CN108920153B - Docker container dynamic scheduling method based on load prediction - Google Patents
Docker container dynamic scheduling method based on load prediction Download PDFInfo
- Publication number
- CN108920153B CN108920153B CN201810535697.0A CN201810535697A CN108920153B CN 108920153 B CN108920153 B CN 108920153B CN 201810535697 A CN201810535697 A CN 201810535697A CN 108920153 B CN108920153 B CN 108920153B
- Authority
- CN
- China
- Prior art keywords
- docker container
- load
- docker
- container
- load 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.)
- Active
Links
Images
Classifications
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06F—ELECTRIC DIGITAL DATA PROCESSING
- G06F8/00—Arrangements for software engineering
- G06F8/60—Software deployment
-
- 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
- G06F9/505—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 load
Landscapes
- Engineering & Computer Science (AREA)
- Software Systems (AREA)
- Theoretical Computer Science (AREA)
- General Engineering & Computer Science (AREA)
- Physics & Mathematics (AREA)
- General Physics & Mathematics (AREA)
- Debugging And Monitoring (AREA)
Abstract
The invention provides a Docker container dynamic scheduling method based on load prediction. Firstly, a Docker container load monitoring module acquires load data of a Docker container; then storing the load data in a time sequence database; then, a Docker container load prediction module acquires and processes load data from a time sequence database, analyzes the load data by combining an ARIMA model, and generates a load prediction value; and finally, the Docker container scheduling module automatically expands the number of Docker containers in the Docker container cluster according to the load predicted value. The invention applies the load prediction technology to the Docker container scheduling field, realizes the automatic expansion of the Docker container cluster according to the real-time load condition, effectively solves the problem that the Docker container cluster cannot automatically adjust the resources due to the load condition and has insufficient load capacity in the operation stage, and improves the availability of the Docker container cluster.
Description
Technical Field
The invention belongs to the technical field of dynamic Docker container scheduling, and particularly provides a dynamic Docker container scheduling method based on load prediction.
Background
The micro-service architecture divides an application into a plurality of independent services with service attributes, each service runs in an independent process, and the services cooperate with each other through a lightweight communication mechanism, so that service value is provided for a terminal user. The deployment mode of the application based on the micro-service architecture includes a manual deployment mode, a script deployment mode and the like based on a cloud platform, and the mirror deployment mode based on a Docker container is a mainstream mode at present.
As an open source application container engine, a Docker container enables developers to package applications and their dependent packages into a portable Docker container and then publish the Docker container on a Linux machine in a Docker container environment. The Docker container realizes the isolation of resources by virtue of the virtualization of an operating system layer, and shares an operating system with a host machine, so that the resource utilization rate can be greatly improved, and the performance in the aspects of I/O (input/output) and the like can be improved.
In order to maintain the availability of applications deployed based on Docker containers, building a highly available Docker container cluster based on redundant backup technology and extending the Docker container cluster by horizontal extension technology is a common solution. A Docker container cluster constructed based on the existing Docker container arranging and deploying technology cannot automatically adjust resources according to the real-time Docker container load condition, and therefore the load capacity of the Docker container cluster in the operation stage is insufficient. The process of deploying the Docker container includes a series of operations such as downloading the mirror image, deploying the mirror image, and starting the Docker container. When finding that the current Docker container resource cannot meet the load requirement, the resource application is started, and the process of deploying a new Docker container is very time-consuming due to operations such as downloading mirror images, so that the availability of the application cannot be guaranteed in the time.
Therefore, the invention provides a dynamic Docker container scheduling method based on load prediction, and the dynamic Docker container scheduling method has high execution efficiency, reduces human intervention and can efficiently maintain the availability of application.
Disclosure of Invention
The invention aims to apply a load prediction technology to dynamic Docker container scheduling, solve the problem that the current Docker container cluster cannot automatically adjust resources due to load conditions and has insufficient load capacity in the operation stage, and provide a load prediction-based dynamic Docker container scheduling method for users of the Docker container cluster.
The purpose of the invention is realized by at least one of the following technical solutions.
