CN112860403A - Cluster load resource scheduling method, device, equipment, medium and product - Google Patents
Cluster load resource scheduling method, device, equipment, medium and product Download PDFInfo
- Publication number
- CN112860403A CN112860403A CN202110196241.8A CN202110196241A CN112860403A CN 112860403 A CN112860403 A CN 112860403A CN 202110196241 A CN202110196241 A CN 202110196241A CN 112860403 A CN112860403 A CN 112860403A
- Authority
- CN
- China
- Prior art keywords
- load
- load data
- scheduling
- cluster
- resources
- Prior art date
- Legal status (The legal status is an assumption and is not a legal conclusion. Google has not performed a legal analysis and makes no representation as to the accuracy of the status listed.)
- Granted
Links
- 238000000034 method Methods 0.000 title claims abstract description 64
- 238000012549 training Methods 0.000 claims abstract description 37
- 230000015654 memory Effects 0.000 claims description 13
- 230000006870 function Effects 0.000 claims description 9
- 238000004590 computer program Methods 0.000 claims description 8
- 230000008859 change Effects 0.000 description 6
- 238000010586 diagram Methods 0.000 description 6
- 238000004891 communication Methods 0.000 description 4
- 238000012544 monitoring process Methods 0.000 description 4
- 230000008569 process Effects 0.000 description 4
- 238000012545 processing Methods 0.000 description 4
- 238000013528 artificial neural network Methods 0.000 description 2
- 230000000694 effects Effects 0.000 description 2
- 230000009467 reduction Effects 0.000 description 2
- 230000006403 short-term memory Effects 0.000 description 2
- 230000006978 adaptation Effects 0.000 description 1
- 230000001413 cellular effect Effects 0.000 description 1
- 230000004069 differentiation Effects 0.000 description 1
- 230000003993 interaction Effects 0.000 description 1
- 238000012986 modification Methods 0.000 description 1
- 230000004048 modification Effects 0.000 description 1
- 238000003062 neural network model Methods 0.000 description 1
- 238000011160 research 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
- G06F9/46—Multiprogramming arrangements
- G06F9/48—Program initiating; Program switching, e.g. by interrupt
- G06F9/4806—Task transfer initiation or dispatching
- G06F9/4843—Task transfer initiation or dispatching by program, e.g. task dispatcher, supervisor, operating system
- G06F9/4881—Scheduling strategies for dispatcher, e.g. round robin, multi-level priority queues
-
- 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
-
- 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]
- G06F9/5083—Techniques for rebalancing the load in a distributed system
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06N—COMPUTING ARRANGEMENTS BASED ON SPECIFIC COMPUTATIONAL MODELS
- G06N3/00—Computing arrangements based on biological models
- G06N3/02—Neural networks
- G06N3/04—Architecture, e.g. interconnection topology
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06N—COMPUTING ARRANGEMENTS BASED ON SPECIFIC COMPUTATIONAL MODELS
- G06N3/00—Computing arrangements based on biological models
- G06N3/02—Neural networks
- G06N3/08—Learning methods
-
- Y—GENERAL TAGGING OF NEW TECHNOLOGICAL DEVELOPMENTS; GENERAL TAGGING OF CROSS-SECTIONAL TECHNOLOGIES SPANNING OVER SEVERAL SECTIONS OF THE IPC; TECHNICAL SUBJECTS COVERED BY FORMER USPC CROSS-REFERENCE ART COLLECTIONS [XRACs] AND DIGESTS
- Y02—TECHNOLOGIES OR APPLICATIONS FOR MITIGATION OR ADAPTATION AGAINST CLIMATE CHANGE
- Y02D—CLIMATE CHANGE MITIGATION TECHNOLOGIES IN INFORMATION AND COMMUNICATION TECHNOLOGIES [ICT], I.E. INFORMATION AND COMMUNICATION TECHNOLOGIES AIMING AT THE REDUCTION OF THEIR OWN ENERGY USE
- Y02D10/00—Energy efficient computing, e.g. low power processors, power management or thermal management
Landscapes
- Engineering & Computer Science (AREA)
- Theoretical Computer Science (AREA)
- Software Systems (AREA)
- Physics & Mathematics (AREA)
- General Engineering & Computer Science (AREA)
- General Physics & Mathematics (AREA)
- Artificial Intelligence (AREA)
- Life Sciences & Earth Sciences (AREA)
- Health & Medical Sciences (AREA)
- Biomedical Technology (AREA)
- Biophysics (AREA)
- Computational Linguistics (AREA)
- Data Mining & Analysis (AREA)
- Evolutionary Computation (AREA)
- General Health & Medical Sciences (AREA)
- Molecular Biology (AREA)
- Computing Systems (AREA)
- Mathematical Physics (AREA)
- Debugging And Monitoring (AREA)
Abstract
The embodiment of the invention provides a method, a device, equipment, a medium and a product for dispatching cluster load resources, wherein the method comprises the following steps: acquiring current load data corresponding to a target cluster in a current time period and historical load data corresponding to a plurality of continuous time periods before the current time period; inputting and training the current load data and the plurality of historical load data into a converged load data prediction model, and determining predicted load data of the next time period; and scheduling load resources according to the current load data and the predicted load data. According to the cluster load resource scheduling method, the load data and the current load data are predicted through the load data prediction model trained to be converged, and the resources are scheduled in advance, so that the load resources can meet the load requirements of the target cluster in the next time period, the load resources of the target cluster can meet the load requirements in real time, and the hysteresis of actually scheduling the load resources is reduced.
Description
Technical Field
The embodiment of the invention relates to the technical field of communication network data, in particular to a method, a device, equipment, a medium and a product for scheduling cluster load resources.
Background
Compared with a 4G network, the 5G network has an order of magnitude higher requirement on data traffic, and meanwhile, the application scene of the 5G network is expanded from the traditional connection between people to the interconnection between people and things and the interconnection between things and things. Three most typical scenarios of 5G: the enhanced mobile bandwidth provides higher speed representation, large-scale mass terminal access, high-reliability and low-delay connection for users. In order to meet the requirement of differentiation and improve the communication data processing efficiency, the 5G core network creates a network slice for each service, namely a micro-service architecture is adopted.
The current 5G core network, due to the popularization and use of the micro service architecture, increases the processing efficiency of communication data, and also gradually increases the load pressure of the 5G core network. At present, aiming at the problem of overlarge load pressure of a 5G core network, generally, when the cluster load pressure of the core network is monitored to be overlarge, load resources in the core network are scheduled to cope with a cluster load high-pressure state. However, the time from monitoring that the load pressure is too large to actually scheduling the load resource to relieve the load pressure is usually long, and there is a serious hysteresis.
Disclosure of Invention
The invention provides a method, a device, equipment, a medium and a product for dispatching cluster load resources, which are used for solving the problem of serious hysteresis of the current mode for dispatching the load resources by clusters in a core network.
