thought, using flexible and extension are convenient, fully meets the demand of things-internet gateway equipment connection request scheduling, the limitation of all scenes can not be applicable in by efficiently solving single routing algorithm.
Description
Technical field
The present invention relates to technical field of the computer network more particularly to internet of things oriented equipment to access the adaptive of processing platform
Answer load balance scheduling mechanism.
Background technology
With the arrival of the Internet of things era, more and more awareness apparatus need access platform system, and pass through network side
Just Internet of Things data information is quickly obtained.In order to solve the concurrent access platform system of large number of equipment, using distributed type assemblies portion
Administration becomes commonplace solution.
Load-balancing technique is accompanied by the appearance of cluster and occurs, it be realize group system key technology it
One, its quality directly affects the performance of whole system.Effect phase is gone so far as from the static load balancing strategy of early stage
To preferable Dynamic Load-Balancing Strategy and from centerized fusion (passive type cluster) to distributed AC servo system (active concentrating)
Etc..
The design philosophy of Dynamic Load-Balancing Strategy is being capable of automatic adaptation system parameter variations and unknown load spy
Property control load distribution, and there is passive type clustered control and active concentrating control in different develop of control algolithm.
According to the different phase that task is distributed, load-balancing algorithm can be divided into two classes, static load balancing algorithm and
Dynamic Load-balancing Algorithm.It is the task requests stage first, system assigns them to collection according to the allocation algorithm pre-set
Some node in group, i.e. original allocation, are called static scheduling algorithm in this stage allocation algorithm.Followed by run in cluster
In the process, when node failure or node overload occurs, task can be reallocated by sub-distribution again, in this stage allocation algorithm
It is called dynamic dispatching algorithm.
The content of the invention
The present invention is to overcome above-mentioned shortcoming, and it is an object of the present invention to provide internet of things oriented equipment accesses oneself of processing platform
Load balance scheduling mechanism is adapted to, using routing policy new dispatching algorithm frame separated with routing algorithm, overcomes existing collection
The defects of single routing algorithm can not be applicable in all scenes in group's system.
The present invention is to reach above-mentioned purpose by the following technical programs:Internet of things oriented equipment accesses the adaptive of processing platform
Load balance scheduling mechanism is answered, is comprised the following steps:
1) routing node sets routing policy according to cluster current operating environment, and associates road required under the routing policy
By algorithm, which comes into force;
2) processing node obtains the loading level measurement index of this node, and loading level measurement index is connected with having distributed
Number is packaged and is periodically sent to routing node;
3) routing node receives new gateway device connection request, and performing corresponding routing algorithm according to routing policy obtains
To candidate target, then destination node is generated from candidate target, connection request is distributed to destination node;
4) routing node receives loading level measurement index and the distribution connection number that processing node is sent, and judges that each node is
No overload, routing node take out the connection of distribution on overload node, are reallocated again according to routing policy
Work keeps the dynamic load leveling of cluster;
5) when running environment changes, step 1) is jumped to.
Preferably, the routing policy and routing algorithm independent separate, routing policy is upper layer module, and routing algorithm is
Lower module.
Preferably, routing policy is:Single algorithm policy or combinational algorithm strategy or adaptation algorithm strategy;
Single algorithm policy is for equipment access situation stabilization and has the running environment of adaptable specific routing algorithm;
Combinational algorithm strategy is for the unknown running environment of equipment access situation;
Adaptation algorithm strategy is obvious and have specific routing algorithm to be applicable in correspondence for equipment access situation alternating rule
The running environment of situation.
Preferably, cluster is divided into different operating modes, adaptation algorithm according to loading condition and distribution connection
Strategy selects the optimal routing algorithm of collocation degree according to the operating mode of cluster.
