CN105760224A - Dynamic resource adjustment method and device - Google Patents
Dynamic resource adjustment method and device Download PDFInfo
- Publication number
- CN105760224A CN105760224A CN201610006553.7A CN201610006553A CN105760224A CN 105760224 A CN105760224 A CN 105760224A CN 201610006553 A CN201610006553 A CN 201610006553A CN 105760224 A CN105760224 A CN 105760224A
- Authority
- CN
- China
- Prior art keywords
- resource
- performance data
- application
- time
- threshold value
- Prior art date
- Legal status (The legal status is an assumption and is not a legal conclusion. Google has not performed a legal analysis and makes no representation as to the accuracy of the status listed.)
- Pending
Links
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
- G06F9/50—Allocation of resources, e.g. of the central processing unit [CPU]
- G06F9/5061—Partitioning or combining of resources
Landscapes
- Engineering & Computer Science (AREA)
- Software Systems (AREA)
- Theoretical Computer Science (AREA)
- Physics & Mathematics (AREA)
- General Engineering & Computer Science (AREA)
- General Physics & Mathematics (AREA)
- Debugging And Monitoring (AREA)
Abstract
The invention provides a dynamic resource adjustment method and device.The method comprises the steps that for a resource configured for an application, performance data corresponding to the resource is obtained; the performance data trend situation with a follow-up period of time is predicted through a neural network by means of the performance data; when the predicted performance data trend situation meets the condition that predicted performance data is larger than a preset first threshold from first time, resources are increased for the application before the first time, or the predicted performance data is smaller than a preset second threshold from second time, and spare resources configured for the application exist currently, the spare resources configured for the application are decreased before the second time.By means of the dynamic resource adjustment method and device, dynamic resource adjustment can be conducted in advance, and therefore the phenomena of temporary access blocking and tardiness and the like are relieved, and the user application experience is improved.
Description
Technical field
The present invention relates to communication technical field, particularly relate to dynamic adjusting method and the device of a kind of resource.
Background technology
Along with the development of Internet technology, especially the developing rapidly of cloud computing technology, can be a multiple resource of application configuration at present, and such as example resource, an example resource can be exactly a virtual machine, and these multiple resources complete the process of this application jointly.Such as, apply for web (WWW), for web application configuration resource 1, resource 2 and resource 3, these 3 resources jointly realize web application.
The performance used due to application constantly changes, when the performance that application uses is more, CPU (CentralProcessingUnit such as resource, central processing unit) utilization rate reached 90%, then resource will be unable to ensure the properly functioning of application, need to distribute more multiple resource to application, it is ensured that the availability of application.
In order to realize this process, generally the performance data of resource is monitored, and judges that whether this performance data is more than predetermined threshold value, if it is, distribute more multiple resource for this application.Such as, jointly realize in the process of web application in resource 1, resource 2 and resource 3, if the CPU usage of resource 1, resource 2 and resource 3 is all higher than predetermined threshold value 85%, then distribute new resource 4 for web application.
But, the process owing to distributing new resources needs certain time, and when monitoring performance data more than predetermined threshold value, the performance that resource is used is higher, therefore before completing new resources distribution, can cause that Caton phenomenon occurs in application, reduce user and apply experience.And, if system has not had available resources, then needing to notify that manager creates new resources, the process of distribution new resources takes longer for.
Summary of the invention
The present invention provides the dynamic adjusting method of a kind of resource, said method comprising the steps of:
For the resource for application configuration, it is thus achieved that the performance data that described resource is corresponding；
Utilize described performance data, use the performance data tendency situation in neural network prediction follow-up a period of time；
When prediction performance data tendency situation be: from the very first time, it was predicted that performance data more than preset first threshold value, then before the described very first time, increase the resource of described application；Or, from the second time, it was predicted that performance data less than default Second Threshold, and there is currently the idling-resource into described application configuration, then, before described second time, be reduced to the idling-resource of described application configuration；Wherein, described preset first threshold value is more than described default Second Threshold.
The process of the performance data that the described resource of described acquisition is corresponding, specifically includes:
Configuration monitoring agency in described resource
Receive the performance data that described resource that described monitoring agent reports is corresponding.
Described utilize described performance data, the process of the performance data tendency situation in use neural network prediction follow-up a period of time, specifically include: use neutral net that described performance data is trained, go forward side by side line parameter optimizing, to obtain the training pattern of described neutral net, and described training pattern is utilized to predict the performance data tendency situation in follow-up a period of time；Described neutral net specifically includes support vector machines neutral net.
The process of the resource of the described application of described increase, specifically includes:
If currently there is no available resources, then notify that manager created new resources before the described very first time, and after new resources have created, increase the resource of described application.
Described method also includes: after obtaining the performance data that described resource is corresponding, if described performance data is more than default 3rd threshold value, then increase the resource of described application；When increasing the resource of described application, if currently there is no available resources, then notify that manager creates new resources, and after new resources have created, increase the resource of described application；Or, if described performance data is less than default 4th threshold value, then reduce the resource of described application；Described the 3rd threshold value of presetting presets the 4th threshold value more than described.