A first aspect of the present invention provides a method for scheduling cluster load resources, including:
acquiring current load data corresponding to a target cluster in a current time period and historical load data corresponding to a plurality of continuous time periods before the current time period;
inputting and training the current load data and the plurality of historical load data into a converged load data prediction model, and determining predicted load data of the next time period;
and scheduling load resources according to the current load data and the predicted load data.
Further, the method as described above, the target cluster comprising a plurality of servers; the server comprises a plurality of containers;
the obtaining of the current load data corresponding to the target cluster in the current time period includes:
collecting current load data of each server in a target cluster and each container in the server in the current time period;
the scheduling load resources according to the current load data and the predicted load data comprises:
and calculating the load data variation between the predicted load data and the current load data, and scheduling load resources according to the load data variation.
Further, according to the method described above, the load resources include load resources corresponding to servers in the cluster and load resources corresponding to containers in the servers;
the scheduling of load resources according to the load data variation includes:
judging whether the load data variation is greater than or equal to a preset scheduling threshold value or not;
if the load data variation is larger than or equal to a preset scheduling threshold, controlling a cloud control server to schedule corresponding load resources according to the load data variation;
if the load data variation is smaller than a preset scheduling threshold, judging whether the load data variation is smaller than or equal to a preset release threshold;
and if the load data variation is smaller than or equal to a preset release threshold, controlling the cloud control server to release the load resources of the corresponding server and/or container in the target cluster according to the load data variation.
Further, the method as described above, the preset scheduling threshold includes a first threshold and a second threshold;
if the load data variation is greater than or equal to a preset scheduling threshold, controlling a cloud control server to schedule corresponding load resources according to the load data variation, including:
if the load data variation is larger than or equal to a first threshold value, controlling a cloud control server to schedule data interfaces of a plurality of servers corresponding to a plurality of clusters according to the load data variation so as to enable load resources of the plurality of servers of the plurality of clusters to be added into a target cluster;
redeploying a plurality of server load resources of a plurality of clusters and load resources of a target cluster through a resource deployment server to complete load resource configuration of the target cluster;
if the load data variation is smaller than the first threshold and larger than or equal to a second threshold, controlling a cloud control server to schedule a data interface of a plurality of servers corresponding to another cluster according to the load data variation so as to enable load resources of the plurality of servers of another cluster to be added into a target cluster;
and the load resources of the plurality of servers of the other cluster and the load resources of the target cluster are redeployed through the resource deployment server to complete the load resource configuration of the target cluster.
Further, in the method as described above, the preset scheduling threshold further includes a third threshold;
if the load data variation is greater than or equal to a preset scheduling threshold, controlling the cloud control server to schedule the corresponding load resource according to the load data variation comprises:
and if the load data variation is smaller than the second threshold and larger than or equal to a third threshold, controlling data interfaces of a plurality of idle containers in a scheduling target cluster of the cloud control server according to the load data variation, so that load resources of the idle containers are redeployed through the resource deployment server, and the load corresponds to the load data variation.
Further, before inputting the current load data and a plurality of historical load data into a preset load data prediction model, the method further includes:
obtaining a training sample, wherein the training sample is a training sample corresponding to a load data prediction model, and the training sample comprises: historical load data and future load data for a plurality of consecutive time periods of the target cluster;
inputting the training samples into a preset load data prediction model to train the load data prediction model;
judging whether the load data prediction model meets a convergence condition or not by adopting an error function;
and if the load data prediction model meets the convergence condition, determining the load data prediction model meeting the convergence condition as a load data prediction model trained to be converged.
A second aspect of the present invention provides a cluster load resource scheduling apparatus, including:
the acquisition module is used for acquiring current load data corresponding to a target cluster in a current time period and historical load data corresponding to a plurality of continuous time periods before the current time period;
the determining module is used for inputting and training the current load data and the plurality of historical load data into a converged load data prediction model and determining predicted load data of the next time period;
and the scheduling module is used for scheduling the load resources according to the current load data and the predicted load data.
Further, the apparatus as described above, the target cluster comprising a plurality of servers; the server comprises a plurality of containers;
the obtaining module is specifically configured to, when obtaining current load data corresponding to a target cluster in a current time period:
collecting current load data of each server in a target cluster and each container in the server in the current time period;
the scheduling module is specifically configured to:
and calculating the load data variation between the predicted load data and the current load data, and scheduling load resources according to the load data variation.
Further, the apparatus as described above, the load resource includes a load resource corresponding to a server in a cluster and a load resource corresponding to a container in a server;
the scheduling module is specifically configured to, when scheduling the load resource according to the load data variation:
judging whether the load data variation is greater than or equal to a preset scheduling threshold value or not; if the load data variation is larger than or equal to a preset scheduling threshold, controlling a cloud control server to schedule corresponding load resources according to the load data variation; if the load data variation is smaller than a preset scheduling threshold, judging whether the load data variation is smaller than or equal to a preset release threshold; and if the load data variation is smaller than or equal to a preset release threshold, controlling the cloud control server to release the load resources of the corresponding server and/or container in the target cluster according to the load data variation.
Further, the apparatus as described above, the preset scheduling threshold includes a first threshold and a second threshold;
if the load data variation is greater than or equal to a preset scheduling threshold, the scheduling module is specifically configured to, when controlling, according to the load data variation, the cloud control server to schedule the corresponding load resource:
if the load data variation is larger than or equal to a first threshold value, controlling a cloud control server to schedule data interfaces of a plurality of servers corresponding to a plurality of clusters according to the load data variation so as to enable load resources of the plurality of servers of the plurality of clusters to be added into a target cluster; redeploying a plurality of server load resources of a plurality of clusters and load resources of a target cluster through a resource deployment server to complete load resource configuration of the target cluster; if the load data variation is smaller than the first threshold and larger than or equal to a second threshold, controlling a cloud control server to schedule a data interface of a plurality of servers corresponding to another cluster according to the load data variation so as to enable load resources of the plurality of servers of another cluster to be added into a target cluster; and the load resources of the plurality of servers of the other cluster and the load resources of the target cluster are redeployed through the resource deployment server to complete the load resource configuration of the target cluster.
Further, in the apparatus as described above, the preset scheduling threshold further includes a third threshold;
if the load data variation is greater than or equal to a preset scheduling threshold, the scheduling module is specifically configured to, when controlling, according to the load data variation, the cloud control server to schedule the corresponding load resource:
and if the load data variation is smaller than the second threshold and larger than or equal to a third threshold, controlling data interfaces of a plurality of idle containers in a scheduling target cluster of the cloud control server according to the load data variation, so that load resources of the idle containers are redeployed through the resource deployment server, and the load corresponds to the load data variation.
Further, the apparatus as described above, further comprising:
a training module, configured to obtain a training sample, where the training sample is a training sample corresponding to a load data prediction model, and the training sample includes: historical load data and future load data for a plurality of consecutive time periods of the target cluster; inputting the training samples into a preset load data prediction model to train the load data prediction model; judging whether the load data prediction model meets a convergence condition or not by adopting an error function; and if the load data prediction model meets the convergence condition, determining the load data prediction model meeting the convergence condition as a load data prediction model trained to be converged.