A Docker container dynamic scheduling method based on load prediction comprises the following steps:
firstly, collecting load data of a Docker container; then storing the load data in a time sequence database; then, load data is obtained and processed from a time sequence database, and the load data is analyzed by combining an ARIMA (Autoregressive Integrated Moving Average model) to generate a load predicted value; and finally, expanding the number of Docker containers in the Docker container cluster according to the load predicted value.
Furthermore, the load data collection is completed by a Docker container load monitoring module, the Docker container load monitoring module collects consumption conditions of the Docker container on four types of resources, namely a CPU, a memory, a disk I/O and a network, as the load data of the Docker container by inquiring a pseudo file according to a timing polling strategy, and then the load data is associated with information of the server node and the Docker container and stored in a time sequence database.
Further, load data of the Docker container is obtained from the time sequence database through the Docker container load prediction module, the load data is abstracted by combining an ARIMA model, an online Docker container load prediction model is built, and then the load data of the Docker container is analyzed through the model to generate a load prediction value.
Further, a load predicted value generated by the Docker container load prediction module is obtained through the Docker container scheduling module, and when the load predicted value exceeds a highest load threshold of a Docker container designated by a Docker container cluster user, the container is automatically expanded according to a step length designated by the Docker container cluster user until the load predicted value regenerated by the Docker container load prediction module is lower than the highest load threshold of the Docker container.
Furthermore, the load prediction is applied to the field of dynamic dispatching of the Docker container to realize automatic expansion of the Docker container cluster according to real-time load conditions, so that the problem that the Docker container cluster cannot automatically adjust resources due to the load conditions and has insufficient load capacity in the operation stage is solved, and the availability of the Docker container cluster is improved.
Compared with the prior art, the invention has the following advantages and technical effects:
according to the invention, a load prediction technology is applied to the field of dynamic dispatching of Docker containers, Docker container cluster users can realize automatic expansion of Docker container clusters according to real-time load conditions through simple configuration, the problem that the Docker container clusters constructed based on the existing Docker container arrangement and deployment technology cannot automatically adjust resources due to the load conditions and have insufficient load capacity in an operation stage is solved, and the availability of the Docker container clusters is improved.
Drawings
FIG. 1 is a general architecture diagram of a Docker container dynamic scheduling method based on load prediction;
FIG. 2 is a flowchart of the Docker container load prediction module work;
FIG. 3 is a flowchart of the Docker container scheduling module's work flow.
Detailed Description
Embodiments of the present invention will be described in detail below with reference to the accompanying drawings and examples, but the invention is not limited thereto, and it should be noted that those skilled in the art can implement the embodiments without specific details below.
The example provides a dynamic Docker container scheduling method based on load prediction, and the overall architecture is shown in FIG. 1, and specifically includes four processes of load data acquisition, load data storage, load data analysis and dynamic Docker container scheduling.
Load data collection
The Docker container cluster user firstly deploys a Docker container load monitoring module at each server node, specifies a time interval for the Docker container load monitoring module to collect load data, and then starts the Docker container load monitoring module.
The Docker container load monitoring module collects the consumption conditions of the Docker container in four types of resources, namely a CPU, an internal memory, a disk I/O and a network, and the consumption conditions are used as load data of the Docker container. In the CPU resource consumption situation of the Docker container, the CPU indexes are divided into two categories: one type of the system embodies the use condition of the Docker container in CPU resources, including user CPU consumption, system CPU consumption, each CPU consumption and total CPU consumption; one type embodies the saturation degree of the use of the Docker container in the CPU resource, including the CPU limit execution number and the total time of the limited CPU utilization rate. The Docker container load monitoring module collects the use condition of the Docker container in the CPU resource consumption aspect.
The core technologies of the Docker container are the control group technology and the namespace technology. The control group is composed of a plurality of control group subsystems, wherein the CPU access subsystem is used for counting the CPU usage of each control group, and comprises the following three interfaces:
stat: the CPU time consumed by the control group in the user mode and the kernel mode, respectively, is reported.
Use: the total CPU time consumed by the control group is reported.
Use _ percpu: the CPU time consumed by the control group on each CPU of the host is reported, and the sum is the value of CPU access.
