CN113110914A - Internet of things platform construction method based on micro-service architecture - Google Patents
Internet of things platform construction method based on micro-service architecture Download PDFInfo
- Publication number
- CN113110914A CN113110914A CN202110229566.1A CN202110229566A CN113110914A CN 113110914 A CN113110914 A CN 113110914A CN 202110229566 A CN202110229566 A CN 202110229566A CN 113110914 A CN113110914 A CN 113110914A
- Authority
- CN
- China
- Prior art keywords
- module
- internet
- micro
- service
- prediction
- Prior art date
- Legal status (The legal status is an assumption and is not a legal conclusion. Google has not performed a legal analysis and makes no representation as to the accuracy of the status listed.)
- Pending
Links
- 238000010276 construction Methods 0.000 title claims abstract description 21
- 238000012544 monitoring process Methods 0.000 claims abstract description 55
- 238000004422 calculation algorithm Methods 0.000 claims abstract description 25
- 230000006870 function Effects 0.000 claims abstract description 15
- 238000004088 simulation Methods 0.000 claims abstract description 8
- 238000000034 method Methods 0.000 claims description 43
- 230000009467 reduction Effects 0.000 claims description 15
- 238000007689 inspection Methods 0.000 claims description 13
- 238000007726 management method Methods 0.000 claims description 11
- 238000004458 analytical method Methods 0.000 claims description 10
- 230000008569 process Effects 0.000 claims description 10
- 238000012545 processing Methods 0.000 claims description 7
- 238000009825 accumulation Methods 0.000 claims description 5
- 238000004590 computer program Methods 0.000 claims description 5
- 238000012360 testing method Methods 0.000 claims description 5
- 238000013480 data collection Methods 0.000 claims description 4
- 238000013459 approach Methods 0.000 claims description 3
- 238000012550 audit Methods 0.000 claims description 3
- 238000004364 calculation method Methods 0.000 claims description 3
- 239000011159 matrix material Substances 0.000 claims description 3
- 238000005070 sampling Methods 0.000 claims description 3
- 238000006467 substitution reaction Methods 0.000 claims description 3
- 230000002087 whitening effect Effects 0.000 claims description 3
- 238000013439 planning Methods 0.000 claims description 2
- 230000007246 mechanism Effects 0.000 abstract description 4
- 230000008602 contraction Effects 0.000 abstract description 2
- 238000010586 diagram Methods 0.000 description 5
- 230000008878 coupling Effects 0.000 description 4
- 238000010168 coupling process Methods 0.000 description 4
- 238000005859 coupling reaction Methods 0.000 description 4
- 238000005516 engineering process Methods 0.000 description 4
- 230000004044 response Effects 0.000 description 4
- 230000008859 change Effects 0.000 description 3
- 230000007547 defect Effects 0.000 description 3
- 238000011161 development Methods 0.000 description 3
- 230000008901 benefit Effects 0.000 description 2
- 238000004891 communication Methods 0.000 description 2
- 238000012423 maintenance Methods 0.000 description 2
- 230000003068 static effect Effects 0.000 description 2
- 239000002699 waste material Substances 0.000 description 2
- 238000013500 data storage Methods 0.000 description 1
- 239000000835 fiber Substances 0.000 description 1
- 230000006872 improvement Effects 0.000 description 1
- 238000012986 modification Methods 0.000 description 1
- 230000004048 modification Effects 0.000 description 1
- 230000006855 networking Effects 0.000 description 1
- 230000003287 optical effect Effects 0.000 description 1
- 238000005457 optimization Methods 0.000 description 1
- 230000008092 positive effect Effects 0.000 description 1
- 238000013468 resource allocation Methods 0.000 description 1
- 239000004065 semiconductor Substances 0.000 description 1
- 239000007787 solid Substances 0.000 description 1
- 238000000638 solvent extraction Methods 0.000 description 1
Images
Classifications
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06F—ELECTRIC DIGITAL DATA PROCESSING
- G06F9/00—Arrangements for program control, e.g. control units
- G06F9/06—Arrangements for program control, e.g. control units using stored programs, i.e. using an internal store of processing equipment to receive or retain programs
- G06F9/44—Arrangements for executing specific programs
- G06F9/455—Emulation; Interpretation; Software simulation, e.g. virtualisation or emulation of application or operating system execution engines
- G06F9/45533—Hypervisors; Virtual machine monitors
- G06F9/45558—Hypervisor-specific management and integration aspects
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06F—ELECTRIC DIGITAL DATA PROCESSING
- G06F11/00—Error detection; Error correction; Monitoring
- G06F11/30—Monitoring
- G06F11/3003—Monitoring arrangements specially adapted to the computing system or computing system component being monitored
- G06F11/3006—Monitoring arrangements specially adapted to the computing system or computing system component being monitored where the computing system is distributed, e.g. networked systems, clusters, multiprocessor systems
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06F—ELECTRIC DIGITAL DATA PROCESSING
- G06F11/00—Error detection; Error correction; Monitoring
- G06F11/30—Monitoring
- G06F11/3003—Monitoring arrangements specially adapted to the computing system or computing system component being monitored
- G06F11/301—Monitoring arrangements specially adapted to the computing system or computing system component being monitored where the computing system is a virtual computing platform, e.g. logically partitioned systems
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06F—ELECTRIC DIGITAL DATA PROCESSING
- G06F9/00—Arrangements for program control, e.g. control units
- G06F9/06—Arrangements for program control, e.g. control units using stored programs, i.e. using an internal store of processing equipment to receive or retain programs
- G06F9/46—Multiprogramming arrangements
- G06F9/50—Allocation of resources, e.g. of the central processing unit [CPU]
- G06F9/5005—Allocation of resources, e.g. of the central processing unit [CPU] to service a request
- G06F9/5027—Allocation of resources, e.g. of the central processing unit [CPU] to service a request the resource being a machine, e.g. CPUs, Servers, Terminals
- G06F9/5038—Allocation of resources, e.g. of the central processing unit [CPU] to service a request the resource being a machine, e.g. CPUs, Servers, Terminals considering the execution order of a plurality of tasks, e.g. taking priority or time dependency constraints into consideration
-
- G—PHYSICS
- G16—INFORMATION AND COMMUNICATION TECHNOLOGY [ICT] SPECIALLY ADAPTED FOR SPECIFIC APPLICATION FIELDS
- G16Y—INFORMATION AND COMMUNICATION TECHNOLOGY SPECIALLY ADAPTED FOR THE INTERNET OF THINGS [IoT]
- G16Y30/00—IoT infrastructure
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06F—ELECTRIC DIGITAL DATA PROCESSING
- G06F9/00—Arrangements for program control, e.g. control units
- G06F9/06—Arrangements for program control, e.g. control units using stored programs, i.e. using an internal store of processing equipment to receive or retain programs
- G06F9/44—Arrangements for executing specific programs
- G06F9/455—Emulation; Interpretation; Software simulation, e.g. virtualisation or emulation of application or operating system execution engines
- G06F9/45533—Hypervisors; Virtual machine monitors
- G06F9/45558—Hypervisor-specific management and integration aspects
- G06F2009/45562—Creating, deleting, cloning virtual machine instances
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06F—ELECTRIC DIGITAL DATA PROCESSING
- G06F9/00—Arrangements for program control, e.g. control units
- G06F9/06—Arrangements for program control, e.g. control units using stored programs, i.e. using an internal store of processing equipment to receive or retain programs
- G06F9/44—Arrangements for executing specific programs
- G06F9/455—Emulation; Interpretation; Software simulation, e.g. virtualisation or emulation of application or operating system execution engines
- G06F9/45533—Hypervisors; Virtual machine monitors
- G06F9/45558—Hypervisor-specific management and integration aspects
- G06F2009/4557—Distribution of virtual machine instances; Migration and load balancing
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06F—ELECTRIC DIGITAL DATA PROCESSING
- G06F9/00—Arrangements for program control, e.g. control units
- G06F9/06—Arrangements for program control, e.g. control units using stored programs, i.e. using an internal store of processing equipment to receive or retain programs
- G06F9/44—Arrangements for executing specific programs
- G06F9/455—Emulation; Interpretation; Software simulation, e.g. virtualisation or emulation of application or operating system execution engines
- G06F9/45533—Hypervisors; Virtual machine monitors