A third aspect of embodiments of the present invention provides an electronic device, including: a memory, a processor, and a transceiver;
the processor, the memory and the transceiver are interconnected through a circuit;
the memory stores computer-executable instructions; the transceiver is used for outputting a scheduling control instruction to the cloud control server and receiving load data sent by the target cluster;
wherein the processor is configured to perform the method of cluster load resource scheduling according to any of the first aspect.
A fourth aspect of the present invention provides a computer-readable storage medium, where a computer executable instruction is stored, and when the computer executable instruction is executed by a processor, the computer executable instruction is configured to implement the cluster load resource scheduling method according to any one of the first aspect.
A fifth aspect of embodiments of the present invention provides a computer program product, which includes a computer program, and when the computer program is executed by a processor, the computer program implements the cluster load resource scheduling method according to any one of the first aspect.
The embodiment of the invention provides a method, a device, equipment, a medium and a product for dispatching cluster load resources, wherein the method comprises the following steps: acquiring current load data corresponding to a target cluster in a current time period and historical load data corresponding to a plurality of continuous time periods before the current time period; inputting and training the current load data and the plurality of historical load data into a converged load data prediction model, and determining predicted load data of the next time period; and scheduling load resources according to the current load data and the predicted load data. According to the cluster load resource scheduling method, the current load data corresponding to the target cluster in the current time period and the historical load data corresponding to the continuous time periods before the current time period are obtained, and the historical load data and the current load data are input and trained to the converged load data prediction model, so that the predicted load data of the next time period are determined. And finally, scheduling resources in advance according to the predicted load data and the current load data so that the load resources of the target cluster can meet the load requirements of the target cluster in the next time period, thereby reducing the hysteresis of actually scheduling the load resources.
Drawings
The accompanying drawings, which are incorporated in and constitute a part of this specification, illustrate embodiments consistent with the invention and together with the description, serve to explain the principles of the invention.
Fig. 1 is a scene diagram of a cluster load resource scheduling method that can implement the embodiment of the present invention;
fig. 2 is a schematic flowchart of a cluster load resource scheduling method according to a first embodiment of the present invention;
fig. 3 is a schematic flowchart of a cluster load resource scheduling method according to a second embodiment of the present invention;
fig. 4 is a schematic flowchart of a cluster load resource scheduling method according to a third embodiment of the present invention;
fig. 5 is a schematic structural diagram of a load data prediction model of a cluster load resource scheduling method according to a third embodiment of the present invention;
fig. 6 is a schematic structural diagram of a cluster load resource scheduling device according to a fourth embodiment of the present invention;
fig. 7 is a schematic structural diagram of an electronic device according to a fifth embodiment of the present invention.
With the above figures, certain embodiments of the invention have been illustrated and described in more detail below. The drawings and the description are not intended to limit the scope of the inventive concept in any way, but rather to illustrate it by those skilled in the art with reference to specific embodiments.
Detailed Description
Reference will now be made in detail to the exemplary embodiments, examples of which are illustrated in the accompanying drawings. When the following description refers to the accompanying drawings, like numbers in different drawings represent the same or similar elements unless otherwise indicated. The embodiments described in the following exemplary embodiments do not represent all embodiments consistent with the present invention. Rather, they are merely examples of apparatus and methods consistent with certain aspects of the invention, as detailed in the appended claims.
The technical solution of the present invention will be described in detail below with specific examples. The following several specific embodiments may be combined with each other, and details of the same or similar concepts or processes may not be repeated in some embodiments. Embodiments of the present invention will be described below with reference to the accompanying drawings.
For a clear understanding of the technical solutions of the present application, a detailed description of the prior art solutions is first provided. The current 5G core network adopts a micro-service architecture. The micro-service architecture is characterized in that a plurality of functions are converged into one device in the past, the device is modified to be distributed into a plurality of corresponding small devices according to the functions, and mutual relation exists among the small devices. The microservice architecture is more efficient than before due to the fact that data are processed through data interaction among a plurality of small devices, but at the same time, data traffic is larger, and therefore load pressure of each cluster of the 5G core network is overlarge. At present, aiming at the problem of overlarge load pressure of each cluster of a 5G core network, generally, when the cluster load pressure of the core network is monitored to be overlarge, load resources in the core network are scheduled to deal with a cluster load high-pressure state. In the current mode, when the cluster load pressure of the core network is monitored to be overlarge, the load resource in the core network is scheduled, and a long time is required from the monitoring of the load pressure to the actual completion of the scheduling of the load resource. Thus, there is a severe hysteresis in this way of resource scheduling.
Therefore, in order to solve the technical problems that the time from the monitoring of the load pressure to the actual scheduling of the load resource to the load pressure reduction is usually long and serious hysteresis exists in the prior art, the inventor finds that in the research, in order to solve the problems that the time from the monitoring of the load pressure to the actual scheduling of the load resource to the load pressure reduction is usually long and serious hysteresis exists at present, a load data prediction model can be constructed in advance, and the predicted load data of the next time period can be predicted through the load data prediction model. Firstly, current load data corresponding to a target cluster in a current time period and historical load data corresponding to a plurality of continuous time periods before the current time period are obtained, and the historical load data and the current load data are input and trained to a converged load data prediction model, so that predicted load data of the next time period are determined. And finally, scheduling resources in advance according to the predicted load data and the current load data so that the load resources can meet the load requirements of the target cluster in the next time period, thereby enabling the load resources of the target cluster to meet the load requirements in real time and reducing the hysteresis of actually scheduling the load resources.
The inventor proposes a technical scheme of the application based on the creative discovery.
An application scenario of the cluster load resource scheduling method provided in the embodiment of the present invention is described below. As shown in fig. 1, 1 is an electronic device, 2 is a target cluster, 3 is another cluster, 4 is a cloud control server, and 5 is a resource deployment server. The network architecture of the application scenario corresponding to the cluster load resource scheduling method provided by the embodiment of the invention comprises the following steps: the system comprises an electronic device 1, a target cluster 2, other clusters 3, a cloud control server 4 and a resource deployment server 5. The target cluster 1 is a cluster with excessive load pressure and needs to schedule load resources to join. The other clusters 3 are clusters that are less load-stressed and provide load resources required for scheduling. The electronic device 1 monitors the target cluster 2 in real time and obtains current load data corresponding to a current time period of the target cluster 2 and historical load data corresponding to a plurality of continuous time periods before the current time period. The electronic device 1 then trains the current load data and the plurality of historical load data inputs to the converged load data prediction model, determining predicted load data for the next time period. At this time, the electronic device 1 generates and transmits a scheduling instruction to the cloud control server 4 according to the predicted load data, so that the cloud control server 4 schedules load resources from other clusters 3 in advance. Finally, the scheduled load resources are deployed and added into the target cluster 2 through the resource deployment server 5, so that the load resources can meet the load requirements of the target cluster in the next time period, the load resources of the target cluster can meet the load requirements in real time, and the hysteresis of actually scheduling the load resources is reduced.