Preferably, the loading level measurement index is loading level saturation value, calculated by following steps:
(1) the load information sampling period set is obtained when handling node initializing;
(2) processing node gathers a variety of machine performance indexs of current the machine when sampling time node reaches, and makes including CPU
With rate, memory usage, disk utilization rate and bandwidth utilization rate, and each desired value is normalized;
(3) handle node and the disposal of gentle filter is carried out to a variety of machine performance indexs, reduce " noise " in measured data
It influences;
(4) model is calculated by multiple input single output mathematical model as input parameter in a variety of machine performance indexs
The scalar from 0 to 1 is enclosed, which is expressed as loading level saturation value.
Preferably, the determination methods of overload are:Loading level saturation value be more than setting load control threshold or
Difference in person's loading level saturation value and cluster between minimum load degree saturation value is more than the threshold value of setting.
The beneficial effects of the present invention are:(1) routing policy new dispatching algorithm frame separated with routing algorithm is used,
Making the division of labor of routing algorithm and routing policy, definitely responsibility becomes apparent from;(2) routing algorithm can be used as unit member into
Row combination and adaptation, the time of running, which changes routing algorithm, can effectively evade the limitation of single algorithm;(3) three kinds of routing policies
Complement each other, can adapt to environment of internet of things complicated and changeable, and add in distribution target the survival of the fittest Pre-Evaluation mechanism into
The better adaptive scheduling of row, keeps cluster dynamic load leveling;(4) platform can be easily integrated new routing algorithm, evaluation
Function also can spread, the requirement of open type software architecture design is fully achieved.
Description of the drawings
Fig. 1 is the principle of the present invention structure diagram;
Fig. 2 is the stream for the self-adapted load balance scheduling mechanism that the present invention realizes internet of things oriented equipment access processing platform
Journey schematic diagram;
Fig. 3 is to handle the update flow diagram that node obtains loading level measurement index;
Fig. 4 is the flow diagram of progress dispatching distribution after routing node accepting device connection request of the invention;
Fig. 5 is the single algorithm policy flow diagram in routing policy of the present invention;
Fig. 6 is the combinational algorithm strategic process schematic diagram in routing policy of the present invention;
Fig. 7 is the adaptation algorithm strategic process schematic diagram in routing policy of the present invention;
Fig. 8 is that the routing node of the present invention keeps the flow diagram of cluster dynamic load leveling;
Fig. 9 is the extension adaptation algorithm strategic process schematic diagram of the present invention.
Specific embodiment
With reference to specific embodiment, the present invention is described further, but protection scope of the present invention is not limited in
This:
Embodiment 1:As shown in Figure 1, the self-adapted load balance scheduling of the present invention is realized by routing node, processing node,
Routing node is responsible for the load balancing between the dispatching distribution and clustered node of newly-built equipment connection request, and processing node is responsible in real time
It monitors the load state of this node and carries out the business processing of all access devices of this section point.When gateway device access platform,
Processing of the dispatching distribution to some processing node progress data and business, when cluster load imbalance, road are responsible for by routing node
It is responsible for handle on the processing node of the equipment connection migration on heavier loads node to other light loads by node.
As shown in Fig. 2, the self-adapted load balance scheduling mechanism of internet of things oriented equipment access processing platform, using routing
The load condition of node real time monitoring cluster processing node and distribution condition, strategy and the calculation of adaptive adjusting device access processing
The separated thought of method, scheduling mechanism can carry out maintenance collection on higher abstraction hierarchy in a manner of being most preferably adapted to running environment
The work of group's load balancing, so as to ensure the overall performance of platform and stable operation.
It realizes above-mentioned self-adapted load balance scheduling mechanism, comprises the following steps:
1) routing node sets routing policy according to cluster current operating environment, and associates road required under the routing policy
By algorithm, which comes into force;
2) processing node obtains the loading level measurement index of this node, and loading level measurement index is connected with having distributed
Number is packaged and is periodically sent to routing node;
3) routing node receives new gateway device connection request, and performing corresponding routing algorithm according to routing policy obtains
To candidate target, then destination node is generated from candidate target, connection request is distributed to destination node;
4) routing node receives loading level measurement index and the distribution connection number that processing node is sent, and judges that each node is
No overload, routing node take out the connection of distribution on overload node, are reallocated again according to routing policy
Work keeps the dynamic load leveling of cluster;
5) when running environment changes, step 1) is jumped to.