The control group interface is realized based on a pseudo file system, so that the Docker container load acquisition module acquires the real-time CPU load condition in a file operation mode through the control group subsystem of the CPU act.
And the Docker container load monitoring module acquires the load data of the Docker container by inquiring the control group pseudo file according to a time interval specified by a Docker container cluster user by a timed polling strategy.
Load data storage
The load data collected by the Docker container load monitoring module each time is a new entry to be recorded, which is related to the timestamp, and the load data needs to be stored so that the Docker container load prediction module can analyze the load data. And the Docker container load monitoring module stores load data by adopting an InfluxDB time sequence database. The data stored by the InfluxDB is logically composed of a measurement, a tag, a field and a time stamp.
The Measurement can be compared with a data table in a relational database, and the Docker container load monitoring module establishes the Measurement respectively according to load monitoring indexes of user CPU consumption, system CPU consumption, total CPU consumption, memory resource consumption, disk I/O resource consumption and network resource consumption, and also represents the load data type of the Docker container.
The Tag group is composed of a group of key-value pair data structures and represents a series of attribute information corresponding to each record in the database. In order to more quickly acquire load data of a required Docker container so as to perform data analysis and load prediction, a Docker container load monitoring module takes a Docker container name and a server node name where the Docker container is located as a tag group of the load data, and the Docker container name not only serves as a unique identifier of the Docker container, but also contains information of application services deployed in the Docker container.
The file group is also composed of a key-value pair data structure and represents the concrete content of each record in the database. The Docker container load monitoring module stores the specific load value of the Docker container in the Field group.
The time stamp is used for recording the generation time of the load data of the Docker container, namely the time when the Docker container load monitoring module acquires the piece of load data.
Load data analysis
The load data of the Docker container has a time sequence which represents that the load data is a time sequence in nature. The analysis and prediction of the load data is actually the analysis and prediction of a set of time series data. ARIMA is a commonly used time sequence prediction model at present, and introduces differential transformation on ARMA (autoregressive moving average model) to convert an original unstable time sequence into a stable time sequence so as to perform data analysis by using the ARMA model. The container load data is correlated with time, which indicates that the container load data is actually a set of time series data, while past container load data can provide reference for the value range of the load prediction data in the short term in the future of the container, so that the container load data can be subjected to data analysis and calculated to obtain a reasonable prediction value through an ARIMA (automatic data mining and mapping) time series prediction model.
The flow of the Docker container load prediction module analyzing the load data and generating a load prediction is shown in FIG. 2. The Docker container load prediction module firstly receives basic data of a server node name, a Docker container name, a load data type needing to be predicted, an analyzed time period and a predicted time period which are specified by a Docker container cluster user, then obtains load data of a Docker container from an InfluxDB time sequence database by taking the basic data as query conditions, and generates a time sequence.
And then, the Docker container load prediction module adopts a unit root test method to test whether the load data has stationarity or not on the time sequence, and if the load data does not have stationarity, the Docker container load prediction module carries out differential operation on the load data until the stationarity condition is met.
And then, judging the order p and the order q of the ARIMA model by using a Bayesian information criterion, and constructing a load prediction model of the Docker container by combining the differential order. And then, the Docker container load prediction module applies a load prediction model according to a prediction time period specified by a Docker container cluster user to generate a load prediction value of the Docker container in the time period.
Finally, if the load data of the Docker container is subjected to differential transformation before the load prediction model is constructed, the load prediction value needs to be reduced, and the finally required load prediction value of the Docker container can be obtained.
Dynamic scheduling of containers
The process of the Docker container scheduling module implementing dynamic scheduling on the Docker container according to the load prediction value of the Docker container is shown in fig. 3. The Docker container scheduling module firstly receives a highest load threshold and an expansion step size of a Docker container specified by a Docker container cluster user. The highest load threshold of the Docker container is a decision criterion on whether or not dynamic scheduling is performed on the Docker container. When the load predicted value of the Docker container exceeds the highest load threshold of the Docker container, the Docker container scheduling module determines that the load of the Docker container exceeds the limit, and the Docker container cannot support the existing access request, and the access request to the service must be shared by other Docker containers. Otherwise, the Docker container scheduling module determines that the Docker container still has sufficient machine resources to maintain a normal availability state for service.