The present invention provides the dynamic adjusting device of a kind of resource, and described device specifically includes:
Obtain module, for for the resource for application configuration, it is thus achieved that the performance data that described resource is corresponding；
Prediction module, is used for utilizing described performance data, uses the performance data tendency situation in neural network prediction follow-up a period of time；
Adjusting module, for when prediction performance data tendency situation be: from the very first time, it was predicted that performance data more than preset first threshold value, then before the described very first time, increase the resource of described application；Or, from the second time, it was predicted that performance data less than default Second Threshold, and there is currently the idling-resource into described application configuration, then, before described second time, be reduced to the idling-resource of described application configuration；Wherein, described preset first threshold value is more than described default Second Threshold.
Described acquisition module, specifically in the process obtaining performance data corresponding to described resource, configuration monitoring agency in described resource, and receive the performance data that described resource that described monitoring agent reports is corresponding.
Described prediction module, specifically for utilizing described performance data, use in the process of the performance data tendency situation in neural network prediction follow-up a period of time, use neutral net that described performance data is trained, go forward side by side line parameter optimizing, to obtain the training pattern of described neutral net, and described training pattern is utilized to predict the performance data tendency situation in follow-up a period of time；Described neutral net specifically includes SVM neutral net.
Described adjusting module, specifically for, in increasing the process of resource of described application, if currently not had available resources, then notifying that manager created new resources before the described very first time, and after new resources have created, increase the resource of described application.
Described adjusting module, is additionally operable to, after obtaining the performance data that described resource is corresponding, if described performance data is more than default 3rd threshold value, then increase the resource of described application；When increasing the resource of described application, if currently there is no available resources, then notify that manager creates new resources, and after new resources have created, increase the resource of described application；Or, if described performance data is less than default 4th threshold value, then reduce the resource of described application；Described the 3rd threshold value of presetting presets the 4th threshold value more than described.
Based on technique scheme, in the embodiment of the present invention, can based on current performance data, predict the performance data tendency situation in follow-up a period of time, when the performance data tendency situation of prediction is: from the very first time, the performance data of prediction is more than preset first threshold value, then before the first time, increase the resource of application；Or, from the second time, it was predicted that performance data less than default Second Threshold, and there is currently the idling-resource into application configuration, then, before the second time, be reduced to the idling-resource of application configuration；Based on aforesaid way, it is possible to use the performance data tendency situation of prediction, the resource of application is dynamically adjusted, such that it is able to before performance data is more than threshold value, increase the resource of application, thus reaching the purpose of resource automatic governing so that the utilization rate of resource reaches the highest.And, performance data tendency situation based on prediction, carry out the dynamic adjustment of resource in advance, namely when increasing resource, performance data is also not up to threshold value, thus solve provisional access card pause, the phenomenon such as slow, provided the user more smooth, resources control accurately, improved user and apply experience.And, if currently there is no available resources, then prior notice manager can create new resources, without just notifying that until performance data reaches threshold value manager creates new resources, therefore, when performance data reaches threshold value, there have been new resources can distribute to related application.
Accompanying drawing explanation
Fig. 1 is the application scenarios schematic diagram in one embodiment of the present invention；
Fig. 2 is the flow chart of the dynamic adjusting method of the resource in one embodiment of the present invention；
Fig. 3 is the performance data tendency situation comparison schematic diagram with actual performance data tendency situation of the prediction in one embodiment of the present invention；
Fig. 4 is the hardware structure diagram controlling equipment in one embodiment of the present invention；
Fig. 5 is the structure chart of the dynamic adjusting device of the resource in one embodiment of the present invention.
Detailed description of the invention
For problems of the prior art, the embodiment of the present invention proposes the dynamic adjusting method of a kind of resource, the method can apply at least to include in the system of control equipment and multiple resource, and the dynamic adjusting method of this resource can be applied on this control equipment.Wherein, these multiple resources are the multiple resources configured for an application (as web applies), such as example resource, resource can be storage resource, calculate resource, Internet resources etc., one example resource can be exactly a virtual machine, and these multiple resources have been used for the process of this application, and these multiple resources are called the resource pool of this application.With Fig. 1 application scenarios schematic diagram being the embodiment of the present invention, for web application configuration resource 1, resource 2 and resource 3, these 3 resources all can realize web application, it is contemplated that the utilization rate of resource, only resource 1 and resource 2 are processing web application at present, and resource 3 is in idle condition, untreated web applies, and resource 1, resource 2 and resource 3 are the resource pools of web application.Under above-mentioned application scenarios, as in figure 2 it is shown, the dynamic adjusting method of this resource specifically may comprise steps of:
Step 201, for the resource for application configuration, it is thus achieved that the performance data that this resource is corresponding.
In the embodiment of the present invention, being configured with monitoring agent in resource, this monitoring agent periodically collects the performance data of this resource, and by current collection to performance data be sent to control equipment.On this basis, control equipment and obtain the process of performance data corresponding to resource, specifically can include but not limited to following manner: control equipment receives the performance data that this resource that monitoring agent reports is corresponding.
Wherein, resource specifically may refer to example resource, and in actual applications, resource can be storage resource, calculating resource, Internet resources etc., performance data corresponding to these resources specifically can include but not limited to that CPU usage, memory usage, number of sessions, web connect number etc..Describe in order to convenient, follow-up connect number for web and illustrate.