The embodiments of the present invention will be described with reference to the accompanying drawings.
Fig. 2 is a schematic flowchart of a method for scheduling cluster load resources according to a first embodiment of the present invention, and as shown in fig. 2, in this embodiment, an execution subject of the embodiment of the present invention is a cluster load resource scheduling device, and the cluster load resource scheduling device may be integrated in an electronic device. The method for scheduling cluster load resources provided by this embodiment includes the following steps:
step S101, obtaining current load data corresponding to a target cluster in a current time period and historical load data corresponding to a plurality of continuous time periods before the current time period.
In this embodiment, the time period may be set to 30 minutes, 1 hour, or other time periods actually required, which is not limited in this embodiment. For example, the current time period may be a time period from 1 pm to 2 pm, and the historical load data corresponding to a plurality of consecutive time periods before the current time period may be the historical load data corresponding to 12 consecutive time periods between 1 am and 1 pm. The plurality of continuous time periods in the present embodiment may be set according to actual requirements, and if the prediction accuracy is improved, the number of continuous time periods may be increased to 20, 30, and so on.
In this embodiment, the current load data refers to data currently loaded by the target cluster, and the load data may include various service data, communication data, and the like, and may have various types according to actual applications.
Step S102, inputting and training the current load data and a plurality of historical load data into a converged load data prediction model, and determining predicted load data of the next time period.
In this embodiment, the load data prediction model is a neural network model that can be trained and continuously learned. Preferably, the load data prediction model may adopt an LSTM (full name of Long Short-Term Memory, chinese is a Long Short-Term Memory network), and the neural network is a time-cycle neural network, which can improve the accuracy of predicting load data.
In this embodiment, the next time period corresponds to the current time period, and refers to the time period next to the current time period.
And step S103, scheduling load resources according to the current load data and the predicted load data.
In this embodiment, the load resource may be scheduled by the scheduling device according to the current load data and the predicted load data, or may be directly scheduled. The load resources may be load resources in the target cluster and load resources in other clusters, and during scheduling, corresponding scheduling may be performed according to a load pressure condition of the target cluster, so that a utilization rate of the load resources may be improved.
The embodiment of the invention provides a cluster load resource scheduling method, which comprises the following steps: and acquiring current load data corresponding to the target cluster in the current time period and historical load data corresponding to a plurality of continuous time periods before the current time period. And inputting and training the current load data and a plurality of historical load data into a converged load data prediction model, and determining predicted load data of the next time period. And scheduling the load resources according to the current load data and the predicted load data. According to the cluster load resource scheduling method, the current load data corresponding to the target cluster in the current time period and the historical load data corresponding to the continuous time periods before the current time period are obtained, and the historical load data and the current load data are input and trained to the converged load data prediction model, so that the predicted load data of the next time period are determined. And finally, scheduling the resources in advance according to the predicted load data so that the load resources can meet the load requirements of the target cluster in the next time period, thereby enabling the load resources of the target cluster to meet the load requirements in real time and reducing the hysteresis of actually scheduling the load resources.
Fig. 3 is a schematic flow chart of a cluster load resource scheduling method according to a second embodiment of the present invention, and as shown in fig. 3, the cluster load resource scheduling method according to this embodiment is further detailed in each step on the basis of the cluster load resource scheduling method according to the previous embodiment of the present invention. The method for scheduling cluster load resources provided by this embodiment includes the following steps.
It should be noted that the target cluster includes a plurality of servers. The server includes a plurality of containers. A container is a tiny unit of load-running load data belonging to a server.
Step S201, collecting current load data of each server and each container in the server in the target cluster in the current time period and historical load data corresponding to a plurality of continuous time periods before the current time period.
In this embodiment, the current load data and the historical load data of each server in the target cluster and each container in the server in the current time period are collected to monitor the load data change of each fine place in the target cluster, so as to provide a basis for performing corresponding load resource scheduling according to the fine load data change.
Step S202, inputting and training the current load data and a plurality of historical load data into a converged load data prediction model, and determining predicted load data of the next time period.
In this embodiment, the implementation manner of step 202 is similar to that of step 102 in the previous embodiment of the present invention, and is not described herein again.
Step S203, calculating the load data variation between the predicted load data and the current load data, and scheduling the load resource according to the load data variation.
In this embodiment, the load data variation is load data variation between the predicted load data and the current load data, and if the load data variation is small, the load data variation may be handled by scheduling load resources in a vacant container in the target cluster. If the load data variation is large, the load resources of the servers in the target cluster can be scheduled, or the load resources of the servers in other clusters can be scheduled. If the load data variation is very large, load resources of other clusters can be scheduled to join the target cluster to deal with the load data variation. Therefore, the load data variation is a reference for scheduling, and the scheduling according to the load data variation can be set according to actual situations, which is not limited by the embodiment.
The method for dispatching the cluster load resources provided by the embodiment of the invention firstly collects the current load data of each server and each container in the server in the target cluster of the current time period and the historical load data corresponding to a plurality of continuous time periods before the current time period, thereby providing a basis for dispatching the corresponding load resources according to slight load data change in the follow-up process. The current load data and historical load data are then input into a predictive model trained to converge to predict load data for the next time period. And finally, scheduling resources in advance according to the load data variation between the predicted load data and the current load data so that the load resources can meet the load requirements of the target cluster in the next time period, thereby enabling the load resources of the target cluster to meet the load requirements in real time and reducing the hysteresis of actually scheduling the load resources. Meanwhile, when the load resources are scheduled, the load resources can be scheduled in a targeted manner according to the numerical condition of the load data variation, so that the utilization rate of the load resources is improved.
Fig. 4 is a flowchart illustrating a cluster load resource scheduling method according to a third embodiment of the present invention. As shown in fig. 4, the method for scheduling cluster load resources provided in this embodiment is further refined in step 203 on the basis of the method for scheduling cluster load resources provided in the previous embodiment of the present invention. The method for scheduling cluster load resources provided by this embodiment includes the following steps.
It should be noted that the load resource includes a load resource corresponding to a server in the cluster and a load resource corresponding to a container in the server.
In step S2031, the load data variation amount between the predicted load data and the current load data is calculated.
In this embodiment, the implementation manner of step 2031 is similar to that of step 203 in the previous embodiment of the present invention, and is not described in detail here.
Step S2032, determining whether the load data variation is greater than or equal to a preset scheduling threshold, if yes, going to step S2033, and if not, going to step S2034.
In this embodiment, the preset scheduling threshold may be set according to the actual conditions by classification according to the cluster level, the server level, and the container level, so as to more accurately utilize the load resources.