Processing node mainly from but be not limited to the load condition that following four performance indicator investigates this node, the load condition
Referred to as loading level measurement index, including:CPU usage, memory usage, disk utilization rate, network bandwidth utilization rate, and lead to
It crosses function model and loading level saturation value is calculated uniformly to weigh the current load information of calculate node, is i.e. loading level weighs
Figureofmerit is represented with loading level saturation value.
CPU usage, the CPU usage of present sample timing node.
Memory usage, the CPU usage of present sample timing node.
Disk utilization rate, the disk utilization rate of present sample timing node.
Network bandwidth utilization rate, the network bandwidth utilization rate of present sample timing node.
Loading level saturation value, the abstract representation of the machine loading information of present sample timing node, value range [0,
1], 0 represent completely non-loaded, 1 represents complete full load.
As shown in figure 3, loading level measurement index by processing node by periodically acquiring, more new technological process detailed process bag
Include following steps:
Step 301:The TCP long connections of node initializing and routing node are handled, and the load information sampling period is set;
Step 302:Processing node is periodically executed step 303 to step 306 according to the sampling period;
Step 303:Processing node gathers a variety of machine performance indexs of current the machine when sampling time node reaches, and right
Each desired value is normalized;
Step 304:It handles node and low-pass filtering treatment is carried out to a variety of machine performance indexs, reduce in measured data
" noise " influences, and obtains smooth data output;
Step 305:The load function model meter that a variety of machine performance indexs pass through multiple input single output as input parameter
Calculation obtains load information, which is scalar of the scope from 0 to 1, which is expressed as loading level saturation value, bears
It carries function model and can be used but not limited to and performance index value is converted by multiple medians and then linear by positive and negative function pair
Weighting obtains unique output valve;
Step 306:Loading level saturation value is reported to routing node by processing node by TCP long connections;
Present invention sampling routing policy and the new dispatching algorithm frame of routing algorithm independent separate, that is, pass through two layers of design
Module, lower module are routing algorithms, and upper layer module is routing policy.Candidate when routing algorithm mainly performs gateway device access
The evaluation work of target is distributed, final distribution target is unified to be assessed to obtain by routing policy.Provide three kinds of routing policies
The sudden and randomness of equipment connection request in true application environment is tackled, including single algorithm policy, combinational algorithm strategy
With adaptation algorithm strategy.Routing algorithm includes a variety of current main-stream routing algorithms, figures method including weighting Smallest connection, poll is calculated
Method, distributed AC servo system algorithm and ant group algorithm can carry out free expansion in platform.
When routing node receives new gateway device connection request, performing corresponding routing algorithm according to routing policy obtains
To candidate target, then destination node is generated from candidate target, connection request is distributed to destination node.Detailed process include with
Lower step, as shown in Figure 4:
Step 401:New gateway device access or when having distributing equipment connection and distributing again, routing node starts distribution
Task;
Step 402:Routing node obtain configuration file in routing policy with associate the configuration informations such as routing algorithm, the configuration
Information can change at any time;
Step 403:Routing node generates corresponding distribution target according to different routing policy patterns:
The use number of single algorithm policy limit algorithm is 1 but does not limit using which kind of algorithm, can be run in platform
Moment carries out replacing used specific algorithm, causes certain special algorithm adaptable suitable for equipment access situation stabilization
Scene.Single algorithm policy flow diagram is as shown in figure 5, single algorithm policy performs the algorithm A1 set, point of generation
It is candidate target with target, which is final destination node.When cluster running environment changes, algorithm A1 is uncomfortable
With current environment, the algorithm B1 for adapting to current operating environment is replaced into algorithm A1 by setting.