Wherein, system can also include load-balancing device, when multiple resources process application simultaneously, load-balancing device is when receiving the request accessing this application, based on load balancing, this request can be assigned in a resource and process, therefore, the performance data of multiple resources is essentially identical, the performance data of one resource also just represents the performance data of other resource, therefore, the follow-up performance data for a resource illustrates, and the process of the performance data of multiple resources is similar, follow-up repeats no more.
Wherein, equipment is controlled after obtaining the performance data that resource is corresponding, the performance data that storage resource is corresponding.When performance data corresponding to resource is not up to predetermined number, then controls equipment and be directly based upon the performance data of current acquisition and carry out the dynamic adjustment of resource.When the performance data that resource is corresponding reaches predetermined number, then control equipment and carry out the dynamic adjustment of resource based on the performance data tendency situation in follow-up a period of time, namely perform subsequent step 202 and step 203.The value of predetermined number can be configured according to practical situation, is to ensure that the minimum data amount of the performance data tendency situation can predicted in follow-up a period of time.
In the embodiment of the present invention, control equipment is directly based upon the current performance data obtained and carries out the process dynamically adjusted of resource, specifically can include but not limited to following manner: after obtaining the performance data that resource is corresponding, if this performance data is more than default 3rd threshold value, then increase the resource of application；Wherein, when increasing the resource of application, if currently there is no available resources, then notify that manager creates new resources, and after new resources have created, increase the resource of application.Or, if performance data is less than default 4th threshold value, then reduce the resource of application.Wherein, the 3rd threshold value is preset more than default 4th threshold value.Or, if performance data is be more than or equal to default 4th threshold value, and less than or equal to default 3rd threshold value, does not then increase the resource of application, do not reduce the resource of application yet, and the resource quantity being to maintain application is constant.
Such as, providing the web system of application for a set of, the web of monitoring agent in resource 1 periodically the collections resource 1 of (as every 5 minutes) connects number, and by current collection to web connection number be sent to control equipment.The web of resource 1, after the web receiving resource 1 connects number, is connected number and preserves by control equipment.At most baseline, owing to web connects the quantity of number not up to predetermined number, therefore control equipment and be directly based upon the performance data of current acquisition and carry out the dynamic adjustment of resource.After collecting enough a number of web connection numbers, then the quantity of web connection number is up to predetermined number, therefore controls equipment and carries out the dynamic adjustment of resource, execution subsequent step 202 and step 203 based on the performance data tendency situation in follow-up a period of time.Wherein, the web that control equipment preserves connects number can form following array [101011121314151820], and numerical value (10,10,11 etc.) is web and connects number.
Step 202, utility data, use the performance data tendency situation in neural network prediction follow-up a period of time.
Wherein, this performance data refers to the performance data reaching predetermined number that control equipment preserves.
In the embodiment of the present invention, utility data, the process of the performance data tendency situation in use neural network prediction follow-up a period of time, specifically can include but not limited to following manner: use neutral net that performance data is trained, go forward side by side line parameter optimizing, to obtain the training pattern of this neutral net, and this training pattern is utilized to predict the performance data tendency situation in follow-up a period of time.Wherein, this neutral net specifically can include but not limited to: SVM (SupportVectorMachine, support vector machine) neutral net.
Wherein, SVM neutral net is a learning model having supervision, is commonly used to carry out pattern recognition, classification, regression analysis.SVM neutral net is by a nonlinear mapping p, is mapped to by sample space in a higher-dimension or infinite dimensional feature space so that the Nonlinear separability problem in sample space, is converted into the linear separability problem in feature space.Briefly, it is simply that rise peacekeeping linearisation.Liter dimension refers to make sample to higher dimensional space and maps, and in the feature space of higher-dimension, realizes linear partition (or recurrence) by a linear hyperplane, the data of low dimensional is raised to high latitude and carries out classifying, returning.
On this basis, in the embodiment of the present invention, after having collected a number of performance data, it is possible to periodically using SVM neutral net that performance data is trained, line parameter optimizing of going forward side by side, to obtain the training pattern of this SVM neutral net.Further, utilizing this training pattern to predict the performance data tendency situation (namely utilizing this training pattern that performance data at no distant date is returned) in follow-up a period of time, the performance data tendency situation in follow-up a period of time of namely predicting is periodically to carry out.
Wherein, using SVM neutral net, performance data is trained and parameter optimization, to obtain in the process of the training pattern of SVM neutral net, RBF (RadialBasisFunction can be selected, RBF) core is as kernel function, the core parameter C (penalty coefficient) and gamma (attribute of input data or characteristic number) of RBF core is carried out parameter optimization, find out C value and the gamma value of optimum, make training pattern error in recurrence minimum, so that the prediction accuracy of training pattern is the highest.
Such as, controlling equipment after preserving web and connecting number [101011121314151820], web connects number [101011121314151820] and brings in RBF core and be trained, line parameter optimizing of going forward side by side, thus obtaining the training pattern of SVM neutral net.Based on this training pattern, regression (recurrence) characteristic of SVM neutral net can be utilized, predict the performance data tendency situation in follow-up a period of time, be [2122232425] as predicted that the web in follow-up a period of time connects the tendency situation of number.