The Docker container scheduling module sets the expansion step length to 1 to 3, that is, the Docker container scheduling module can only increase 1 to 3 Docker containers each time a Docker container cluster increases a copy of a Docker container for a service. The capacity expansion step length is set to prevent a Docker container scheduling module from increasing too many Docker containers at one time when the Docker containers are horizontally expanded, so that one service has too many Docker container copies, and machine resources are wasted.
When the Docker container scheduling module determines that the Docker container needs to be horizontally expanded, a server node with a small load in the Docker container cluster is selected, and a Docker container deployment request is distributed to the node. And after the newly-added Docker container is started successfully, the Docker container scheduling module requests the Docker container load prediction module to analyze the real-time load data of the Docker container again and returns a new load prediction value. Then, the Docker container scheduling module compares the load predicted value with the highest load threshold of the Docker container specified by the Docker container cluster user again, and judges whether the Docker container needs to be expanded continuously or not.
Claims (5)
1. A Docker container dynamic scheduling method based on load prediction is characterized by comprising the following steps: firstly, collecting load data of a Docker container; then storing the load data in a time sequence database; acquiring and processing load data from a time sequence database, and analyzing the load data by combining an ARIMA model to generate a load predicted value; finally, expanding the number of Docker containers in the Docker container cluster according to the load predicted value; the method comprises the following steps that a Docker container cluster user firstly deploys a Docker container load monitoring module at each server node, specifies a time interval for the Docker container load monitoring module to acquire load data, and then starts the Docker container load monitoring module;
the Docker container scheduling module realizes a dynamic scheduling process for the Docker container according to the load prediction value of the Docker container; the method comprises the steps that a Docker container scheduling module firstly receives a highest load threshold and an expansion step length of a Docker container specified by a Docker container cluster user; the highest load threshold of the Docker container is a judgment standard for judging whether the Docker container is dynamically scheduled or not; when the load predicted value of the Docker container exceeds the highest load threshold of the Docker container, the Docker container scheduling module determines that the load of the Docker container exceeds the limit and cannot support the existing access request, and the access request to the service must be shared by other Docker containers; otherwise, the Docker container scheduling module determines that the Docker container still has sufficient machine resources to maintain the normal available state of the service;
the Docker container scheduling module sets the expansion step length to be 1 to 3, namely the Docker container scheduling module can only increase 1 to 3 Docker containers each time when the Docker container cluster increases Docker container copies for services; the capacity expansion step length is set to prevent a Docker container scheduling module from increasing too many Docker containers at one time when the Docker containers are horizontally expanded, so that one service has too many copies of the Docker containers, and machine resources are wasted;
when the Docker container scheduling module determines that the Docker container needs to be horizontally expanded, selecting a server node with a small load in a Docker container cluster, and distributing a Docker container deployment request to the node; after the newly-added Docker container is started successfully, the Docker container scheduling module requests the Docker container load prediction module to analyze the real-time load data of the Docker container again and returns a new load prediction value; then, the Docker container scheduling module compares the load predicted value with the highest load threshold of the Docker container specified by the Docker container cluster user again, and judges whether the Docker container needs to be expanded continuously or not.
2. The method for dynamically scheduling the Docker container based on the load prediction as claimed in claim 1, wherein the load data collection is completed by a Docker container load monitoring module, the Docker container load monitoring module collects consumption conditions of the Docker container on four types of resources, namely a CPU, a memory, a disk I/O and a network, as the load data of the Docker container by querying a dummy file according to a timing polling strategy, and then associates the load data with information of a server node and the Docker container and stores the load data in a time series database.
3. The method for dynamically scheduling the Docker container based on the load prediction as claimed in claim 1, wherein the load data of the Docker container is obtained from the time series database through a Docker container load prediction module, the load data is abstracted by combining with an ARIMA model, an online Docker container load prediction model is constructed, and then the load data of the Docker container is analyzed through the model to generate the load prediction value.