- G06F9/45558—Hypervisor-specific management and integration aspects
- G06F2009/45595—Network integration; Enabling network access in virtual machine instances
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06F—ELECTRIC DIGITAL DATA PROCESSING
- G06F2209/00—Indexing scheme relating to G06F9/00
- G06F2209/50—Indexing scheme relating to G06F9/50
- G06F2209/5021—Priority
Landscapes
- Engineering & Computer Science (AREA)
- Theoretical Computer Science (AREA)
- Physics & Mathematics (AREA)
- Computing Systems (AREA)
- Software Systems (AREA)
- General Engineering & Computer Science (AREA)
- General Physics & Mathematics (AREA)
- Mathematical Physics (AREA)
- Quality & Reliability (AREA)
- Management, Administration, Business Operations System, And Electronic Commerce (AREA)
Abstract
The invention belongs to the technical field of Internet of things platform construction, and discloses an Internet of things platform construction method based on a micro-service architecture, wherein the Internet of things platform construction system comprises: the system comprises a micro-service level division module, a kubernets cluster building module, a mirror image building module, an expansion and contraction module, a module adding module, a data collecting module, a monitoring scheme selecting module, a prediction model building module, a simulation module, a precision checking module, a prediction module and a scheduling module. According to the invention, the complex functions of the platform of the Internet of things can be decoupled, the application of the Internet of things is established on the micro-service set, and in actual deployment, multiple instances of the micro-service can enhance the flexibility and robustness of an application program; the scheduling algorithm in kubernets is improved, resource scheduling is achieved by distinguishing multiple tenants, scheduling decision and resource constraint of a single user are respected, and fairness of a default overall scheduling mechanism is improved.
Description
Technical Field
The invention belongs to the technical field of Internet of things platform construction, and particularly relates to a method for constructing an Internet of things platform based on a micro-service architecture.
Background
At present, most of existing internet of things platforms are single applications deployed in virtual machines, and with continuous development of the internet of things industry, functions of the internet of things platforms are gradually complicated, and redundancy backup and horizontal expansion need to consume more machine resources and time cost for maintenance. Meanwhile, for the application of the traditional single Internet of things, the coupling degree between the modules is high, the functions are operated in the same process in a centralized mode, when the environment of the increasingly complex Internet of things is faced, the whole cluster can only be stretched, and resources are greatly wasted.
The micro-service architecture is widely supported in recent years, and can decouple the application services, reduce the coupling between the services and improve the robustness of the system; docker is an open source application container engine, and supports the deployment of containers to local environments and various mainstream cloud platforms.
Task scheduling and resource allocation in a cloud infrastructure are well-known discussion problems, and the conventional open source architecture kubernets provides functions of scheduling containers and load balancing; the scheduling algorithm belongs to a non-exclusive type, all incoming task requests are processed through one scheduling component, the priority of tasks cannot be realized in the component, all tasks are processed by the principle of first come and first serve, meanwhile, the default scheduling algorithm does not support a preemptive strategy, and the scheduling algorithm cannot better process the task requests in the complex environment of the Internet of things; the scheduler realized by kubernets by default screens out improper nodes through a preselection process, then calculates scores of all nodes through a preference strategy, and after the scores are compared, the pod is operated on the node with the highest score, but when the scheduler schedules different pod copies under the same resource controller, preselection and preference operation can be carried out on the same node, so that repeated work can be carried out each time.
The kubernets can realize load balance of the service, corresponding capacity expansion and capacity reduction are carried out on the service according to preset rules, but the function of the automatic scaling service requires an application program provider to define a parameter set in a self-defining mode, the parameter set is unpredictable in determination and can only be obtained according to experience, and the situations of resource waste or resource shortage are likely to exist. These management parameters are again static and incoming requests do change often, in which case the scaling decision is passive in nature rather than proactive.
Through the above analysis, the problems and defects of the prior art are as follows:
(1) with the continuous development of the internet of things industry, the functions of the internet of things platform are gradually complicated, and the redundant backup and the horizontal expansion need to consume more machine resources and time cost for maintenance. Meanwhile, for the application of the traditional single Internet of things, the coupling degree between the modules is high, the functions are operated in the same process in a centralized mode, when the environment of the increasingly complex Internet of things is faced, the whole cluster can only be stretched, and resources are greatly wasted.
(2) The scheduling algorithm of the conventional open source architecture kubernets belongs to a non-exclusive type, the priority of tasks cannot be realized in a component, and all tasks are processed by a principle of first come and first serve; meanwhile, the default scheduling algorithm does not support a preemptive strategy, and the scheduling algorithm cannot better process task requests in the complex environment of the internet of things.
(3) When a scheduler of the conventional open source architecture kubernets schedules different pod copies under the same resource controller, preselection and optimization operations are performed on the same node, and repeated work is performed each time.
(4) The function of the automatic scaling service of the conventional open source architecture kubernets requires an application program provider to define a parameter set, the determination of the parameter set is unpredictable and can only be obtained according to experience, and the situations of resource waste or resource insufficiency are likely to exist; these management parameters are again static and incoming requests do change often, in which case the scaling decision is passive in nature rather than proactive.
The difficulty in solving the above problems and defects is: compared with the traditional architecture, the micro service architecture reduces the granularity of the service, shortens the development period, reduces the inherent complexity of the large service and improves the scalability by constructing the service with the independent life cycle. However, the cost is expensive performance overhead and complex dynamic resource usage, and an application program usually deployed in the cloud needs to meet certain performance requirements, such as response time, but at the same time, the cost of the cloud resources is also reduced as much as possible. How to reasonably allocate resources and balance load of micro-service programs through a container scheduling system kubernets becomes an industry difficulty.
The significance of solving the problems and the defects is as follows: with the increasing severity of micro-service architecture, resource scheduling gradually becomes a key technology in a cloud platform. And (4) reasonable resource scheduling is carried out. And sufficient resource guarantee can be provided for the application program, and the response time of the service is reduced, so that the service quality is improved. For the micro service management platform kubernets, the resource scheduling mechanism has very important significance as well, and is an indispensable important component in cluster management. The invention provides a resource scheduling strategy based on load prediction on the basis of the original scheduling strategy, the load situation applied at a certain time in the future is predicted according to the running situation of a monitoring program, and then the resource is scheduled in advance according to the prediction result, so that the response time of service is reduced.