Step 203, utilizes the performance data tendency situation of prediction, and the resource of application is dynamically adjusted.
In the embodiment of the present invention, utilize the performance data tendency situation of prediction, the resource of application is carried out the dynamic process adjusted, specifically can include but not limited to following manner: if the performance data tendency situation of prediction is: from the very first time, the performance data of prediction is more than preset first threshold value, then before the first time, the resource of application is increased.Or, if the performance data tendency situation of prediction is: from the second time, it was predicted that performance data less than default Second Threshold, and there is currently the idling-resource into this application configuration, then, before the second time, be reduced to the idling-resource of this application configuration.
Wherein, preset first threshold value is more than default Second Threshold.
Further, if the performance data tendency situation of prediction is: performance data is be more than or equal to default Second Threshold, and less than or equal to preset first threshold value, does not then increase the resource of application, does not also reduce the resource of application, and the resource quantity being to maintain application is constant.Wherein, the idling-resource for this application configuration refers to: the resource being in idle condition existed in the resource pool that this application is corresponding.
Further, when increasing the resource of application, if currently there is no available resources, then notify that manager creates new resources before the first time, and after new resources have created, increase the resource of application.
Such as, number is connected for web, assume that preset first threshold value is 22, if it is [2122232425] that the web of prediction connects number tendency situation, then from the web predicted time (i.e. the very first time) connecting several 23 correspondences, the web of prediction connects number more than preset first threshold value, based on this, before web connects the predicted time of several 23 correspondences, increase the resource of web application, as increased resource 3 for web application, being increased by resource thus connecting before number is actually reached 23 at web, carrying out dilatation operation in advance.
In the embodiment of the present invention, if there is error in the performance data tendency situation of prediction, cause the performance data tendency situation failed according to prediction, the resource of application is dynamically adjusted, then obtaining performance data (the namely current actual performance data that resource is corresponding, rather than the performance data of prediction) after, if this performance data is more than default 3rd threshold value, then increase the resource of application；Wherein, when increasing the resource of application, if currently there is no available resources, then notify that manager creates new resources, and after new resources have created, increase the resource of application.Or, if performance data is less than default 4th threshold value, then reduce the resource of application.Wherein, the 3rd threshold value is preset more than default 4th threshold value.Or, if performance data is be more than or equal to default 4th threshold value, and less than or equal to default 3rd threshold value, does not then increase the resource of application, do not reduce the resource of application yet, and the resource quantity being to maintain application is constant.
Wherein, presetting the 3rd threshold value can be identical with preset first threshold value, it is also possible to different from preset first threshold value.Presetting the 4th threshold value can be identical with default Second Threshold, it is also possible to different from default Second Threshold.
Assume that the performance data of resource 1 and/or resource 2 more than preset first threshold value or presets the 3rd threshold value, owing to currently exist for realizes the available resources 3 of web application, therefore increasing resource 3 for web application, afterwards, resource 1, resource 2 and resource 3 realize web application jointly.Jointly realize in the process of web application in resource 1, resource 2 and resource 3, the performance data assuming to have in resource 1, resource 2 and resource 3 any one or multiple resource more than preset first threshold value or presets the 3rd threshold value, owing to there is currently no the available resources for realizing web application, therefore notify that manager creates new resources, created the available resources 4 for realizing web application by manager, and increase resource 4 for web application.
Assume that the performance data of resource 1 and resource 2 is respectively less than default Second Threshold, owing in resource pool, currently exist for realizes the available resources 3 (i.e. idling-resource) of web application, therefore, it is reduced to the idling-resource of web application configuration, namely resource 3 can be discharged, resource 3 is no longer allocated to web application, and now this resource 3 can be allocated to other application.Assume that the performance data of resource 1 and resource 2 is respectively less than default 4th threshold value, therefore, the resource of web application can be reduced, namely resource 1 is only used to process web application, although now resource 2 and resource 3 do not process web application, but resource 2 and resource 3 are still configured to web application, it is impossible to be allocated to other application, be simply currently at idle condition.
Based on technique scheme, in the embodiment of the present invention, can based on current performance data, predict the performance data tendency situation in follow-up a period of time, when the performance data tendency situation of prediction is: from the very first time, the performance data of prediction is more than preset first threshold value, then before the first time, increase the resource of application；Or, from the second time, it was predicted that performance data less than default Second Threshold, and there is currently the idling-resource into application configuration, then, before the second time, be reduced to the idling-resource of application configuration；Based on aforesaid way, the performance data tendency situation of prediction can be utilized, the resource of application is dynamically adjusted, thus before performance data is more than threshold value, increasing the resource of application, it is achieved Intellisense and prediction, more intelligent realizes resource pre-coordination, reach the purpose of resource automatic governing, make the utilization rate of resource reach the highest.Wherein, the performance data tendency situation of prediction is essentially identical with actual performance data tendency situation, as shown in Figure 3, comparison schematic diagram for the performance data tendency situation predicted and actual performance data tendency situation, dotted line is actual performance data tendency situation, the performance data tendency situation that straight line is then predicted, is the performance data tendency situation of new prediction after vertical line, from figure 3, it can be seen that the performance data tendency situation of prediction is essentially identical with actual performance data tendency situation.And, performance data tendency situation based on prediction, carry out the dynamic adjustment of resource in advance, namely when increasing resource, performance data is also not up to threshold value, thus solve provisional access card pause, the phenomenon such as slow, provided the user more smooth, resources control accurately, improved user and apply experience.And, if currently there is no available resources, then prior notice manager can create new resources, without just notifying that until performance data reaches threshold value manager creates new resources, therefore, when performance data reaches threshold value, there have been new resources can distribute to related application.