4. The method for dynamically scheduling the Docker container based on the load prediction as claimed in claim 1, wherein the load prediction value generated by the Docker container load prediction module is obtained by the Docker container scheduling module, and when the load prediction value exceeds the highest load threshold of the Docker container specified by the Docker container cluster user, the container is automatically expanded according to the step length specified by the Docker container cluster user until the load prediction value regenerated by the Docker container load prediction module is lower than the highest load threshold of the Docker container.
5. The Docker container dynamic scheduling method based on load prediction as claimed in any one of claims 1 to 4, is characterized in that the load prediction is applied to the Docker container dynamic scheduling field to realize automatic expansion of Docker container clusters according to real-time load conditions, solve the problem that the Docker container clusters cannot automatically adjust resources due to the load conditions, and have insufficient load capacity in the operation stage, and improve the availability of the Docker container clusters.
Priority Applications (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN201810535697.0A CN108920153B (en) | 2018-05-29 | 2018-05-29 | Docker container dynamic scheduling method based on load prediction |
Applications Claiming Priority (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN201810535697.0A CN108920153B (en) | 2018-05-29 | 2018-05-29 | Docker container dynamic scheduling method based on load prediction |
Publications (2)
| Publication Number | Publication Date |
|---|---|
| CN108920153A CN108920153A (en) | 2018-11-30 |
| CN108920153B true CN108920153B (en) | 2022-04-22 |
Family
ID=64411128
Family Applications (1)
| Application Number | Title | Priority Date | Filing Date |
|---|---|---|---|
| CN201810535697.0A Active CN108920153B (en) | 2018-05-29 | 2018-05-29 | Docker container dynamic scheduling method based on load prediction |
Country Status (1)
| Country | Link |
|---|---|
| CN (1) | CN108920153B (en) |
Families Citing this family (11)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN111327655A (en) * | 2018-12-14 | 2020-06-23 | 中移（杭州）信息技术有限公司 | Multi-tenant container resource quota prediction method and device and electronic equipment |
| CN111190790A (en) * | 2019-12-17 | 2020-05-22 | 西安交通大学 | Cloud computing cluster monitoring method and system based on peak prediction |
| CN111385142B (en) * | 2020-02-14 | 2022-04-12 | 重庆邮电大学 | Kubernetes-based adaptive web container stretching method |
| CN111427753B (en) * | 2020-03-23 | 2024-04-23 | 上海新炬网络信息技术股份有限公司 | Capacity prediction device based on ARIMA model and control method thereof |
| CN111625196A (en) * | 2020-05-26 | 2020-09-04 | 北京海益同展信息科技有限公司 | Block chain node capacity expansion method and device, computer equipment and storage medium |
| CN113971066A (en) * | 2020-07-22 | 2022-01-25 | 中国科学院深圳先进技术研究院 | Kubernetes cluster resource dynamic adjustment method and electronic equipment |
| CN111949374B (en) * | 2020-08-18 | 2023-11-24 | 广州市优普科技有限公司 | Intelligent processing platform for business data machine |
| CN112187894B (en) * | 2020-09-17 | 2022-06-10 | 杭州谐云科技有限公司 | Container dynamic scheduling method based on load correlation prediction |
| CN113672335A (en) * | 2021-07-08 | 2021-11-19 | 浙江一山智慧医疗研究有限公司 | Container scheduling method, device, electronic device and storage medium |
| US11977922B2 (en) * | 2021-09-30 | 2024-05-07 | International Business Machines Corporation | Resource reuse for pipeline workloads |
| CN117369981A (en) * | 2022-06-30 | 2024-01-09 | 中兴通讯股份有限公司 | Container adjusting method, device and storage medium based on monitor |
Citations (6)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN103425535A (en) * | 2013-06-05 | 2013-12-04 | 浙江大学 | Agile elastic telescoping method in cloud environment |
| CN106095533A (en) * | 2016-06-14 | 2016-11-09 | 中国联合网络通信集团有限公司 | Method of server expansion and device |
| CN106790595A (en) * | 2016-12-29 | 2017-05-31 | 上海理想信息产业（集团）有限公司 | A kind of Docker containers active load balancer and method |