There is no limit can replace it to the number and species that combinational algorithm strategy uses algorithm in the platform time of running
Middle one or more algorithm, it is unknown suitable for equipment access situation and to the scene that single algorithm is caused not adapt to effectively.Group
Hop algorithm strategic process schematic diagram calculates corresponding evaluation index simultaneously as shown in fig. 6, each algorithm generates a candidate target
Value, the candidate target of evaluation index value maximum is as distribution target, which is destination node, and algorithm can be replaced at any time
It changes.
Adaptation algorithm strategy belongs to empirical mode, as shown in fig. 7, adaptation algorithm strategy is selected according to current cluster operating mode
Take the algorithm that collocation degree is best, the candidate target that algorithm generates is to distribute target, and synchronization only exists an algorithm and not
Alternatively.Typical operating mode is including but not limited to following two:Active mode and inactive pattern.Active mode represents flat
The new gateway device connection request of platform is less or more stable, and the business data processing motion frequency of access device compared with
It is high.Inactive pattern represents that the data processing activity frequency of platform access device is relatively low, and new gateway device connection request
More or situation is unknown.
Step 404:Routing node will be in gateway device connection distribution to destination node;
Step 405:Judge that whether all devices connection request is disposed in distribution task, enters if processing terminates
In next step, otherwise it is back to step 402.
Step 406:Routing node continues listening port and waits new distribution task.
In order to keep cluster dynamic load leveling, routing node receive loading level measurement index that processing node sends with
Distribution connection number, judge each node whether overload, routing node take out overload node on distribution connection, again
Reallocation work is carried out according to routing policy, as shown in figure 8, specifically including following steps:
Step 801:The loading level saturation value that each processing node reports by TCP long connections is monitored and received to routing node
The information such as connection number are distributed, specific loading level saturation value more new technological process is referring to Fig. 3;
Step 802:Routing node updates cluster state and operating mode, and according to loading level saturation value size to all
Processing node is ranked up, and finds out the processing node (i.e. the load saturation angle value of the node is maximum) of pack heaviest;
Step 803:Routing node judges whether current cluster meets the condition of state transition, into next if meeting
Step, is otherwise back to step 801, and Rule of judgment is more than the load control threshold of setting or load journey for loading level saturation value
Spend the threshold value that the difference in saturation value and cluster between minimum load degree saturation value is more than setting.
Step 804:Routing node activation processing sub-line journey, starts connection migration task, sets source mesh where migration connection
It is designated as the relevant informations such as the node of pack heaviest and required migration connection number;
Step 805:A gateway device connection is obtained from the target of source and is connected associated facility information and industry with this
Business contextual information etc.;
Step 806:Dispatching distribution workflow is connected by the connection migration to new distribution target according to the equipment of Fig. 4
On;
Step 807:The contextual information of the connection and facility information are synchronized in new distribution node;
Step 808:Judge whether required migration connection has all been disposed in migration task, if processing terminates
(i.e. step 809) is otherwise back to step 805 to dormancy sub-line journey.
Illustrate how flexible expansion is carried out to the new dispatching algorithm frame of the present invention to increase adaptation algorithm strategy newly, expand
Flow is opened up referring to Fig. 9, detailed process comprises the following steps:
Step 901:Platform first passes through the testing results of combinational algorithm strategy and single algorithm policy, finds out working platform mould
The feature of the formula algorithm best with corresponding collocation degree;
Step 902:Judge whether to need to add new routing algorithm to generate candidate target, if need to if realize interface
RouteStrategyService, otherwise into next step;
Step 903:Judge whether to need to change the cluster mode realization for dividing operating mode according to operation characteristic, if needing
Realize interface FamilyClusterStatus<T extends Object>, otherwise into next step;
Step 904:Judge whether need change acquiescence candidate target evaluation criterion realize, if need to if realize interface
ScoreVoter, otherwise into next step;
Step 905:It realizes that interface PolicyVoter completes routing policy and does traffic control, that is, realizes under different mode
Select the traffic control of best-fit algorithm and the candidate target generated according to evaluation criterion filtering algorithm;
Step 906:All realization classes are packaged and are re-deployed in platform running environment, modification configuration file Road
It is newest realization class by algorithm and routing policy, the adaptation algorithm strategy that platform dynamic load increases newly.