Based on the inventive concept same with said method, additionally providing the dynamic adjusting device of a kind of resource in the embodiment of the present invention, the dynamic adjusting device of this resource can be applied on the control device.Wherein, the dynamic adjusting device of this resource can be realized by software, it is also possible to is realized by the mode of hardware or software and hardware combining.Implemented in software for example, as the device on a logical meaning, it is the processor controlling equipment by its place, computer program instructions corresponding in reading non-volatile storage is formed.Say from hardware view, as shown in Figure 4, for the present invention propose resource dynamic adjusting device place control equipment a kind of hardware structure diagram, except the processor shown in Fig. 4, nonvolatile memory, control equipment can also include other hardware, such as the forwarding chip of responsible process message, network interface, internal memory etc.；From hardware configuration, this control equipment is it is also possible that distributed apparatus, it is possible to include multiple interface card, in order to carry out the extension of Message processing at hardware view.
As shown in Figure 5, the structure chart of dynamic adjusting device for the resource that the present invention proposes, it is possible on the control device, the dynamic adjusting device of described resource specifically includes in application: obtain module 11, for for the resource for application configuration, it is thus achieved that the performance data that described resource is corresponding；Prediction module 12, is used for utilizing described performance data, uses the performance data tendency situation in neural network prediction follow-up a period of time；Adjusting module 13, for when prediction performance data tendency situation be: from the very first time, it was predicted that performance data more than preset first threshold value, then before the described very first time, increase the resource of described application；Or, from the second time, it was predicted that performance data less than default Second Threshold, and there is currently the idling-resource into described application configuration, then, before described second time, be reduced to the idling-resource of described application configuration；Wherein, described preset first threshold value is more than described default Second Threshold.
In the embodiment of the present invention, described acquisition module 11, specifically in the process obtaining performance data corresponding to described resource, configuration monitoring agency in described resource, and receive the performance data that described resource that described monitoring agent reports is corresponding.
Described prediction module 12, specifically for utilizing described performance data, use in the process of the performance data tendency situation in neural network prediction follow-up a period of time, use neutral net that described performance data is trained, go forward side by side line parameter optimizing, to obtain the training pattern of described neutral net, and described training pattern is utilized to predict the performance data tendency situation in follow-up a period of time；Described neutral net includes SVM neutral net.
Described adjusting module 13, specifically for, in increasing the process of resource of described application, if currently not had available resources, then notifying that manager created new resources before the described very first time, and after new resources have created, increase the resource of described application.
Described adjusting module 13, is additionally operable to, after obtaining the performance data that described resource is corresponding, if described performance data is more than default 3rd threshold value, then increase the resource of described application；When increasing the resource of described application, if currently there is no available resources, then notify that manager creates new resources, and after new resources have created, increase the resource of described application；Or, if described performance data is less than default 4th threshold value, reduce the resource of described application；Described the 3rd threshold value of presetting presets the 4th threshold value more than described.
Based on technique scheme, in the embodiment of the present invention, can based on current performance data, predict the performance data tendency situation in follow-up a period of time, and utilize the performance data tendency situation of prediction, the resource of application is dynamically adjusted, such that it is able to before performance data is more than threshold value, increase the resource of application, thus reaching the purpose of resource automatic governing so that the utilization rate of resource reaches the highest.And, performance data tendency situation based on prediction, carry out the dynamic adjustment of resource in advance, namely when increasing resource, performance data is also not up to threshold value, thus solve provisional access card pause, the phenomenon such as slow, provided the user more smooth, resources control accurately, improved user and apply experience.And, if currently there is no available resources, then prior notice manager can create new resources, without just notifying that until performance data reaches threshold value manager creates new resources, therefore, when performance data reaches threshold value, there have been new resources can distribute to related application.
Wherein, the modules of apparatus of the present invention can be integrated in one, it is also possible to separates and disposes.Above-mentioned module can merge into a module, it is also possible to is further split into multiple submodule.
Through the above description of the embodiments, those skilled in the art is it can be understood that can add the mode of required general hardware platform by software to the present invention and realize, naturally it is also possible to by hardware, but in a lot of situation, the former is embodiment more preferably.Based on such understanding, the part that prior art is contributed by technical scheme substantially in other words can embody with the form of software product, this computer software product is stored in a storage medium, including some instructions with so that a computer equipment (can be personal computer, server, or the network equipment etc.) perform the method described in each embodiment of the present invention.It will be appreciated by those skilled in the art that accompanying drawing is the schematic diagram of a preferred embodiment, module or flow process in accompanying drawing are not necessarily implemented necessary to the present invention.