Step S2033, controlling the cloud control server to schedule the corresponding load resource according to the load data change amount.
In this embodiment, the cloud control server is a device for scheduling load resources including each cluster, each server in the cluster, and each container in the server. The cloud control server can schedule not only the load resources of other clusters but also the load resources in the target cluster, thereby providing a basis for realizing the redistribution of the load resources.
Optionally, in this embodiment, the preset scheduling threshold includes a first threshold and a second threshold. If the load data variation is greater than or equal to the preset scheduling threshold, controlling the cloud control server to schedule the corresponding load resource according to the load data variation, including:
and if the load data variation is larger than or equal to the first threshold, controlling the cloud control server to schedule the data interfaces of the plurality of servers corresponding to the plurality of clusters according to the load data variation so as to enable the load resources of the plurality of servers of the plurality of clusters to be added into the target cluster.
And redeploying the load resources of the servers of the clusters and the load resources of the target cluster through the resource deployment server to complete the load resource configuration of the target cluster.
And if the load data variation is smaller than the first threshold and larger than or equal to the second threshold, controlling the cloud control server to schedule the data interfaces of the plurality of servers corresponding to the other cluster according to the load data variation so as to enable the load resources of the plurality of servers of the other cluster to be added into the target cluster.
And the load resources of the plurality of servers of the other cluster and the load resources of the target cluster are redeployed through the resource deployment server to complete the load resource configuration of the target cluster.
In this embodiment, the first threshold corresponds to a threshold of a cluster level, and when the load data variation is greater than or equal to the first threshold, it represents that the load resource of the target cluster is completely insufficient to cope with the load pressure of the next time period, and the load resource difference is very large, and the corresponding load resource needs to be scheduled from other clusters and added to the target cluster to cope with the load pressure of the next time period. When the cloud controller server schedules load resources, data connection is realized on a network by scheduling a data interface, such as a data interface of a server in a cluster. When the data interface is scheduled, the load data to be processed can be introduced to the corresponding server through the data interface.
In this embodiment, the resource deployment server is configured to deploy the scheduled load resource and add the load resource to the target cluster. After the data interfaces of the servers in the other clusters are scheduled, the resource deployment server can reset the original load tasks born by the servers of the other clusters through the data interfaces, and then configures the load tasks and the target clusters into a new target cluster. After the configuration is completed, the target cluster includes the servers of other clusters and the servers of the original target cluster.
In this embodiment, the second threshold corresponds to a server-level threshold, and at this time, when the load data variation is smaller than the first threshold and greater than or equal to the second threshold, the load pressure of the target cluster in the next time period is larger. At this time, the load resources of the servers of other clusters are scheduled to join the target cluster, so that the load pressure of the next time period can be met. When configuring the server, the configuration may be performed according to the memory condition, the hard disk capacity, and the like of the server.
Optionally, in this embodiment, the preset scheduling threshold further includes a third threshold.
If the load data variation is greater than or equal to the preset scheduling threshold, controlling the cloud control server to schedule the corresponding load resource according to the load data variation comprises:
and if the load data variation is smaller than the second threshold and larger than or equal to a third threshold, controlling data interfaces of the corresponding idle containers in the cloud control server scheduling target cluster according to the load data variation, so that the load resources of the idle containers are redeployed through the resource deployment server, and the load corresponds to the load data variation.
In this embodiment, the third threshold corresponds to the container level, and when the load data variation is smaller than the second threshold and greater than or equal to the third threshold, it represents that the load pressure of the target cluster in the next time period is smaller at this time, and the scheduling and allocation of the load resource of the target cluster may be performed. The idle container represents a container in an idle state in the target cluster. When the container is configured, the corresponding configuration can be realized by modifying the memory parameters.
Step S2034 is performed to determine whether the load data variation is smaller than or equal to a preset release threshold, if yes, step 2035 is performed, and if not, step 2036 is performed.
In this embodiment, the release threshold refers to a threshold when the target cluster needs to release the resource to improve the resource utilization rate. Generally, when the load data in the next time period changes negatively compared with the load data in the current time period, it is necessary to consider whether the load data change is smaller than or equal to a preset release threshold.
Step S2035, the cloud control server is controlled to release the load resources of the corresponding server and/or container in the target cluster according to the load data change amount.
In this embodiment, the resource may be released according to the load data variation, and the load resource of the server and/or the container may be released correspondingly.
In step S2036, the load resource remains unchanged.
In this embodiment, when the load data variation is greater than the preset release threshold, it indicates that the load data variation does not meet the release standard, and at this time, the load resource may be maintained unchanged.
Meanwhile, in order to better understand the cluster load resource scheduling method of the embodiment, a training process of the load data prediction model will be described in detail below. As shown in fig. 5, the load data prediction model of the present embodiment uses an LSTM network model. The load data prediction model is roughly divided into an input layer, a hidden layer, an LSTM layer and an output layer, wherein the input layer corresponds to input parameter data, the output layer corresponds to output parameter data, and the hidden layer comprises correlation among the input parameter data. After passing through the input layer, the hidden layer, the LSTM layer, the input parameter data will output the result through the output layer.
During training, firstly, a training sample is obtained, wherein the training sample is a training sample corresponding to the load data prediction model, and the training sample comprises: historical load data and future load data for a plurality of consecutive time periods of the target cluster.
And inputting the training samples into a preset load data prediction model to train the load data prediction model.
And judging whether the load data prediction model meets the convergence condition or not by adopting an error function.
And if the load data prediction model meets the convergence condition, determining the load data prediction model meeting the convergence condition as the load data prediction model trained to be converged.
In this embodiment, the error function is used to determine whether an error between the output predicted load data and the future load data satisfies a convergence condition. The future load data is actual load data collected and corresponds to the historical load data in time.
Fig. 6 is a schematic structural diagram of a cluster load resource scheduling apparatus according to a fourth embodiment of the present invention, and as shown in fig. 6, in this embodiment, the cluster load resource scheduling apparatus 300 includes:
an obtaining module 301, configured to obtain current load data corresponding to a target cluster in a current time period and historical load data corresponding to a plurality of consecutive time periods before the current time period.
A determining module 302, configured to input and train the current load data and the plurality of historical load data into a converged load data prediction model, and determine predicted load data of a next time period.
And a scheduling module 303, configured to schedule the load resource according to the current load data and the predicted load data.
The cluster load resource scheduling apparatus provided in this embodiment may execute the technical solution of the method embodiment shown in fig. 2, and the implementation principle and technical effect of the apparatus are similar to those of the method embodiment shown in fig. 2, which are not described in detail herein.
Meanwhile, another embodiment of the cluster load resource scheduling device provided by the present invention further refines the cluster load resource scheduling device 300 on the basis of the cluster load resource scheduling device provided in the previous embodiment.
Optionally, the target cluster includes a plurality of servers. The server includes a plurality of containers.