Disclosure of Invention
Aiming at the problems in the prior art, the invention provides a method for constructing an Internet of things platform based on a micro-service architecture, and particularly relates to a technology for constructing a networking platform on a Kubernets architecture.
The invention is realized in such a way that an Internet of things platform construction method based on a micro-service architecture comprises the following steps:
dividing a micro-service level for an Internet of things platform; the micro-service level comprises a data access layer, a data processing layer, a service layer and an application layer, and is used for decoupling the functions of the platform, reducing the coupling and improving the expansibility;
step two, building a kubernets cluster, and realizing high availability of the cluster, wherein three masters and six slaves are built;
step three, constructing all split projects of the platform into a Docker mirror image file, so that the service can be deployed more quickly;
step four, the deployment module of the micro-service uses a container cluster management tool kubernets, and correspondingly deploys the mirror image file manufactured in the third step on the kubernets cluster, so that the split micro-service architecture Internet of things platform can be managed more conveniently;
step five, when the load balance of the cluster service is realized, intelligent capacity expansion and capacity reduction are realized through monitoring states, grading stages, plan making and execution operations;
adding a monitoring module, a prediction module and a resource dynamic scheduling module on the kubernets architecture, and automatically expanding and reducing the capacity;
collecting the resource use conditions of all containers on the node through a monitoring module; wherein the monitoring module is a monitoring container operating at each node;
step eight, selecting Prometheus + Grafana as a monitoring scheme to more accurately monitor the running state of the application program;
step nine, using historical resource use condition data provided by the monitoring module to establish a gray prediction model, predicting the resource use condition in a period of time in the future, and scheduling the resource in advance, thereby reducing the response time of the service;
step ten, using the cAdvisor to obtain the utilization rate of the CPU and the utilization rate of the memory about the previous instant value of the node, and using a GM (1, 1) prediction algorithm to simulate;
step eleven, carrying out precision inspection on the prediction data in the step seven, so that the prediction result is more accurate;
step twelve, by collecting historical resource use condition data of the application program running on the kubernets platform, predicting the resource use condition in a future period of time by using a grey prediction model, and then calculating the scaling time and predicting the workload in an analysis stage;
step thirteen, the realization of the cluster preemptive scheduling can be carried out on the tasks with high priority;
fourteen, taking the CPU utilization rate into consideration through a scheduling algorithm, and judging a plurality of indexes including the memory utilization rate and the applied network state;
and step fifteen, performing resource specific scheduling based on multiple tenants.
Further, in step eight, the selecting Prometheus + Grafana as the monitoring scheme includes:
(1) integrating Prometous onto the deployed kubernets;
(2) optimizing Prometheus deployment to realize hot loading of configuration information;
(3) and configuring a Prometeus information collecting rule to realize information collection of the running containers in the kubernets cluster.
Further, in step ten, the obtaining, by using cAdvisor, values of the CPU utilization rate and the memory utilization rate about a previous moment of the node, and performing simulation by using a GM (1, 1) prediction algorithm includes:
assuming that the CPU utilization of a node is Uc and the memory utilization is Um, the CPU and Um are obtained from cAdvisor at the previous moment of the node, where Uc ═ { Uc (1), Uc (2), …, Uc (n)) }, Um ═ Um { Um (1), Um (2), …, Um (n)) }, and then GM (1, 1) prediction algorithm is used to simulate it, and then the values of Uc and Um at time n +1 are predicted, the calculation steps are as follows:
(1) assume a time series, x(0)={x(0)(1),x(0)(2),…,x(0)(N), the number of original values is N, and then one is generated by one accumulationNew sequences, i.e. x(1)={x(1)(1),x(1)(2),…,x(1)(N), the summary can be:
according to the grey prediction method, the corresponding whitening differential equation of the GM (1, 1) model can be obtained:
where alpha is called constant, mu is called developed gray number, and the gray number for endogenous control is a constant input to the system, and this equation satisfies the initial conditions,
when t is equal to t0Time x(1)=x(1)(t0) (3)
the approach to the gray model is to accumulate the sequence once (1) to estimate the constants α and μ by the least squares method.
(2) Because of x(1)(1) Left as the initial value, so that x(1)(2),x(1)x(3),…,x(1)(N) is substituted into equation (2) to substitute the differential, and Δ t ═ 1 (t +1) to t ═ 1 (t +1) are obtained by sampling at equal intervals, instead of the differentialIs like thatThen, the formula (2) hasWill ax(1)(i) The term moves to the right and is written as the product of the vector quantities:
due to the fact thatInvolving an accumulation column x(1)Of two time instants, thus x(1)(i) It is more reasonable to take the average substitution of the two moments before and after, namely x(1)(i) Is replaced byWriting equation (5) as a matrix expression:
y is (x)(0)(2),x(0)(3),…,x(0)(N))T。
when k is 1, 2, …, N-1, the result is obtained from equation (8)Is the fitted value; when k is more than or equal to N,for predicting values, this is relative to a once-accumulated sequence x(1)The fitting value of (a) is reduced by a post-subtraction operation, and when k is 1, 2, …, N-1, the original sequence x is obtained(0)Fitting value ofWhen k is more than or equal to N, the original sequence x can be obtained(0)And (6) forecasting values.
Further, in step eleven, the precision checking of the prediction data includes:
(1) residual error test, respectively calculating:
(2) and (3) posterior difference inspection: respectively calculating:
(3) and constructing a prediction precision grade comparison table.
Further, in the step (3), in the prediction accuracy grade comparison table, when P is greater than 0.95 and C is less than 0.35, the prediction accuracy grade is good; when P is greater than 0.80 and C is less than 0.45, the prediction precision grade is qualified; p >0.70 and C <0.50, the prediction accuracy level is marginal; when P is less than or equal to 0.70 and C is more than or equal to 0.65, the prediction precision grade is unqualified.
Further, in step thirteen, the implementation of the cluster preemptive scheduling includes:
firstly, the Pod is divided into a high priority and a low priority, and each priority is added with a sub-optimal priority which can be defined by a user; in the scheduling process, the kubernets cluster can schedule the Pod with high priority in advance, and meanwhile, when cluster resources cannot support the operation condition of the container, the cluster resources can also support the function that the Pod with high priority preempts the Pod with low priority.
Further, in step fifteen, the resource-specific scheduling based on the multiple tenants includes:
(1) before starting, the first step is to audit deleted, stopped or abnormally crashed instances;
(2) internal circulation: selecting a user in the priority ordering, selecting a suspended task in the user's queue, and then determining whether any node can host the task; if no node can bear the task, deleting the user from the list, and internally circulating to continue to the next user in the list;
(3) external circulation: if a match is found between the resource requirements of the task and the available resources of the node, then the user is removed from the inner loop; and then recalculating the scheduling priorities of all the users again to generate a new list, and then performing another round of scheduling through internal circulation.