The above technical principle for being specific embodiments of the present invention and being used, if conception under this invention institute
The change of work during the spirit that generated function is still covered without departing from specification and attached drawing, should belong to the present invention's
Protection domain.
Claims (4)
1. internet of things oriented equipment accesses the self-adapted load balance scheduling mechanism of processing platform, which is characterized in that including following
Step:
1) routing node sets routing policy according to cluster current operating environment, and associates routing required under the routing policy and calculate
Method, the configuration come into force;
2) processing node obtains the loading level measurement index of this node, and loading level measurement index is connected several beat with having distributed
Bag is periodically sent to routing node;
3) routing node receives new gateway device connection request, and performing corresponding routing algorithm according to routing policy is waited
Target is selected, then destination node is generated from candidate target, connection request is distributed to destination node;
4) routing node receives loading level measurement index and the distribution connection number that processing node is sent, and judges whether each node is born
Carry it is overweight, routing node take out overload node on distribution connection, reallocation work is carried out according to routing policy again,
Keep the dynamic load leveling of cluster;Wherein, the processing method after discovery node load is overweight is as follows:
4.1) routing node activation processing sub-line journey, starts connection migration task, and source target where setting migration connection is load
Most heavy node and the relevant information of required migration connection number;
4.2) a gateway device connection is obtained from the target of source and is connected associated facility information and business contexts with this
Information;
4.3) connecting dispatching distribution workflow according to equipment will be on the connection migration to new distribution target;
4.4) contextual information of the connection and facility information are synchronized in new distribution node;
4.5) judge whether required migration connection has all been disposed in migration task, the dormancy sub-line if processing terminates
Otherwise journey returns and performs step 4.2);
5) when running environment changes, step 1) is jumped to.
2. the self-adapted load balance scheduling mechanism of internet of things oriented equipment access processing platform according to claim 1,
It is characterized in that, the routing policy and routing algorithm independent separate, routing policy is upper layer module, and routing algorithm is lower floor's mould
Block.
3. the self-adapted load balance scheduling mechanism of internet of things oriented equipment access processing platform according to claim 1,
It is characterized in that, the loading level measurement index is loading level saturation value, calculated by following steps:
(1) the load information sampling period set is obtained when handling node initializing;
(2) processing node gathers a variety of machine performance indexs of current the machine when sampling time node reaches, and is used including CPU
Rate, memory usage, disk utilization rate and bandwidth utilization rate, and each desired value is normalized;
(3) handle node and the disposal of gentle filter is carried out to a variety of machine performance indexs, " noise " reduced in measured data influences;
(4) a variety of machine performance indexs as input parameter by multiple input single output mathematical model be calculated a scope from
0 to 1 scalar, the scalar are expressed as loading level saturation value.
4. the self-adapted load balance scheduling mechanism of internet of things oriented equipment access processing platform according to claim 3,
It is characterized in that, the determination methods of overload are:Loading level saturation value is more than load control threshold or the load of setting
Difference in degree saturation value and cluster between minimum load degree saturation value is more than the threshold value of setting.