The obtaining module 301, when obtaining current load data corresponding to a target cluster in a current time period, is specifically configured to:
and collecting current load data of each server in the target cluster and each container in the server in the current time period.
The scheduling module 303 is specifically configured to:
and calculating the load data variation between the predicted load data and the current load data, and scheduling the load resources according to the load data variation.
Optionally, in this embodiment, the load resource includes a load resource corresponding to a server in the cluster and a load resource corresponding to a container in the server.
The scheduling module 303 is specifically configured to, when scheduling the load resource according to the load data variation:
and judging whether the load data variation is greater than or equal to a preset scheduling threshold value. And if the load data variation is larger than or equal to the preset scheduling threshold, controlling the cloud control server to schedule the corresponding load resource according to the load data variation. And if the load data variation is smaller than the preset scheduling threshold, judging whether the load data variation is smaller than or equal to the preset release threshold. And if the load data variation is smaller than or equal to a preset release threshold, controlling the cloud control server to release the load resources of the corresponding server and/or container in the target cluster according to the load data variation.
Optionally, in this embodiment, the preset scheduling threshold includes a first threshold and a second threshold.
If the load data variation is greater than or equal to the preset scheduling threshold, the scheduling module 303 is specifically configured to, when controlling, according to the load data variation, the cloud control server to schedule the corresponding load resource:
and if the load data variation is larger than or equal to the first threshold, controlling the cloud control server to schedule the data interfaces of the plurality of servers corresponding to the plurality of clusters according to the load data variation so as to enable the load resources of the plurality of servers of the plurality of clusters to be added into the target cluster. And redeploying the load resources of the servers of the clusters and the load resources of the target cluster through the resource deployment server to complete the load resource configuration of the target cluster. And if the load data variation is smaller than the first threshold and larger than or equal to the second threshold, controlling the cloud control server to schedule the data interfaces of the plurality of servers corresponding to the other cluster according to the load data variation so as to enable the load resources of the plurality of servers of the other cluster to be added into the target cluster. And the load resources of the plurality of servers of the other cluster and the load resources of the target cluster are redeployed through the resource deployment server to complete the load resource configuration of the target cluster.
Optionally, in this embodiment, the preset scheduling threshold further includes a third threshold.
If the load data variation is greater than or equal to the preset scheduling threshold, the scheduling module 303 is specifically configured to, when controlling, according to the load data variation, the cloud control server to schedule the corresponding load resource:
and if the load data variation is smaller than the second threshold and larger than or equal to a third threshold, controlling data interfaces of the corresponding idle containers in the cloud control server scheduling target cluster according to the load data variation, so that the load resources of the idle containers are redeployed through the resource deployment server, and the load corresponds to the load data variation.
Optionally, in this embodiment, the method further includes:
the training module is used for obtaining training samples, the training samples are corresponding to the load data prediction model, and the training samples comprise: historical load data and future load data for a plurality of consecutive time periods of the target cluster. And inputting the training samples into a preset load data prediction model to train the load data prediction model. And judging whether the load data prediction model meets the convergence condition or not by adopting an error function. And if the load data prediction model meets the convergence condition, determining the load data prediction model meeting the convergence condition as the load data prediction model trained to be converged.
The cluster load resource scheduling apparatus provided in this embodiment may execute the technical solutions of the method embodiments shown in fig. 2 to fig. 5, and the implementation principles and technical effects thereof are similar to those of the method embodiments shown in fig. 2 to fig. 5, and are not described in detail here.
The invention also provides an electronic device, a computer readable storage medium and a computer program product according to the embodiments of the invention.
As shown in fig. 7, fig. 7 is a schematic structural diagram of an electronic device according to a fifth embodiment of the present invention. Electronic devices are intended for various forms of digital computers, such as laptops, desktops, workstations, personal digital assistants, servers, blade servers, mainframes, and other appropriate computers. The electronic device may also represent various forms of mobile devices, such as personal digital processing, cellular phones, smart phones, wearable devices, and other similar computing devices. The components shown herein, their connections and relationships, and their functions, are meant to be exemplary only, and are not meant to limit implementations of the inventions described and/or claimed herein.
As shown in fig. 7, the electronic apparatus includes: a processor 401, a memory 402, and a transceiver 403. The various components are interconnected using different buses and may be mounted on a common motherboard or in other manners as desired. The processor may process instructions for execution within the electronic device. The transceiver 403 is configured to output a scheduling control instruction to the cloud control server, and receive load data sent by the target cluster.
The memory 402 is a non-transitory computer readable storage medium provided by the present invention. The memory stores instructions executable by the at least one processor, so that the at least one processor executes the cluster load resource scheduling method provided by the present invention. The non-transitory computer readable storage medium of the present invention stores computer instructions for causing a computer to execute the cluster load resource scheduling method provided by the present invention.
Other embodiments of the invention will be apparent to those skilled in the art from consideration of the specification and practice of the invention disclosed herein. This application is intended to cover any variations, uses, or adaptations of the embodiments of the invention following, in general, the principles of the embodiments of the invention and including such departures from the present disclosure as come within known or customary practice within the art to which the embodiments of the invention pertains. It is intended that the specification and examples be considered as exemplary only, with a true scope and spirit of embodiments of the invention being indicated by the following claims.
It is to be understood that the embodiments of the present invention are not limited to the precise arrangements described above and shown in the drawings, and that various modifications and changes may be made without departing from the scope thereof. The scope of embodiments of the invention is limited only by the appended claims.
Claims (10)
1. A method for scheduling cluster load resources is characterized by comprising the following steps:
acquiring current load data corresponding to a target cluster in a current time period and historical load data corresponding to a plurality of continuous time periods before the current time period;
inputting and training the current load data and the plurality of historical load data into a converged load data prediction model, and determining predicted load data of the next time period;
and scheduling load resources according to the current load data and the predicted load data.
2. The method of claim 1, wherein the target cluster comprises a plurality of servers; the server comprises a plurality of containers;
the obtaining of the current load data corresponding to the target cluster in the current time period includes:
collecting current load data of each server in a target cluster and each container in the server in the current time period;
the scheduling load resources according to the current load data and the predicted load data comprises:
and calculating the load data variation between the predicted load data and the current load data, and scheduling load resources according to the load data variation.
3. The method of claim 2, wherein the load resources comprise load resources corresponding to servers in the cluster and load resources corresponding to containers in the servers;
the scheduling of load resources according to the load data variation includes:
judging whether the load data variation is greater than or equal to a preset scheduling threshold value or not;
if the load data variation is larger than or equal to a preset scheduling threshold, controlling a cloud control server to schedule corresponding load resources according to the load data variation;
if the load data variation is smaller than a preset scheduling threshold, judging whether the load data variation is smaller than or equal to a preset release threshold;
and if the load data variation is smaller than or equal to a preset release threshold, controlling the cloud control server to release the load resources of the corresponding server and/or container in the target cluster according to the load data variation.