Another object of the present invention is to provide an internet of things platform construction system using the internet of things platform construction method based on a micro service architecture, the internet of things platform construction system including:
the micro-service level dividing module is used for dividing micro-service levels for the Internet of things platform; the micro service level comprises a data access layer, a data processing layer, a service layer and an application layer;
the kubernets cluster building module is used for building a kubernets cluster, three masters and six slaves, and high availability of the cluster is achieved;
the mirror image construction module is used for constructing all split projects of the platform into mirror images and carrying out corresponding deployment on the cluster, and the deployment module of the micro-service uses a container cluster management tool kubernets;
the capacity expansion and reduction module is used for realizing intelligent capacity expansion and reduction through monitoring states, grading stages, plan making and execution operations when the load balance of the cluster service is realized;
the module adding module is used for adding a monitoring module, a prediction module, a resource dynamic scheduling module and an automatic capacity expansion and reduction module on a kubernets architecture;
the data collection module is used for collecting the resource use conditions of all containers on the node through the monitoring module; the monitoring module runs a monitoring container at each node;
the monitoring scheme selection module is used for selecting Prometheus + Grafana as a monitoring scheme;
the prediction model establishing module is used for establishing a gray prediction model by utilizing the historical resource use condition data provided by the monitoring module and predicting the resource use condition in a period of time in the future;
the simulation module is used for acquiring the utilization rate of the CPU and the utilization rate of the memory about the previous instant value of the node through the cAdvisor and simulating by using a GM (1, 1) prediction algorithm;
the precision inspection module is used for carrying out precision inspection on the predicted data;
the prediction module is used for predicting the resource use condition in a future period of time by collecting historical resource use condition data of an application program running on the kubernets platform and using a grey prediction model, and then calculating the scaling time and the workload prediction in an analysis stage;
the scheduling module is used for realizing cluster preemptive scheduling; the CPU utilization rate is considered through a scheduling algorithm, and multiple indexes including the memory utilization rate and the applied network state are judged at the same time; and meanwhile, the resources are specifically scheduled based on multiple tenants.
Another object of the present invention is to provide a computer program product stored on a computer readable medium, which includes a computer readable program for providing a user input interface to implement the method for constructing the internet of things platform based on the micro service architecture when the computer program product is executed on an electronic device.
Another object of the present invention is to provide a computer-readable storage medium, which stores instructions that, when executed on a computer, cause the computer to execute the method for constructing a platform of an internet of things based on a microservice architecture.
By combining all the technical schemes, the invention has the advantages and positive effects that: according to the Internet of things platform construction method based on the micro-service architecture, the application program of the existing Internet of things platform is divided into a plurality of small services for reconstruction, the communication between the services adopts an REST method, and an interface calling method of a single platform is reserved to the maximum extent. Meanwhile, the container technology is utilized to deploy the split micro-services in a Docker, a traditional virtual machine virtualizes a set of complete hardware, a complete operating system is operated on the hardware, and then the required application program is operated on the system; and the application program of the container directly runs the kernel of the host, and the container does not have the kernel of the container and does not perform hardware virtualization, so that the container is easier to transplant and has high efficiency.
According to the invention, the complex functions of the platform of the Internet of things can be decoupled, the application of the Internet of things is established on the micro-service set, and in actual deployment, multiple instances of the micro-service can enhance the flexibility and robustness of an application program. The invention also provides a method for realizing resource scheduling by distinguishing multiple tenants, and a scheduler of the kubernets is not suitable for a multi-tenant environment for a group of different tasks and different resource requirements, so that two-stage scheduling is provided, another scheduling layer is integrated on the whole scheduler of the kubernets, the scheduling decision and resource constraint of a single user can be respected, and the fairness of a default whole scheduling mechanism is improved; meanwhile, a kubernets expansion engine is provided, in the engine, the running condition of a container in a kubernets cluster collected by Prometous is utilized, a gray prediction model is used, and prediction analysis is carried out on collected data, so that load balancing in kubernets is faster and more accurate, the dispatching efficiency of a kubernets scheduler is improved, the cluster load balancing capacity is enhanced, and the requirement under the complex and changeable internet of things environment is met more quickly and accurately.
Drawings
In order to more clearly illustrate the technical solutions of the embodiments of the present invention, the drawings needed to be used in the embodiments of the present invention will be briefly described below, and it is obvious that the drawings described below are only some embodiments of the present invention, and it is obvious for those skilled in the art that other drawings can be obtained according to the drawings without creative efforts.
Fig. 1 is a flowchart of a method for constructing an internet of things platform based on a micro-service architecture according to an embodiment of the present invention.
Fig. 2 is a structural block diagram of a platform construction system of the internet of things provided by the embodiment of the invention;
in the figure: 1. a microservice level partitioning module; 2. a kubernets cluster building module; 3. a mirror image building module; 4. a capacity expansion and reduction module; 5. a module adding module; 6. a data collection module; 7. a monitoring scheme selection module; 8. a prediction model building module; 9. a simulation module; 10. a precision inspection module; 11. a prediction module; 12. and a scheduling module.
Fig. 3 is a layered architecture diagram of an internet of things platform according to an embodiment of the present invention.
Fig. 4 is a micro-service orchestration architecture diagram (kubernets architecture diagram) provided by an embodiment of the present invention.
Fig. 5 is a diagram of the overall architecture of scheduling provided by an embodiment of the present invention.
Detailed Description
In order to make the objects, technical solutions and advantages of the present invention more apparent, the present invention is further described in detail with reference to the following embodiments. It should be understood that the specific embodiments described herein are merely illustrative of the invention and are not intended to limit the invention.
Aiming at the problems in the prior art, the invention provides a method for constructing an internet of things platform based on a micro-service architecture, and the invention is described in detail below with reference to the accompanying drawings.
As shown in fig. 1, the method for constructing an internet of things platform based on a micro-service architecture provided by the embodiment of the invention includes the following steps:
s101, dividing a micro-service level for an Internet of things platform; the micro service level comprises a data access layer, a data processing layer, a service layer and an application layer;
s102, building a kubernets cluster, and realizing high availability of the cluster, wherein three masters and six slaves are built;
s103, constructing all split projects of the platform into mirror images, and carrying out corresponding deployment on the cluster;
s104, the deployment module of the micro-service uses a container cluster management tool kubernets;
s105, when the load balance of the cluster service is realized, intelligent capacity expansion and capacity reduction are realized through monitoring states, grading stages, plan making and execution operations;
s106, adding a monitoring module, a prediction module and a resource dynamic scheduling module on the kubernets architecture, and automatically expanding and reducing the capacity;
s107, collecting the resource use conditions of all containers on the node through a monitoring module; the monitoring module runs a monitoring container at each node;
s108, selecting Prometheus + Grafana as a monitoring scheme;
s109, establishing a gray prediction model by using the historical resource use condition data provided by the monitoring module, and predicting the resource use condition in a period of time in the future;
s110, using the cAdvisor to obtain the utilization rate of the CPU and the utilization rate of the memory about the previous instant value of the node, and using a GM (1, 1) prediction algorithm to simulate;
s111, carrying out precision inspection on the prediction data in the S107;
s112, forecasting the resource use condition in a future period of time by collecting historical resource use condition data of an application program running on the kubernets platform by using a grey forecasting model, and then calculating the scaling time and forecasting the workload in an analysis stage;
s113, realizing cluster preemptive scheduling;
s114, the CPU utilization rate is considered through a scheduling algorithm, and multiple indexes including the memory utilization rate and the applied network state are judged at the same time;
and S115, specifically scheduling resources based on multiple tenants.
A person skilled in the art can also implement the method for constructing the platform of the internet of things based on the micro service architecture according to the present invention by using other steps, and the method for constructing the platform of the internet of things based on the micro service architecture according to the present invention shown in fig. 1 is only a specific embodiment.