It will be appreciated by those skilled in the art that the module in the device in embodiment can describe according to embodiment to carry out being distributed in the device of embodiment, it is also possible to carry out respective change and be disposed other than in one or more devices of the present embodiment.The module of above-described embodiment can merge into a module, it is possible to is further split into multiple submodule.The invention described above embodiment sequence number, just to describing, does not represent the quality of embodiment.
The several specific embodiments being only the present invention disclosed above, but, the present invention is not limited to this, and the changes that any person skilled in the art can think of all should fall into protection scope of the present invention.
Claims (10)
1. the dynamic adjusting method of a resource, it is characterised in that said method comprising the steps of:
For the resource for application configuration, it is thus achieved that the performance data that described resource is corresponding；
Utilize described performance data, use the performance data tendency situation in neural network prediction follow-up a period of time；
When prediction performance data tendency situation be: from the very first time, it was predicted that performance data more than preset first threshold value, then before the described very first time, increase the resource of described application；Or, from the second time, it was predicted that performance data less than default Second Threshold, and there is currently the idling-resource into described application configuration, then, before described second time, be reduced to the idling-resource of described application configuration；Wherein, described preset first threshold value is more than described default Second Threshold.
2. method according to claim 1, it is characterised in that the process of the performance data that the described resource of described acquisition is corresponding, specifically includes:
Configuration monitoring agency in described resource；
Receive the performance data that described resource that described monitoring agent reports is corresponding.
3. method according to claim 1, it is characterised in that described utilize described performance data, uses the process of performance data tendency situation in neural network prediction follow-up a period of time, specifically includes:
Use neutral net that described performance data is trained, line parameter optimizing of going forward side by side, to obtain the training pattern of described neutral net, and utilize described training pattern to predict the performance data tendency situation in follow-up a period of time；Wherein, described neutral net specifically includes support vector machines neutral net.
4. method according to claim 1, it is characterised in that the process of the resource of the described application of described increase, specifically includes:
If currently there is no available resources, then notify that manager created new resources before the described very first time, and after new resources have created, increase the resource of described application.
5. method according to claim 1, it is characterised in that described method also includes:
After obtaining the performance data that described resource is corresponding, if described performance data is more than default 3rd threshold value, then increase the resource of described application；Wherein, when increasing the resource of described application, if currently there is no available resources, then notify that manager creates new resources, and after new resources have created, increase the resource of described application；Or, if described performance data is less than default 4th threshold value, then reduce the resource of described application；Wherein, described the 3rd threshold value of presetting presets the 4th threshold value more than described.
6. the dynamic adjusting device of a resource, it is characterised in that described device specifically includes:
Obtain module, for for the resource for application configuration, it is thus achieved that the performance data that described resource is corresponding；
Prediction module, is used for utilizing described performance data, uses the performance data tendency situation in neural network prediction follow-up a period of time；
Adjusting module, for when prediction performance data tendency situation be: from the very first time, it was predicted that performance data more than preset first threshold value, then before the described very first time, increase the resource of described application；Or, from the second time, it was predicted that performance data less than default Second Threshold, and there is currently the idling-resource into described application configuration, then, before described second time, be reduced to the idling-resource of described application configuration；Wherein, described preset first threshold value is more than described default Second Threshold.
7. device according to claim 6, it is characterised in that
Described acquisition module, specifically in the process obtaining performance data corresponding to described resource, configuration monitoring agency in described resource, and receive the performance data that described resource that described monitoring agent reports is corresponding.
8. device according to claim 6, it is characterised in that
Described prediction module, specifically for utilizing described performance data, use in the process of the performance data tendency situation in neural network prediction follow-up a period of time, use neutral net that described performance data is trained, go forward side by side line parameter optimizing, to obtain the training pattern of described neutral net, and described training pattern is utilized to predict the performance data tendency situation in follow-up a period of time；Wherein, described neutral net specifically includes support vector machines neutral net.
9. device according to claim 6, it is characterised in that
Described adjusting module, specifically for, in increasing the process of resource of described application, if currently not had available resources, then notifying that manager created new resources before the described very first time, and after new resources have created, increase the resource of described application.
10. device according to claim 6, it is characterised in that
Described adjusting module, is additionally operable to, after obtaining the performance data that described resource is corresponding, if described performance data is more than default 3rd threshold value, then increase the resource of described application；When increasing the resource of described application, if currently there is no available resources, then notify that manager creates new resources, and after new resources have created, increase the resource of described application；Or, if described performance data is less than default 4th threshold value, then reduce the resource of described application；Described the 3rd threshold value of presetting presets the 4th threshold value more than described.