Priority Applications (1)
Application Number | Priority Date | Filing Date | Title |
---|---|---|---|
CN201410597843.4A CN104270322B (en) | 2014-10-30 | 2014-10-30 | Internet of things oriented equipment accesses the self-adapted load balance scheduling mechanism of processing platform |
Applications Claiming Priority (1)
Application Number | Priority Date | Filing Date | Title |
---|---|---|---|
CN201410597843.4A CN104270322B (en) | 2014-10-30 | 2014-10-30 | Internet of things oriented equipment accesses the self-adapted load balance scheduling mechanism of processing platform |
Publications (2)
Publication Number | Publication Date |
---|---|
CN104270322A CN104270322A (en) | 2015-01-07 |
CN104270322B true CN104270322B (en) | 2018-05-29 |
Family
ID=52161814
Family Applications (1)
Application Number | Title | Priority Date | Filing Date |
---|---|---|---|
CN201410597843.4A Active CN104270322B (en) | 2014-10-30 | 2014-10-30 | Internet of things oriented equipment accesses the self-adapted load balance scheduling mechanism of processing platform |
Country Status (1)
Country | Link |
---|---|
CN (1) | CN104270322B (en) |
Families Citing this family (15)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
CN105099912A (en) * | 2015-08-07 | 2015-11-25 | 浪潮电子信息产业股份有限公司 | Multipath data scheduling method and device |
CN107360067B (en) * | 2017-06-29 | 2021-08-10 | 青岛恒金源电子科技有限公司 | Network load balancing method in intelligent household system |
CN107547424B (en) * | 2017-07-18 | 2020-09-25 | 深圳市盛路物联通讯技术有限公司 | Internet of things equipment mounting control method and equipment |
CN107613456B (en) * | 2017-07-19 | 2020-07-03 | 深圳市盛路物联通讯技术有限公司 | Information management method and equipment based on Internet of things |
CN107547630B (en) * | 2017-07-21 | 2021-01-26 | 深圳市盛路物联通讯技术有限公司 | Equipment mounting control system of Internet of things |
CN107786460A (en) * | 2017-09-08 | 2018-03-09 | 北京科东电力控制系统有限责任公司 | A kind of management of electricity transaction system request and current-limiting method based on token bucket algorithm |
CN107807853B (en) * | 2017-10-16 | 2021-07-02 | 北京航空航天大学 | Node screening method and device based on machine real-time load and task state machine |
CN111083181B (en) * | 2018-10-19 | 2023-02-17 | 北京奇虎科技有限公司 | Equipment scheduling method, gateway and system for large-scale access to Internet of things |
CN110413389B (en) * | 2019-07-24 | 2021-09-28 | 浙江工业大学 | Task scheduling optimization method under resource imbalance Spark environment |
CN110933512B (en) * | 2019-10-23 | 2022-05-06 | 视联动力信息技术股份有限公司 | Load determination method and device based on video network |
CN110795250B (en) * | 2019-10-30 | 2023-05-12 | 腾讯科技(深圳)有限公司 | Load scheduling method, device, equipment and storage medium |
CN112003922B (en) * | 2020-08-18 | 2023-06-06 | 杭州指令集智能科技有限公司 | Data transmission method and device |
CN114297263A (en) * | 2021-12-30 | 2022-04-08 | 浙江邦盛科技股份有限公司 | Load self-adaptive distributed space text stream data processing system |
CN115473753B (en) * | 2022-07-20 | 2024-07-19 | 青岛海尔科技有限公司 | Gateway equipment control method and device, storage medium and electronic device |
CN116827975A (en) * | 2022-12-08 | 2023-09-29 | 慧之安信息技术股份有限公司 | Algorithm application method based on equipment model |
Citations (4)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
CN101631034A (en) * | 2008-07-15 | 2010-01-20 | 华为技术有限公司 | Method, device and system for node management and access in peer-to-peer network |
CN103227754A (en) * | 2013-04-16 | 2013-07-31 | 浪潮(北京)电子信息产业有限公司 | Dynamic load balancing method of high-availability cluster system, and node equipment |
CN103368864A (en) * | 2013-07-31 | 2013-10-23 | 北京华易互动科技有限公司 | Intelligent load balancing method based on c/s (Client/Server) architecture |
CN103401939A (en) * | 2013-08-08 | 2013-11-20 | 中国航天科工集团第三研究院第八三五七研究所 | Load balancing method adopting mixing scheduling strategy |
Family Cites Families (1)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