As shown in fig. 2, the internet of things platform construction system provided by the embodiment of the present invention includes:
the micro-service level dividing module 1 is used for dividing micro-service levels for the platform of the Internet of things; the micro service level comprises a data access layer, a data processing layer, a service layer and an application layer;
the kubernets cluster building module 2 is used for building a kubernets cluster, three masters and six slaves, and high availability of the cluster is achieved;
a mirror image construction module 3, configured to construct mirror images of all split projects of the platform, and perform corresponding deployment on the cluster, where the deployment module of the micro-service uses a container cluster management tool kubernets;
the capacity expansion and reduction module 4 is used for realizing intelligent capacity expansion and reduction through monitoring states, grading stages, plan making and execution operations when the load balance of the cluster service is realized;
the module adding module 5 is used for adding a monitoring module, a prediction module and a resource dynamic scheduling module on the kubernets architecture and automatically expanding and reducing the capacity;
the data collection module 6 is used for collecting the resource use conditions of all containers on the node through the monitoring module; the monitoring module runs a monitoring container at each node;
a monitoring scheme selection module 7, configured to select Prometheus + Grafana as a monitoring scheme;
the prediction model establishing module 8 is used for establishing a gray prediction model by utilizing the historical resource use condition data provided by the monitoring module, and predicting the resource use condition in a future period of time;
the simulation module 9 is used for acquiring the values of the utilization rate of the CPU and the utilization rate of the memory about the previous moment of the node through the cAdvisor and performing simulation by using a GM (1, 1) prediction algorithm;
the precision inspection module 10 is used for carrying out precision inspection on the prediction data;
the prediction module 11 is used for predicting the resource usage in a future period of time by collecting historical resource usage data of an application program running on the kubernets platform and using a grey prediction model, and then calculating the scaling time and the workload prediction in an analysis stage;
the scheduling module 12 is configured to implement cluster preemptive scheduling; the CPU utilization rate is considered through a scheduling algorithm, and multiple indexes including the memory utilization rate and the applied network state are judged at the same time; and meanwhile, the resources are specifically scheduled based on multiple tenants.
The technical solution of the present invention is further described with reference to the following examples.
1. According to the invention, the application program of the existing Internet of things platform is divided into a plurality of small services for reconstruction, the REST method is adopted for communication among the services, and the interface calling method of the single platform is reserved to the maximum extent. Meanwhile, the container technology is utilized to deploy the split micro-services in a Docker, a traditional virtual machine virtualizes a set of complete hardware, a complete operating system is operated on the hardware, and then the required application program is operated on the system; and the application program of the container directly runs the kernel of the host, and the container does not have the kernel of the container and does not perform hardware virtualization, so that the container is easier to transplant and has high efficiency.
According to the invention, the complex functions of the platform of the Internet of things can be decoupled, the application of the Internet of things is established on the micro-service set, and in actual deployment, multiple instances of the micro-service can enhance the flexibility and robustness of an application program. The invention also provides a method for realizing resource scheduling by distinguishing multiple tenants, and a scheduler of the kubernets is not suitable for a multi-tenant environment for a group of different tasks and different resource requirements, so that two-stage scheduling is provided, another scheduling layer is integrated on the whole scheduler of the kubernets, the scheduling decision and resource constraint of a single user can be respected, and the fairness of a default whole scheduling mechanism is improved; meanwhile, a kubernets expansion engine is provided, in the engine, the running condition of a container in a kubernets cluster collected by Prometous is utilized, a gray prediction model is used, and prediction analysis is carried out on collected data, so that load balancing in kubernets is faster and more accurate, the dispatching efficiency of a kubernets scheduler is improved, the cluster load balancing capacity is enhanced, and the requirement under the complex and changeable internet of things environment is met more quickly and accurately.
2. Aiming at the problems in the prior art, the invention provides a method for constructing an Internet of things platform based on a micro-service architecture. The invention is realized in such a way that the construction method of the Internet of things platform based on the micro-service architecture comprises the following steps:
(1) and dividing a micro service level into a data access layer, a data processing layer, a service layer and an application layer for the platform of the Internet of things.
(2) Building a kubernetes cluster, three masters and six slaves to realize the high availability of the cluster;
(3) constructing all split projects of the platform into a mirror image, and carrying out corresponding deployment on the cluster;
(4) the deployment module of the micro-service uses a container cluster management tool kubernets;
(5) when realizing the load balance of the cluster service, in order to be able to more intelligent expansion and contraction, the process that needs to be realized probably has: monitoring state, grading stage, planning and executing operation;
(6) a monitoring module, a prediction module and a resource dynamic scheduling module are added on a kubernets architecture, and an automatic capacity expansion and reduction module is added;
(7) the monitoring module runs a monitoring container on each node to collect the resource use condition of all containers on the node;
(8) the monitoring scheme selects Prometheus + Grafana, and the specific implementation steps are as follows:
the method comprises the following steps: prometous is integrated on the well-deployed kubernets, and the high availability of Prometous can be guaranteed due to the characteristics of kubernets
Step two: optimizing Prometheus deployment to realize hot loading of configuration information
Step three: configuring a Prometeus information collecting rule to realize information collection of containers in operation in the kubernets cluster;
(9) at each worker node, the usage of the resource changes over time. Therefore, the relation between the change trend of the monitored node and the time is researched, and a prediction model is built by utilizing the historical resource use condition data provided by the monitoring module so as to predict the resource use condition in a future period of time. The gray prediction model is chosen for use in this patent because it can be adapted for small data modeling without regard to its internal factors.
(10) Since the original scheduling policy of kubernets only calculates the resources of CPU and memory, therefore, assuming that the CPU utilization of a node is Uc and the memory utilization is Um, cAdvisor is used to obtain the previous moment of Uc and Um about the node, where Uc ═ { Uc (1), Uc (2), …, Uc (n)) }, Um ═ Um { Um (1), Um (2), …, Um (n)) }, and then GM (1, 1) prediction algorithm is used to simulate it, and then the values of Uc and Um at time n +1 are predicted, the main calculation steps are as follows:
the method comprises the following steps: assume a time series, x(0)={x(0)(1),x(0)(2),…,x(0)(N), the number of original values is N, and then a new sequence, x, is generated by a single accumulation(1)={x(1)(1),x(1)(2),…,x(1)(N), the summary can be:
according to the grey prediction method, the corresponding whitening differential equation of the GM (1, 1) model can be obtained:
where alpha is called constant, mu is called developed gray number, and the gray number for endogenous control is a constant input to the system, and this equation satisfies the initial conditions,
when t is equal to t0Time x(1)=x(1)(t0) (3)
the approach to the gray model is to accumulate the sequence once (1) to estimate the constants α and μ by the least squares method.
Step two: because of x(1)(1) Left as the initial value, so that x(1)(2),x(1)(3),…,x(1)(N) is substituted into equation (2) to substitute the differential, and Δ t ═ 1 (t +1) to t ═ 1 (t +1) are obtained by sampling at equal intervals, instead of the differentialIs like thatThen, the formula (2) hasWill ax(1)(i) The term moves to the right and is written as the product of the vector quantities:
due to the fact thatInvolving an accumulation column x(1)Of two time instants, thus x(1)(i) It is more reasonable to take the average substitution of the two moments before and after, namely x(1)(i) Is replaced byWrite (5) as a matrix expression:
y is (x)(0)(2),x(0)(3),…,x(0)(N))T。
step four: estimate the value Substituting the equation (4) to obtain a corresponding time equation:
when k is 1, 2, …, N-1, the result is obtained from equation (8)Is the fitted value; when k is more than or equal to N,for predicting values, this is relative to a once-accumulated sequence x(1)The fitting value of (a) is reduced by a post-subtraction operation, and when k is 1, 2, …, N-1, the original sequence x is obtained(0)Fitting value ofWhen k is more than or equal to N, the original sequence x can be obtained(0)And (6) forecasting values.