| CN106961351A (en) * | 2017-03-03 | 2017-07-18 | 南京邮电大学 | Intelligent elastic telescopic method based on Docker container clusters |
| CN107045455A (en) * | 2017-06-19 | 2017-08-15 | 华中科技大学 | A kind of Docker Swarm cluster resource method for optimizing scheduling based on load estimation |
| CN107370816A (en) * | 2017-07-26 | 2017-11-21 | 郑州云海信息技术有限公司 | A kind of dispositions method and device of Web applications |
-
2018
- 2018-05-29 CN CN201810535697.0A patent/CN108920153B/en active Active
Patent Citations (6)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN103425535A (en) * | 2013-06-05 | 2013-12-04 | 浙江大学 | Agile elastic telescoping method in cloud environment |
| CN106095533A (en) * | 2016-06-14 | 2016-11-09 | 中国联合网络通信集团有限公司 | Method of server expansion and device |
| CN106790595A (en) * | 2016-12-29 | 2017-05-31 | 上海理想信息产业（集团）有限公司 | A kind of Docker containers active load balancer and method |
| CN106961351A (en) * | 2017-03-03 | 2017-07-18 | 南京邮电大学 | Intelligent elastic telescopic method based on Docker container clusters |
| CN107045455A (en) * | 2017-06-19 | 2017-08-15 | 华中科技大学 | A kind of Docker Swarm cluster resource method for optimizing scheduling based on load estimation |
| CN107370816A (en) * | 2017-07-26 | 2017-11-21 | 郑州云海信息技术有限公司 | A kind of dispositions method and device of Web applications |
Also Published As
| Publication number | Publication date |
|---|---|
| CN108920153A (en) | 2018-11-30 |
Similar Documents
| Publication | Publication Date | Title |
|---|---|---|
| CN108920153B (en) | Docker container dynamic scheduling method based on load prediction | |
| CN109150987B (en) | Two-layer container cluster elastic expansion method based on host layer and container layer | |
| CN111752965B (en) | Real-time database data interaction method and system based on micro-service | |
| Ju et al. | iGraph: an incremental data processing system for dynamic graph | |
| CN104050042B (en) | The resource allocation methods and device of ETL operations | |
| CN111355606B (en) | Web application-oriented container cluster self-adaptive expansion and contraction system and method | |
| CN106982356B (en) | Distributed large-scale video stream processing system | |
| CN103533058A (en) | HDFS (Hadoop distributed file system)/Hadoop storage cluster-oriented resource monitoring system and HDFS/Hadoop storage cluster-oriented resource monitoring method | |
| CN114064211A (en) | Video stream analysis system and method based on end-edge-cloud computing architecture | |
| Wang et al. | A cluster autoscaler based on multiple node types in kubernetes | |
| CN115587118A (en) | Task data dimension table association processing method and device and electronic equipment | |
| CN116450355A (en) | Multi-cluster model training method, device, equipment and medium | |
| CN114389955B (en) | Method for managing heterogeneous resource pool of embedded platform | |
| CN114666335A (en) | DDS-based distributed system load balancing device | |
| CN111984393A (en) | Distributed large-scale real-time data scheduling engine system and data scheduling method thereof | |
| CN111858656A (en) | Static data query method and device based on distributed architecture | |
| CN115858499A (en) | Database partition processing method and device, computer equipment and storage medium | |
| CN112667393B (en) | Method and device for building distributed task computing scheduling framework and computer equipment | |
| CN109005071A (en) | A kind of decision and deployment method and controlling equipment | |
| CN113886111A (en) | Workflow-based data analysis model calculation engine system and operation method | |
| CN114840343A (en) | Task scheduling method and system based on distributed system | |
| CN111506407B (en) | Resource management and job scheduling method and system combining Pull mode and Push mode | |
| CN109302723A (en) | A kind of multinode real-time radio pyroelectric monitor control system Internet-based and control method | |
| CN105677853A (en) | Data storage method and device based on big data technology framework | |
| CN110704159B (en) | Integrated cloud operating system based on OpenStack |
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 | ||
| GR01 | Patent grant | ||
| GR01 | Patent grant |