US9077640B2 (en) * | 2012-01-13 | 2015-07-07 | Board Of Regents, The University Of Texas System | Method and system of congestion control in a mobile virtual network |
-
2014
- 2014-10-30 CN CN201410597843.4A patent/CN104270322B/en active Active
Patent Citations (4)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
CN101631034A (en) * | 2008-07-15 | 2010-01-20 | 华为技术有限公司 | Method, device and system for node management and access in peer-to-peer network |
CN103227754A (en) * | 2013-04-16 | 2013-07-31 | 浪潮(北京)电子信息产业有限公司 | Dynamic load balancing method of high-availability cluster system, and node equipment |
CN103368864A (en) * | 2013-07-31 | 2013-10-23 | 北京华易互动科技有限公司 | Intelligent load balancing method based on c/s (Client/Server) architecture |
CN103401939A (en) * | 2013-08-08 | 2013-11-20 | 中国航天科工集团第三研究院第八三五七研究所 | Load balancing method adopting mixing scheduling strategy |
Also Published As
Publication number | Publication date |
---|---|
CN104270322A (en) | 2015-01-07 |
Similar Documents
Publication | Publication Date | Title |
---|---|---|
CN104270322B (en) | Internet of things oriented equipment accesses the self-adapted load balance scheduling mechanism of processing platform | |
Praveen et al. | An Adaptive Load Balancing Technique for Multi SDN Controllers | |
Yadav et al. | Smart healthcare: RL-based task offloading scheme for edge-enable sensor networks | |
Assefa et al. | A survey of energy efficiency in SDN: Software-based methods and optimization models | |
Zheng et al. | An approach for cloud resource scheduling based on Parallel Genetic Algorithm | |
US8510747B2 (en) | Method and device for implementing load balance of data center resources | |
Yakhchi et al. | Proposing a load balancing method based on Cuckoo Optimization Algorithm for energy management in cloud computing infrastructures | |
Qilin et al. | A load balancing method based on SDN | |
EP3549312B1 (en) | A master node, a local node and respective methods performed thereby for predicting one or more metrics associated with a communication network | |
CN108182105A (en) | Local dynamic station moving method and control system based on Docker container techniques | |
CN103957280B (en) | Connection allocation and scheduling method of sensor network in Internet of things | |
CN109271257A (en) | A kind of method and apparatus of virtual machine (vm) migration deployment | |
CN108200156A (en) | The dynamic load balancing method of distributed file system under a kind of cloud environment | |
CN106330558A (en) | Controller load prediction system and method applied to software defined network | |
Seung et al. | Cloudflex: Seamless scaling of enterprise applications into the cloud | |
Fröhlich et al. | Smart SDN management of fog services | |
CN105247834B (en) | The distribution method of Internet resources, composer and manager in virtual network function | |
CN108319501B (en) | Elastic resource supply method and system based on micro-service gateway | |
CN106059940B (en) | A kind of flow control methods and device | |
CN109960579B (en) | Method and device for adjusting service container | |
CN111865817A (en) | Load balancing control method, device and equipment for remote measuring collector and storage medium | |
Dalla-Costa et al. | Orchestra: A customizable split-aware NFV orchestrator for dynamic cloud radio access networks | |
Mada et al. | Latency-aware service placement and live migrations in 5G and beyond mobile systems | |
Pusztai et al. | Polaris scheduler: SLO-and topology-aware microservices scheduling at the edge | |
Qin et al. | Research on nginx dynamic load balancing 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 | ||
CB02 | Change of applicant information |
Address after: Yuhang District, Hangzhou City, Zhejiang Province, 311121 West No. 1500 Building 1 room 311 Applicant after: Zhong electricity Haikang Group Co.,Ltd Address before: Ma Cheng Road Hangzhou City, Zhejiang province 310012 No. 36 Applicant before: Zhong electricity Haikang Group Co.,Ltd |
|
CB02 | Change of applicant information | ||
GR01 | Patent grant | ||
GR01 | Patent grant |