4. The method of claim 3, wherein the preset scheduling threshold comprises a first threshold and a second threshold;
if the load data variation is greater than or equal to a preset scheduling threshold, controlling a cloud control server to schedule corresponding load resources according to the load data variation, including:
if the load data variation is larger than or equal to a first threshold value, controlling a cloud control server to schedule data interfaces of a plurality of servers corresponding to a plurality of clusters according to the load data variation so as to enable load resources of the plurality of servers of the plurality of clusters to be added into a target cluster;
redeploying a plurality of server load resources of a plurality of clusters and load resources of a target cluster through a resource deployment server to complete load resource configuration of the target cluster;
if the load data variation is smaller than the first threshold and larger than or equal to a second threshold, controlling a cloud control server to schedule a data interface of a plurality of servers corresponding to another cluster according to the load data variation so as to enable load resources of the plurality of servers of another cluster to be added into a target cluster;
and the load resources of the plurality of servers of the other cluster and the load resources of the target cluster are redeployed through the resource deployment server to complete the load resource configuration of the target cluster.
5. The method of claim 4, wherein the preset scheduling threshold further comprises a third threshold;
if the load data variation is greater than or equal to a preset scheduling threshold, controlling the cloud control server to schedule the corresponding load resource according to the load data variation comprises:
and if the load data variation is smaller than the second threshold and larger than or equal to a third threshold, controlling data interfaces of a plurality of idle containers in a scheduling target cluster of the cloud control server according to the load data variation, so that load resources of the idle containers are redeployed through the resource deployment server, and the load corresponds to the load data variation.
6. The method according to any one of claims 1-5, wherein before inputting the current load data and the plurality of historical load data into a predetermined load data prediction model, further comprising:
obtaining a training sample, wherein the training sample is a training sample corresponding to a load data prediction model, and the training sample comprises: historical load data and future load data for a plurality of consecutive time periods of the target cluster;
inputting the training samples into a preset load data prediction model to train the load data prediction model;
judging whether the load data prediction model meets a convergence condition or not by adopting an error function;
and if the load data prediction model meets the convergence condition, determining the load data prediction model meeting the convergence condition as a load data prediction model trained to be converged.
7. A cluster load resource scheduling apparatus, comprising:
the acquisition module is used for acquiring current load data corresponding to a target cluster in a current time period and historical load data corresponding to a plurality of continuous time periods before the current time period;
the determining module is used for inputting and training the current load data and the plurality of historical load data into a converged load data prediction model and determining predicted load data of the next time period;
and the scheduling module is used for scheduling the load resources according to the current load data and the predicted load data.
8. An electronic device, comprising: a memory, a processor, and a transceiver;
the processor, the memory and the transceiver are interconnected through a circuit;
the memory stores computer-executable instructions; the transceiver is used for outputting a scheduling control instruction to the cloud control server and receiving load data sent by the target cluster;
wherein the processor is configured to perform the method of cluster load resource scheduling according to any of claims 1 to 6 by the processor.
9. A computer-readable storage medium, having stored thereon computer-executable instructions for implementing the cluster load resource scheduling method according to any one of claims 1 to 6 when executed by a processor.
10. A computer program product comprising a computer program, characterized in that the computer program, when being executed by a processor, is adapted to carry out the method of cluster load resource scheduling according to any of the claims 1-6.
Priority Applications (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN202110196241.8A CN112860403B (en) | 2021-02-22 | 2021-02-22 | Cluster load resource scheduling method, device, equipment, medium and product |
Applications Claiming Priority (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN202110196241.8A CN112860403B (en) | 2021-02-22 | 2021-02-22 | Cluster load resource scheduling method, device, equipment, medium and product |
Publications (2)
| Publication Number | Publication Date |
|---|---|
| CN112860403A true CN112860403A (en) | 2021-05-28 |
| CN112860403B CN112860403B (en) | 2023-11-07 |
Family
ID=75988471
Family Applications (1)
| Application Number | Title | Priority Date | Filing Date |
|---|---|---|---|
| CN202110196241.8A Active CN112860403B (en) | 2021-02-22 | 2021-02-22 | Cluster load resource scheduling method, device, equipment, medium and product |
Country Status (1)
| Country | Link |
|---|---|
| CN (1) | CN112860403B (en) |
Cited By (9)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN113568754A (en) * | 2021-08-03 | 2021-10-29 | 安天科技集团股份有限公司 | Resource allocation method and device, computing equipment and storage medium |
| CN113904940A (en) * | 2021-09-03 | 2022-01-07 | 深圳市雷鸟网络传媒有限公司 | Resource adjusting method and device, electronic equipment and computer readable storage medium |
| CN114189482A (en) * | 2021-12-14 | 2022-03-15 | 郑州阿帕斯数云信息科技有限公司 | Control method, device and system for cluster resources |
| CN114416280A (en) * | 2021-12-20 | 2022-04-29 | 深信服科技股份有限公司 | Demand resource prediction method, device, equipment and storage medium |
| WO2022252895A1 (en) * | 2021-05-31 | 2022-12-08 | 腾讯科技（深圳）有限公司 | Resource allocation method and apparatus, and readable medium, electronic device and program product |
| CN115514996A (en) * | 2021-06-22 | 2022-12-23 | 武汉斗鱼鱼乐网络科技有限公司 | Method and device for determining state of live transcoding machine |
| CN115629858A (en) * | 2022-10-17 | 2023-01-20 | 南京航空航天大学 | Self-adaptive method for number of function examples in server-free background and application |
| CN115865684A (en) * | 2023-02-03 | 2023-03-28 | 浪潮电子信息产业股份有限公司 | Host scheduling method, system, equipment and computer readable storage medium |
| CN117421131A (en) * | 2023-12-18 | 2024-01-19 | 武汉泽塔云科技股份有限公司 | Intelligent scheduling method and system for monitoring power consumption load of server |
Citations (9)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| EP3033675A1 (en) * | 2013-08-13 | 2016-06-22 | Intel Corporation | Power balancing to increase workload density and improve energy efficiency |