Priority Applications (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN201610006553.7A CN105760224A (en) | 2016-01-06 | 2016-01-06 | Dynamic resource adjustment method and device |
Applications Claiming Priority (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN201610006553.7A CN105760224A (en) | 2016-01-06 | 2016-01-06 | Dynamic resource adjustment method and device |
Publications (1)
| Publication Number | Publication Date |
|---|---|
| CN105760224A true CN105760224A (en) | 2016-07-13 |
Family
ID=56342351
Family Applications (1)
| Application Number | Title | Priority Date | Filing Date |
|---|---|---|---|
| CN201610006553.7A Pending CN105760224A (en) | 2016-01-06 | 2016-01-06 | Dynamic resource adjustment method and device |
Country Status (1)
| Country | Link |
|---|---|
| CN (1) | CN105760224A (en) |
Cited By (11)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN106775936A (en) * | 2016-11-30 | 2017-05-31 | 华为技术有限公司 | The management method and device of a kind of virtual machine |
| CN107196869A (en) * | 2017-07-14 | 2017-09-22 | 北京知道创宇信息技术有限公司 | The adaptive load balancing method, apparatus and system of Intrusion Detection based on host actual loading |
| CN108090025A (en) * | 2018-01-19 | 2018-05-29 | 福州瑞芯微电子股份有限公司 | The neutral net SOC chip of dynamic multichannel and its channel resource distribution method |
| WO2018121282A1 (en) * | 2016-12-26 | 2018-07-05 | 华为技术有限公司 | Data processing method, end device, cloud device, and end-cloud collaboration system |
| CN108984369A (en) * | 2018-07-13 | 2018-12-11 | 厦门美图移动科技有限公司 | Caton prediction method and device and mobile terminal |
| CN109032793A (en) * | 2018-07-11 | 2018-12-18 | Oppo广东移动通信有限公司 | Method, apparatus, terminal and the storage medium of resource distribution |
| CN109960581A (en) * | 2017-12-26 | 2019-07-02 | 广东欧珀移动通信有限公司 | Hardware resource configuration method, device, mobile terminal and storage medium |
| CN109995573A (en) * | 2018-01-02 | 2019-07-09 | 中国移动通信有限公司研究院 | Determine method, the computational resource allocation method and device for predicting the sample space of computing resource |
| CN109996247A (en) * | 2019-03-27 | 2019-07-09 | 中国电子科技集团公司信息科学研究院 | Networked resource concocting method, device, equipment and storage medium |
| CN111989653A (en) * | 2018-04-19 | 2020-11-24 | 微软技术许可有限责任公司 | Utilization based dynamic resource reallocation using two thresholds and a time window |
| CN114564313A (en) * | 2022-03-07 | 2022-05-31 | 中国农业银行股份有限公司 | Load adjustment method and device, electronic equipment and storage medium |
Citations (5)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN101442807A (en) * | 2008-12-30 | 2009-05-27 | 北京邮电大学 | Method and system for distribution of communication system resource |
| CN102279771A (en) * | 2011-09-02 | 2011-12-14 | 北京航空航天大学 | Method and system for adaptively allocating resources as required in virtualization environment |
| US20120173709A1 (en) * | 2011-01-05 | 2012-07-05 | Li Li | Seamless scaling of enterprise applications |
| CN103870341A (en) * | 2014-03-12 | 2014-06-18 | 汉柏科技有限公司 | Method and system of adjusting resources of virtual machine |
| CN103970587A (en) * | 2014-04-21 | 2014-08-06 | 华为技术有限公司 | Resource scheduling method, device and system |
-
2016
- 2016-01-06 CN CN201610006553.7A patent/CN105760224A/en active Pending
Patent Citations (5)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN101442807A (en) * | 2008-12-30 | 2009-05-27 | 北京邮电大学 | Method and system for distribution of communication system resource |
| US20120173709A1 (en) * | 2011-01-05 | 2012-07-05 | Li Li | Seamless scaling of enterprise applications |
| CN102279771A (en) * | 2011-09-02 | 2011-12-14 | 北京航空航天大学 | Method and system for adaptively allocating resources as required in virtualization environment |
| CN103870341A (en) * | 2014-03-12 | 2014-06-18 | 汉柏科技有限公司 | Method and system of adjusting resources of virtual machine |
| CN103970587A (en) * | 2014-04-21 | 2014-08-06 | 华为技术有限公司 | Resource scheduling method, device and system |
Cited By (18)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN106775936A (en) * | 2016-11-30 | 2017-05-31 | 华为技术有限公司 | The management method and device of a kind of virtual machine |
| CN106775936B (en) * | 2016-11-30 | 2020-06-26 | 华为技术有限公司 | Virtual machine management method and device |
| US11861499B2 (en) | 2016-12-26 | 2024-01-02 | Huawei Technologies Co., Ltd. | Method, terminal-side device, and cloud-side device for data processing and terminal-cloud collaboration system |
| WO2018121282A1 (en) * | 2016-12-26 | 2018-07-05 | 华为技术有限公司 | Data processing method, end device, cloud device, and end-cloud collaboration system |
| CN107196869B (en) * | 2017-07-14 | 2019-11-15 | 北京知道创宇信息技术股份有限公司 | The adaptive load balancing method, apparatus and system of Intrusion Detection based on host actual loading |
| CN107196869A (en) * | 2017-07-14 | 2017-09-22 | 北京知道创宇信息技术有限公司 | The adaptive load balancing method, apparatus and system of Intrusion Detection based on host actual loading |
| CN109960581A (en) * | 2017-12-26 | 2019-07-02 | 广东欧珀移动通信有限公司 | Hardware resource configuration method, device, mobile terminal and storage medium |