(11) And (5) carrying out precision test on the prediction data in the step (7), wherein the implementation mode is as follows:
the method comprises the following steps: residual error test, respectively calculating:
step two: and (3) posterior difference inspection: respectively calculating:
step three: prediction accuracy grade comparison table (see Table 1)
TABLE 1 prediction accuracy grade LUT
(12) By collecting historical resource use condition data of an application program running on a kubernets platform, predicting the resource use condition in a future period of time by using a grey prediction model, and then calculating the scaling time and the workload prediction in an analysis stage;
(13) and (3) realizing cluster preemptive scheduling: firstly, the Pod is divided into a high priority and a low priority, and each priority is added with a sub-optimal priority which can be defined by a user; in the scheduling process, the kubernets cluster can schedule the Pod with high priority in advance, and meanwhile, when cluster resources cannot support the operation condition of the container, the cluster resources can also support the function that the Pod with high priority preempts the Pod with low priority.
(14) In the scheduling algorithm, not only the CPU utilization rate is considered, but also a plurality of indexes such as the memory utilization rate, the applied network state and the like are judged, so that a better basis is provided for the scheduling of tasks.
(15) The specific scheduling of resources is performed based on multiple tenants, and the whole scheduling process is as follows:
1. before starting, the first step is to audit deleted, stopped or abnormally crashed instances;
2. internal circulation: selecting a user in the priority ordering, selecting a suspended task in the user's queue, and then determining whether any node can host the task; if no node can bear the task, deleting the user from the list, and internally circulating to continue to the next user in the list;
3. external circulation: if a match is found between the resource requirements of the task and the available resources of the node, the user is removed from the inner loop. Then, the scheduling priorities of all users are recalculated again, a new list is generated, and then another round of scheduling is performed through an internal loop.
In the above embodiments, the implementation may be wholly or partially realized by software, hardware, firmware, or any combination thereof. When used in whole or in part, can be implemented in a computer program product that includes one or more computer instructions. When loaded or executed on a computer, cause the flow or functions according to embodiments of the invention to occur, in whole or in part. The computer may be a general purpose computer, a special purpose computer, a network of computers, or other programmable device. The computer instructions may be stored in a computer readable storage medium or transmitted from one computer readable storage medium to another, for example, the computer instructions may be transmitted from one website site, computer, server, or data center to another website site, computer, server, or data center via wire (e.g., coaxial cable, fiber optic, Digital Subscriber Line (DSL), or wireless (e.g., infrared, wireless, microwave, etc.)). The computer-readable storage medium can be any available medium that can be accessed by a computer or a data storage device, such as a server, a data center, etc., that includes one or more of the available media. The usable medium may be a magnetic medium (e.g., floppy Disk, hard Disk, magnetic tape), an optical medium (e.g., DVD), or a semiconductor medium (e.g., Solid State Disk (SSD)), among others.
The above description is only for the purpose of illustrating the present invention and the appended claims are not to be construed as limiting the scope of the invention, which is intended to cover all modifications, equivalents and improvements that are within the spirit and scope of the invention as defined by the appended claims.
Claims (10)
1. A method for constructing an Internet of things platform based on a micro-service architecture is characterized by comprising the following steps:
dividing micro service levels for the platform of the Internet of things; the micro service level comprises a data access layer, a data processing layer, a service layer and an application layer;
building a kubernetes cluster, three masters and six slaves to realize the high availability of the cluster;
constructing all split projects of the platform into a mirror image, and carrying out corresponding deployment on the cluster;
the deployment module of the micro-service uses a container cluster management tool kubernets;
when the load balance of the cluster service is realized, intelligent capacity expansion and capacity reduction are realized through monitoring states, grading stages, planning and executing operations;
a monitoring module, a prediction module and a resource dynamic scheduling module are added on a kubernets architecture, and an automatic capacity expansion and reduction module is added;
collecting the resource use conditions of all containers on the node through a monitoring module; the monitoring module runs a monitoring container at each node;
selecting Prometheus + Grafana as a monitoring scheme;
establishing a gray prediction model by using historical resource use condition data provided by a monitoring module, and predicting the resource use condition in a future period of time;
using the cAdvisor to obtain the utilization rate of the CPU and the utilization rate of the memory about the previous instant value of the node, and using a GM (1, 1) prediction algorithm to simulate;
carrying out precision inspection on the predicted data;
by collecting historical resource use condition data of an application program running on a kubernets platform, predicting the resource use condition in a future period of time by using a grey prediction model, and then calculating the scaling time and the workload prediction in an analysis stage;
realizing cluster preemptive scheduling;
the CPU utilization rate is considered through a scheduling algorithm, and multiple indexes including the memory utilization rate and the applied network state are judged at the same time;
and performing resource specific scheduling based on multiple tenants.
2. The Internet of things platform construction method based on the micro-service architecture as claimed in claim 1, wherein the selecting Prometheus + Grafana as a monitoring scheme comprises:
(1) integrating Prometous onto the deployed kubernets;
(2) optimizing Prometheus deployment to realize hot loading of configuration information;
(3) and configuring a Prometeus information collecting rule to realize information collection of the running containers in the kubernets cluster.
3. The method for constructing the platform of the internet of things based on the micro-service architecture according to claim 1, wherein the obtaining of the values of the utilization rate of the CPU and the utilization rate of the memory with respect to the previous moment of the node by using the cAdvisor and the simulation by using a GM (1, 1) prediction algorithm comprise:
assuming that the CPU utilization of a node is Uc and the memory utilization is Um, cAdvisor is used to obtain Uc and Um about the previous moment of the node, where Uc ═ Uc (1), Uc (2),.., Uc (n) }, Um ═ Um (1), Um (2),. once, Um (n) }, and then GM (1, 1) prediction algorithm is used to simulate it, and then the values of Uc and Um at time n +1 are predicted, the calculation steps are as follows:
(1) assume a time series, x(0)={x(0)(1),x(0)(2),…,x(0)(N), the number of original values is N, and then a new sequence, x, is generated by a single accumulation(1)={x(1)(1),x(1)(2),…,x(1)(N), the summary can be:
according to the grey prediction method, the corresponding whitening differential equation of the GM (1, 1) model can be obtained:
where alpha is called constant, mu is called developed gray number, and the gray number for endogenous control is a constant input to the system, and this equation satisfies the initial conditions,
when t is equal to t0Time x(1)=x(1)(t0) (3)
the approach to the gray model is to accumulate the sequence once (1) to estimate the constants α and μ by the least squares method;
(2) because of x(1)(1) Left as the initial value, so that x(1)(2),x(1)(3),...,x(1)(N) is substituted into equation (2) to substitute the differential, and Δ t ═ 1 (t +1) to t ═ 1 (t +1) are obtained by sampling at equal intervals, instead of the differentialIs like thatThen, the formula (2) hasWill ax(1)(i) The term moves to the right and is written as the product of the vector quantities:
due to the fact thatInvolving an accumulation column x(1)Of two time instants, thus x(1)(i) It is more reasonable to take the average substitution of the two moments before and after, namely x(1)(i) Is replaced byWriting equation (5) as a matrix expression:
y is (x)(0)(2),x(0)(3),…,x(0)(N))T;
when k is 1, 2, …, N-1, the result is obtained from equation (8)Is the fitted value; when k is more than or equal to N,for predicting values, this is relative to a once-accumulated sequence x(1)The fitting value of (a) is reduced by a post-subtraction operation, and when k is 1, 2, …, N-1, the original sequence x is obtained(0)Fitting value ofWhen k is more than or equal to N, the original sequence x can be obtained(0)And (6) forecasting values.