| 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 |
| CN108009016A (en) * | 2016-10-31 | 2018-05-08 | 华为技术有限公司 | A kind of balancing resource load control method and colony dispatching device |
| CN108063784A (en) * | 2016-11-08 | 2018-05-22 | 阿里巴巴集团控股有限公司 | The methods, devices and systems of application cluster resource allocation under a kind of cloud environment |
| CN110221917A (en) * | 2019-05-23 | 2019-09-10 | 阿里巴巴集团控股有限公司 | For distributing the method and device of stream data |
| CN111491006A (en) * | 2020-03-03 | 2020-08-04 | 天津大学 | Load-aware cloud computing resource elastic distribution system and method |
| CN111581174A (en) * | 2019-02-18 | 2020-08-25 | 北京搜狗科技发展有限公司 | Resource management method and device based on distributed cluster system |
| CN112087504A (en) * | 2020-08-31 | 2020-12-15 | 浪潮通用软件有限公司 | Dynamic load balancing method and device based on working load characteristics |
| CN112100024A (en) * | 2020-08-14 | 2020-12-18 | 北京浪潮数据技术有限公司 | Resource load abnormity detection method, device and equipment |
-
2021
- 2021-02-22 CN CN202110196241.8A patent/CN112860403B/en active Active
Patent Citations (9)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| EP3033675A1 (en) * | 2013-08-13 | 2016-06-22 | Intel Corporation | Power balancing to increase workload density and improve energy efficiency |
| CN108009016A (en) * | 2016-10-31 | 2018-05-08 | 华为技术有限公司 | A kind of balancing resource load control method and colony dispatching device |
| CN108063784A (en) * | 2016-11-08 | 2018-05-22 | 阿里巴巴集团控股有限公司 | The methods, devices and systems of application cluster resource allocation under a kind of cloud environment |
| 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 |
| CN111581174A (en) * | 2019-02-18 | 2020-08-25 | 北京搜狗科技发展有限公司 | Resource management method and device based on distributed cluster system |
| CN110221917A (en) * | 2019-05-23 | 2019-09-10 | 阿里巴巴集团控股有限公司 | For distributing the method and device of stream data |
| CN111491006A (en) * | 2020-03-03 | 2020-08-04 | 天津大学 | Load-aware cloud computing resource elastic distribution system and method |
| CN112100024A (en) * | 2020-08-14 | 2020-12-18 | 北京浪潮数据技术有限公司 | Resource load abnormity detection method, device and equipment |
| CN112087504A (en) * | 2020-08-31 | 2020-12-15 | 浪潮通用软件有限公司 | Dynamic load balancing method and device based on working load characteristics |
Non-Patent Citations (4)
| Title |
|---|
| HAO ZHANG等: "An Adaptive Task Scheduling System Based on Real-time Clustering and NetFlow Prediction", 《2019 IEEE 5TH INTERNATIONAL CONFERENCE ON COMPUTER AND COMMUNICATIONS (ICCC)》, pages 77 - 80 * |
| 宋彩华: "云资源监控中的数据传输模型研究", 《中国优秀硕士学位论文全文数据库信息科技辑》, pages 139 - 246 * |
| 林韧昊等: "Openstack云平台自适应认证系统负载均衡策略研究", OPENSTACK云平台自适应认证系统负载均衡策略研究, vol. 36, no. 9, pages 302 - 309 * |
| 陈强: "基于ARMA模型预测的云计算资源调度策略研究", 《中国优秀硕士学位论文全文数据库信息科技辑》, pages 137 - 22 * |
Cited By (11)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| WO2022252895A1 (en) * | 2021-05-31 | 2022-12-08 | 腾讯科技（深圳）有限公司 | Resource allocation method and apparatus, and readable medium, electronic device and program product |
| CN115514996A (en) * | 2021-06-22 | 2022-12-23 | 武汉斗鱼鱼乐网络科技有限公司 | Method and device for determining state of live transcoding machine |
| CN113568754A (en) * | 2021-08-03 | 2021-10-29 | 安天科技集团股份有限公司 | Resource allocation method and device, computing equipment and storage medium |
| CN113904940A (en) * | 2021-09-03 | 2022-01-07 | 深圳市雷鸟网络传媒有限公司 | Resource adjusting method and device, electronic equipment and computer readable storage medium |
| CN114189482A (en) * | 2021-12-14 | 2022-03-15 | 郑州阿帕斯数云信息科技有限公司 | Control method, device and system for cluster resources |
| CN114416280A (en) * | 2021-12-20 | 2022-04-29 | 深信服科技股份有限公司 | Demand resource prediction method, device, equipment and storage medium |
| CN115629858A (en) * | 2022-10-17 | 2023-01-20 | 南京航空航天大学 | Self-adaptive method for number of function examples in server-free background and application |
| CN115865684A (en) * | 2023-02-03 | 2023-03-28 | 浪潮电子信息产业股份有限公司 | Host scheduling method, system, equipment and computer readable storage medium |
| CN115865684B (en) * | 2023-02-03 | 2023-05-23 | 浪潮电子信息产业股份有限公司 | Host scheduling method, system, equipment and computer readable storage medium |
| CN117421131A (en) * | 2023-12-18 | 2024-01-19 | 武汉泽塔云科技股份有限公司 | Intelligent scheduling method and system for monitoring power consumption load of server |
| CN117421131B (en) * | 2023-12-18 | 2024-03-26 | 武汉泽塔云科技股份有限公司 | Intelligent scheduling method and system for monitoring power consumption load of server |
Also Published As
| Publication number | Publication date |
|---|---|
| CN112860403B (en) | 2023-11-07 |
Similar Documents
| Publication | Publication Date | Title |
|---|---|---|
| CN112860403B (en) | Cluster load resource scheduling method, device, equipment, medium and product | |
| Bhattacharjee et al. | Barista: Efficient and scalable serverless serving system for deep learning prediction services | |
| US11032212B2 (en) | Systems and methods for provision of a guaranteed batch | |
| CN109408205B (en) | Task scheduling method and device based on hadoop cluster | |
| US11436050B2 (en) | Method, apparatus and computer program product for resource scheduling | |
| EP3462705B1 (en) | Provisioning of software applications on edge devices in an internet-of-things environment | |
| CN108052384B (en) | Task processing method, service platform and electronic equipment | |
| CN112380020A (en) | Computing power resource allocation method, device, equipment and storage medium | |
| CN110990138B (en) | Resource scheduling method, device, server and storage medium | |
| CN113946431B (en) | Resource scheduling method, system, medium and computing device | |
| CN112445575A (en) | Multi-cluster resource scheduling method, device and system | |
| CN112540841A (en) | Task scheduling method and device, processor and electronic equipment | |
| CN115421930B (en) | Task processing method, system, device, equipment and computer readable storage medium | |
| CN110442428A (en) | The coordination approach of Docker container | |
| CN113515382A (en) | Cloud resource allocation method and device, electronic equipment and program product | |
| CN113672345A (en) | IO prediction-based cloud virtualization engine distributed resource scheduling method | |
| CN114911615B (en) | Intelligent prediction scheduling method and application during micro-service running | |
| CN112685160A (en) | Scheduling method and device of timing task, terminal equipment and computer storage medium | |
| CN114741200A (en) | Data center station-oriented computing resource allocation method and device and electronic equipment | |
| CN107203256B (en) | Energy-saving distribution method and device under network function virtualization scene | |
| CN114490048A (en) | Task execution method and device, electronic equipment and computer storage medium | |
| CN116402318B (en) | Multi-stage computing power resource distribution method and device for power distribution network and network architecture | |
| CN112463361A (en) | Method and equipment for distributing elastic resources of distributed computation | |
| CN113254172A (en) | Task scheduling method and device | |
| CN115827232A (en) | Method, device, system and equipment for determining configuration for service model |
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 |