| CN109960581B (en) * | 2017-12-26 | 2021-06-01 | Oppo广东移动通信有限公司 | Hardware resource allocation method and device, mobile terminal and storage medium |
| CN109995573A (en) * | 2018-01-02 | 2019-07-09 | 中国移动通信有限公司研究院 | Determine method, the computational resource allocation method and device for predicting the sample space of computing resource |
| CN109995573B (en) * | 2018-01-02 | 2022-04-01 | 中国移动通信有限公司研究院 | Method for determining sample space for predicting computing resource, computing resource allocation method and device |
| CN108090025A (en) * | 2018-01-19 | 2018-05-29 | 福州瑞芯微电子股份有限公司 | The neutral net SOC chip of dynamic multichannel and its channel resource distribution method |
| CN108090025B (en) * | 2018-01-19 | 2021-08-27 | 瑞芯微电子股份有限公司 | Dynamic multichannel neural network SOC chip and channel resource allocation method thereof |
| CN111989653A (en) * | 2018-04-19 | 2020-11-24 | 微软技术许可有限责任公司 | Utilization based dynamic resource reallocation using two thresholds and a time window |
| CN109032793A (en) * | 2018-07-11 | 2018-12-18 | Oppo广东移动通信有限公司 | Method, apparatus, terminal and the storage medium of resource distribution |
| CN108984369A (en) * | 2018-07-13 | 2018-12-11 | 厦门美图移动科技有限公司 | Caton prediction method and device and mobile terminal |
| CN109996247B (en) * | 2019-03-27 | 2022-03-11 | 中国电子科技集团公司信息科学研究院 | Networked resource allocation method, device, equipment and storage medium |
| CN109996247A (en) * | 2019-03-27 | 2019-07-09 | 中国电子科技集团公司信息科学研究院 | Networked resource concocting method, device, equipment and storage medium |
| CN114564313A (en) * | 2022-03-07 | 2022-05-31 | 中国农业银行股份有限公司 | Load adjustment method and device, electronic equipment and storage medium |
Similar Documents
| Publication | Publication Date | Title |
|---|---|---|
| CN105760224A (en) | Dynamic resource adjustment method and device | |
| CN106776005B (en) | Resource management system and method for containerized application | |
| KR102137741B1 (en) | Dynamic thermal budget allocation for memory array | |
| US11061580B2 (en) | Storage device and controllers included in storage device | |
| US10430723B1 (en) | Storage system with machine learning based skew prediction | |
| CN104881325A (en) | Resource scheduling method and resource scheduling system | |
| CN104767689A (en) | Method of controlling network access amount and server | |
| CN101116310A (en) | Apparatus and method for autonomic adjustment of connection keep-alives | |
| US9641431B1 (en) | System and methods for utilization-based balancing of traffic to an information retrieval system | |
| CN108063784B (en) | Method, device and system for distributing application cluster resources in cloud environment | |
| US11163352B2 (en) | Multicore processor and method for dynamically adjusting a supply voltage and a clock speed | |
| DE112020002189T5 (en) | CONTAINER-BASED APPLICATIONS | |
| CN111669787B (en) | Resource allocation method and device based on time delay sensitive network slice | |
| CN115543577B (en) | Covariate-based Kubernetes resource scheduling optimization method, storage medium and device | |
| CN110018781B (en) | Disk flow control method and device and electronic equipment | |
| CN103699433A (en) | Method and system for performing dynamic adjustment on number of tasks in Hadoop platform | |
| CN103713935A (en) | Method and device for managing Hadoop cluster resources in online manner | |
| CN104243531A (en) | Data processing method, device and system | |
| CN116149846A (en) | Application performance optimization method and device, electronic equipment and storage medium | |
| CN113946450A (en) | Self-adaptive authorized polling load balancing system for K8S micro service framework | |
| CN103841129A (en) | Cloud computing resource information acquisition server, cloud computing resource information acquisition client and information processing method | |
| CN114035962A (en) | Resource allocation method and device, computer equipment and storage medium | |
| CN110688993B (en) | Spark operation-based computing resource determination method and device | |
| CN104571931A (en) | I/O (input/output) request combination scheduling system and method based on system resources | |
| US10289958B1 (en) | Storage system with machine learning based skew prediction |
Legal Events
| Date | Code | Title | Description |
|---|---|---|---|
| C06 | Publication | ||
| PB01 | Publication | ||
| C10 | Entry into substantive examination | ||
| SE01 | Entry into force of request for substantive examination | ||
| CB02 | Change of applicant information | ||
| CB02 | Change of applicant information |
Address after: 310052 Binjiang District Changhe Road, Zhejiang, China, No. 466, No. Applicant after: Xinhua three Technology Co., Ltd. Address before: 310052 Binjiang District Changhe Road, Zhejiang, China, No. 466, No. Applicant before: Huasan Communication Technology Co., Ltd. |
|
| RJ01 | Rejection of invention patent application after publication | ||
| RJ01 | Rejection of invention patent application after publication |
Application publication date: 20160713 |