4. The method for constructing the platform of the internet of things based on the micro-service architecture as claimed in claim 1, wherein the performing precision test on the prediction data comprises:
(1) residual error test, respectively calculating:
(2) and (3) posterior difference inspection: respectively calculating:
(3) and constructing a prediction precision grade comparison table.
5. The method for constructing the platform of the internet of things based on the micro-service architecture as claimed in claim 4, wherein in the step (3), when P is greater than 0.95 and C is less than 0.35 in the prediction precision level comparison table, the prediction precision level is good; when P is more than 0.80 and C is less than 0.45, the prediction precision grade is qualified; when P is more than 0.70 and C is less than 0.50, the prediction precision level is marginal; when P is less than or equal to 0.70 and C is more than or equal to 0.65, the prediction precision grade is unqualified.
6. The method for constructing the platform of the internet of things based on the micro-service architecture as claimed in claim 1, wherein the implementation of the cluster preemptive scheduling comprises: firstly, the Pod is divided into a high priority and a low priority, and each priority is added with a sub-optimal priority which can be defined by a user; in the scheduling process, the kubeemets cluster can schedule the Pod with high priority in advance, and meanwhile, when cluster resources cannot support the condition that the container runs, the function that the Pod with high priority preempts the Pod with low priority can be supported.
7. The method for constructing the platform of the internet of things based on the micro-service architecture as claimed in claim 1, wherein the resource-specific scheduling based on the multi-tenant comprises:
(1) before starting, the first step is to audit deleted, stopped or abnormally crashed instances;
(2) internal circulation: selecting a user in the priority ordering, selecting a suspended task in the user's queue, and then determining whether any node can host the task; if no node can bear the task, deleting the user from the list, and internally circulating to continue to the next user in the list;
(3) external circulation: if a match is found between the resource requirements of the task and the available resources of the node, then the user is removed from the inner loop; and then recalculating the scheduling priorities of all the users again to generate a new list, and then performing another round of scheduling through internal circulation.
8. An internet of things platform construction system applying the internet of things platform construction method based on the micro-service architecture according to any one of claims 1 to 7, wherein the internet of things platform construction system comprises:
the micro-service level dividing module is used for dividing micro-service levels for the Internet of things platform; the micro service level comprises a data access layer, a data processing layer, a service layer and an application layer;
the kubernets cluster building module is used for building a kubernets cluster, three masters and six slaves, and high availability of the cluster is achieved;
the mirror image construction module is used for constructing all split projects of the platform into mirror images and carrying out corresponding deployment on the cluster, and the deployment module of the micro-service uses a container cluster management tool kubernets;
the capacity expansion and reduction module is used for realizing intelligent capacity expansion and reduction through monitoring states, grading stages, plan making and execution operations when the load balance of the cluster service is realized;
the module adding module is used for adding a monitoring module, a prediction module, a resource dynamic scheduling module and an automatic capacity expansion and reduction module on a kubernets architecture;
the data collection module is used for collecting the resource use conditions of all containers on the node through the monitoring module; the monitoring module runs a monitoring container at each node;
the monitoring scheme selection module is used for selecting Prometheus + Grafana as a monitoring scheme;
the prediction model establishing module is used for establishing a gray prediction model by utilizing the historical resource use condition data provided by the monitoring module and predicting the resource use condition in a period of time in the future;
the simulation module is used for acquiring the utilization rate of the CPU and the utilization rate of the memory about the previous instant value of the node through the cAdvisor and simulating by using a GM (1, 1) prediction algorithm;
the precision inspection module is used for carrying out precision inspection on the predicted data;
the prediction module is used for predicting the resource use condition in a future period of time by collecting historical resource use condition data of an application program running on the kubernets platform and using a grey prediction model, and then calculating the scaling time and the workload prediction in an analysis stage;
the scheduling module is used for realizing cluster preemptive scheduling; the CPU utilization rate is considered through a scheduling algorithm, and multiple indexes including the memory utilization rate and the applied network state are judged at the same time; and meanwhile, the resources are specifically scheduled based on multiple tenants.
9. A computer program product stored on a computer readable medium, comprising a computer readable program for providing a user input interface to implement the method for constructing a platform of an internet of things based on a microservice architecture according to any one of claims 1 to 7 when the computer program product is executed on an electronic device.
10. A computer-readable storage medium storing instructions which, when executed on a computer, cause the computer to execute the method for constructing a platform of internet of things based on a micro-service architecture according to any one of claims 1 to 7.
Priority Applications (1)
Application Number | Priority Date | Filing Date | Title |
---|---|---|---|
CN202110229566.1A CN113110914A (en) | 2021-03-02 | 2021-03-02 | Internet of things platform construction method based on micro-service architecture |
Applications Claiming Priority (1)
Application Number | Priority Date | Filing Date | Title |
---|---|---|---|
CN202110229566.1A CN113110914A (en) | 2021-03-02 | 2021-03-02 | Internet of things platform construction method based on micro-service architecture |
Publications (1)
Publication Number | Publication Date |
---|---|
CN113110914A true CN113110914A (en) | 2021-07-13 |
Family
ID=76709642
Family Applications (1)
Application Number | Title | Priority Date | Filing Date |
---|---|---|---|
CN202110229566.1A Pending CN113110914A (en) | 2021-03-02 | 2021-03-02 | Internet of things platform construction method based on micro-service architecture |
Country Status (1)
Country | Link |
---|---|
CN (1) | CN113110914A (en) |
Cited By (8)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
CN113791863A (en) * | 2021-08-10 | 2021-12-14 | 北京中电飞华通信有限公司 | Virtual container-based power internet of things agent resource scheduling method and related equipment |
CN114048021A (en) * | 2021-09-30 | 2022-02-15 | 河北嘉朗科技有限公司 | Internet of things multilayer multi-rule hybrid computing power automatic distribution technology |
CN114500400A (en) * | 2022-01-04 | 2022-05-13 | 西安电子科技大学 | Large-scale network real-time simulation method based on container technology |
CN115237570A (en) * | 2022-07-29 | 2022-10-25 | 陈魏炜 | Strategy generation method based on cloud computing and cloud platform |
WO2024007849A1 (en) * | 2023-04-26 | 2024-01-11 | 之江实验室 | Distributed training container scheduling for intelligent computing |
CN117453493A (en) * | 2023-12-22 | 2024-01-26 | 山东爱特云翔信息技术有限公司 | GPU computing power cluster monitoring method and system for large-scale multi-data center |
CN117791613A (en) * | 2024-02-27 | 2024-03-29 | 浙电(宁波北仑)智慧能源有限公司 | Decision method and system based on resource cluster regulation and control |
CN117453493B (en) * | 2023-12-22 | 2024-05-31 | 山东爱特云翔信息技术有限公司 | GPU computing power cluster monitoring method and system for large-scale multi-data center |
Citations (4)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
CN108829494A (en) * | 2018-06-25 | 2018-11-16 | 杭州谐云科技有限公司 | Container cloud platform intelligence method for optimizing resources based on load estimation |
CN110149396A (en) * | 2019-05-20 | 2019-08-20 | 华南理工大学 | A kind of platform of internet of things construction method based on micro services framework |
US20200019444A1 (en) * | 2018-07-11 | 2020-01-16 | International Business Machines Corporation | Cluster load balancing based on assessment of future loading |
CN112199150A (en) * | 2020-08-13 | 2021-01-08 | 北京航空航天大学 | Online application dynamic capacity expansion and contraction method based on micro-service calling dependency perception |
-
2021
- 2021-03-02 CN CN202110229566.1A patent/CN113110914A/en active Pending
Patent Citations (4)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
CN108829494A (en) * | 2018-06-25 | 2018-11-16 | 杭州谐云科技有限公司 | Container cloud platform intelligence method for optimizing resources based on load estimation |
US20200019444A1 (en) * | 2018-07-11 | 2020-01-16 | International Business Machines Corporation | Cluster load balancing based on assessment of future loading |
CN110149396A (en) * | 2019-05-20 | 2019-08-20 | 华南理工大学 | A kind of platform of internet of things construction method based on micro services framework |
CN112199150A (en) * | 2020-08-13 | 2021-01-08 | 北京航空航天大学 | Online application dynamic capacity expansion and contraction method based on micro-service calling dependency perception |
Non-Patent Citations (6)
Title |
---|
CHIA-CHEN CHANG: "A Kubernetes-Based Monitoring Platform for Dynamic Cloud Resource Provisioning", 《2017 IEEE GLOBAL COMMUNICATIONS CONFERENCE》, 15 January 2018 (2018-01-15) * |
沈玉龙: "无线异构网络中的切换预测算法", 《通信学报》 * |
沈玉龙: "无线异构网络中的切换预测算法", 《通信学报》, 31 October 2009 (2009-10-31) * |
王天泽: "基于灰色模型的云资源动态伸缩功能研究", 《软件导刊》 * |
王天泽: "基于灰色模型的云资源动态伸缩功能研究", 《软件导刊》, no. 04, 15 April 2018 (2018-04-15) * |
青鸟英谷教育科技股份有限公司: "《云计算框架与应用》", 西安电子科技大学出版社, pages: 163 - 165 * |
Cited By (11)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
CN113791863A (en) * | 2021-08-10 | 2021-12-14 | 北京中电飞华通信有限公司 | Virtual container-based power internet of things agent resource scheduling method and related equipment |
CN113791863B (en) * | 2021-08-10 | 2024-01-23 | 北京中电飞华通信有限公司 | Virtual container-based power Internet of things proxy resource scheduling method and related equipment |
CN114048021A (en) * | 2021-09-30 | 2022-02-15 | 河北嘉朗科技有限公司 | Internet of things multilayer multi-rule hybrid computing power automatic distribution technology |
CN114500400A (en) * | 2022-01-04 | 2022-05-13 | 西安电子科技大学 | Large-scale network real-time simulation method based on container technology |
CN114500400B (en) * | 2022-01-04 | 2023-09-08 | 西安电子科技大学 | Large-scale network real-time simulation method based on container technology |
CN115237570A (en) * | 2022-07-29 | 2022-10-25 | 陈魏炜 | Strategy generation method based on cloud computing and cloud platform |
CN115237570B (en) * | 2022-07-29 | 2023-06-16 | 上海佑瞻智能科技有限公司 | Policy generation method based on cloud computing and cloud platform |
WO2024007849A1 (en) * | 2023-04-26 | 2024-01-11 | 之江实验室 | Distributed training container scheduling for intelligent computing |
CN117453493A (en) * | 2023-12-22 | 2024-01-26 | 山东爱特云翔信息技术有限公司 | GPU computing power cluster monitoring method and system for large-scale multi-data center |
CN117453493B (en) * | 2023-12-22 | 2024-05-31 | 山东爱特云翔信息技术有限公司 | GPU computing power cluster monitoring method and system for large-scale multi-data center |
CN117791613A (en) * | 2024-02-27 | 2024-03-29 | 浙电(宁波北仑)智慧能源有限公司 | Decision method and system based on resource cluster regulation and control |
Similar Documents
Publication | Publication Date | Title |
---|---|---|
CN113110914A (en) | Internet of things platform construction method based on micro-service architecture | |
US11656911B2 (en) | Systems, methods, and apparatuses for implementing a scheduler with preemptive termination of existing workloads to free resources for high priority items | |
Zhu et al. | Task scheduling for multi-cloud computing subject to security and reliability constraints | |
CN108829494B (en) | Container cloud platform intelligent resource optimization method based on load prediction | |
US10514951B2 (en) | Systems, methods, and apparatuses for implementing a stateless, deterministic scheduler and work discovery system with interruption recovery | |
US11294726B2 (en) | Systems, methods, and apparatuses for implementing a scalable scheduler with heterogeneous resource allocation of large competing workloads types using QoS | |
CN113806018B (en) | Kubernetes cluster resource mixed scheduling method based on neural network and distributed cache | |
US11579933B2 (en) | Method for establishing system resource prediction and resource management model through multi-layer correlations | |
CN112685153A (en) | Micro-service scheduling method and device and electronic equipment | |
CN115297112A (en) | Dynamic resource quota and scheduling component based on Kubernetes | |
CN115220916B (en) | Automatic calculation scheduling method, device and system of video intelligent analysis platform | |
CN113391913A (en) | Distributed scheduling method and device based on prediction | |
JP5515889B2 (en) | Virtual machine system, automatic migration method and automatic migration program | |
CN115543626A (en) | Power defect image simulation method adopting heterogeneous computing resource load balancing scheduling | |
CN109614210B (en) | Storm big data energy-saving scheduling method based on energy consumption perception | |
Lu et al. | InSTechAH: Cost-effectively autoscaling smart computing hadoop cluster in private cloud | |
CN111367632B (en) | Container cloud scheduling method based on periodic characteristics | |
CN112130927A (en) | Reliability-enhanced mobile edge computing task unloading method | |
CN111124619A (en) | Container scheduling method for secondary scheduling | |
Li et al. | On scheduling of high-throughput scientific workflows under budget constraints in multi-cloud environments | |
CN115562841A (en) | Cloud video service self-adaptive resource scheduling system and method | |
Yakubu et al. | Priority based delay time scheduling for quality of service in cloud computing networks | |
CN115269140A (en) | Container-based cloud computing workflow scheduling method, system and equipment | |
CN115061811A (en) | Resource scheduling method, device, equipment and storage medium | |
Du et al. | A combined priority scheduling method for distributed machine learning |
Legal Events
Date | Code | Title | Description |
---|---|---|---|
PB01 | Publication | ||
PB01 | Publication | ||
SE01 | Entry into force of request for substantive examination | ||
SE01 | Entry into force of request for substantive examination | ||
RJ01 | Rejection of invention patent application after publication |
Application publication date: 20210713 |
|
RJ01 | Rejection of invention patent application after publication |