CN109905255A - A kind of system for cloud computing method for predicting and device based on timing statistical sectional - Google Patents
A kind of system for cloud computing method for predicting and device based on timing statistical sectional Download PDFInfo
- Publication number
- CN109905255A CN109905255A CN201711286862.5A CN201711286862A CN109905255A CN 109905255 A CN109905255 A CN 109905255A CN 201711286862 A CN201711286862 A CN 201711286862A CN 109905255 A CN109905255 A CN 109905255A
- Authority
- CN
- China
- Prior art keywords
- sequence
- statistical
- time
- cloud computing
- virtual machine
- 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
- 238000004364 calculation method Methods 0.000 title claims abstract description 10
- 230000011218 segmentation Effects 0.000 claims description 14
- 238000000034 method Methods 0.000 claims description 12
- 230000007246 mechanism Effects 0.000 claims description 8
- 238000001514 detection method Methods 0.000 claims description 3
- 125000002015 acyclic group Chemical group 0.000 abstract description 2
- 238000007405 data analysis Methods 0.000 abstract description 2
- 230000008569 process Effects 0.000 description 3
- 238000002955 isolation Methods 0.000 description 2
- 230000008859 change Effects 0.000 description 1
- 230000002045 lasting effect Effects 0.000 description 1
- 230000007774 longterm Effects 0.000 description 1
- 230000005012 migration Effects 0.000 description 1
- 238000013508 migration Methods 0.000 description 1
- 230000004044 response Effects 0.000 description 1
- 230000009466 transformation Effects 0.000 description 1
- 230000007704 transition Effects 0.000 description 1
Landscapes
- Data Exchanges In Wide-Area Networks (AREA)
Abstract
The present invention relates to a kind of system for cloud computing method for predicting based on timing statistical sectional, resource occupation amount statistics at times is carried out to the virtual machine in system for cloud computing, Markov state model is established according to statistical result, the predicting network flow of subsequent period is carried out according to model.Compared with prior art, the present invention is for acyclic network behavior, time series data analysis and prediction are carried out by the result to network boundary bandwidth monitor, identify the discharge characteristic of different business and timing trend in network, and then a possibility that judging network " adjacent flow interference " hidden danger in following a period of time, to optimize resource adjusting strategies, avoid excessively frequently dispatching and adjusting.
Description
Technical field
The present invention relates to a kind of network flow prediction methods, more particularly, to a kind of cloud computing based on timing statistical sectional
Network flow prediction method and device.
Background technique
Since the business carried in cloud computing environment and application are many kinds of, and different application is for the demand of Internet resources
Different, particularly, in the cloud computing environment that nowadays mostly application coexists, a distinct issues are " adjacent flow interference "
(Noisy Neigbour) problem, since the isolation and scheduling of resource are incomplete, leads to some that is, while shared resource
It applies to generate adjacent and relevant other application when generating vast resources demand or generating a large amount of network flows and do
It disturbs, influences the stability of other application.
In order to solve adjacent flow interference on Internet resources to the influence of other users, current countermeasure mainly has
The modes such as empty machine scheduling, network scheduling, resource isolation, but these have regular hour expense, i.e., cannot accomplish to become network
The real-time response of change, while certain influence, such as virtual machine (vm) migration itself may also be generated to network performance itself and may require that
Certain network bandwidth, therefore it is unable to frequent progress adjustment.
Therefore, want in actual operations within the following certain time network flow and interference traffic conditions into
The certain prediction of row avoids excessively frequently dispatching and adjusting to optimize resource adjusting strategies.
Summary of the invention
It is an object of the present invention to overcome the above-mentioned drawbacks of the prior art and provide one kind to be counted based on timing
The system for cloud computing method for predicting and device of segmentation.
The purpose of the present invention can be achieved through the following technical solutions:
A kind of system for cloud computing method for predicting based on timing statistical sectional, to the virtual machine in system for cloud computing into
The resource occupation amount statistics of row at times, establishes Markov state model according to statistical result, carries out lower a period of time according to model
The predicting network flow of section.
The resource occupation amount includes flow and/or bandwidth usage.
Further, the method includes following below scheme:
Traffic statistics, the flow of virtual machine changes with time in statistical server node, obtains current-time sequence；
The sequence of unsegmented in window is carried out current-time sequence using sliding window mechanism by sequence segment
Segmentation makes every section of resource occupation amount statistical property difference be not less than given threshold, then window sliding to current point in time, after
It is continuous that the sequence of unsegmented is segmented；
Sequence statistic carries out the network bandwidth occupancy characteristic statistics based on segmentation, obtains each to the sequence being segmented
The holding time ratio in heterogeneous networks bandwidth section of the virtual machine within its corresponding period；
Piecewise prediction describes the state transfer spy of each user, each virtual machine according to the statistical nature in the period
Property, the hiding Markov model of building state transfer carries out the prediction of next period according to established model, predicts
Content includes duration and resource occupation amount.
In the sequence segment, using top-down sequence segment method, specifically: sequence is first divided into two sections, is made
Statistical property difference between two sections is maximum, the statistical property difference then to every section of progress recursive segmentation, between two sections
Reach setting range.
A kind of system for cloud computing volume forecasting device based on timing statistical sectional, comprising:
Flow statistical module, the flow of virtual machine changes with time in statistical server node, obtains current-time sequence
Column；
Sequence segment module, for current-time sequence, using sliding window mechanism, by the sequence of unsegmented in window
It is segmented, so that every section of resource occupation amount statistical property difference is not less than given threshold, then window sliding to current time
Point continues to be segmented the sequence of unsegmented；
Sequence statistic module carries out the network bandwidth occupancy characteristic statistics based on segmentation, obtains to the sequence being segmented
The holding time ratio in heterogeneous networks bandwidth section of each virtual machine within its corresponding period；
Piecewise prediction module describes the state transfer of each user, each virtual machine according to the statistical nature in the period
The hiding Markov model of characteristic, building state transfer carries out the prediction of next period, in advance according to established model
Surveying content includes duration and resource occupation amount.
The device further includes confidence interval detection module, carries out real-time statistical check, gained ratio to prediction result
To result feed back input into piecewise prediction module.
Compared with prior art, the present invention is for acyclic network behavior, by network boundary bandwidth monitor
As a result time series data analysis and prediction are carried out, identifies the discharge characteristic of different business and timing trend in network, and then sentence
It a possibility that network " adjacent flow interference " hidden danger in following a period of time of breaking, to optimize resource adjusting strategies, avoids
Excessively frequently scheduling and adjustment.
For current-time sequence, using sliding window mechanism and top-down sequence syncopation, realize dynamic
Model foundation process, the state variable to be differed greatly between each other can accurately find practical tenant's behavior to net
The influence stage of network flow and material time point improve the accuracy of model prediction.
Detailed description of the invention
Fig. 1 is the structure principle chart of the present embodiment prediction meanss；
Fig. 2 is the present embodiment time Series Processing algorithm flow.
Specific embodiment
The present invention is described in detail with specific embodiment below in conjunction with the accompanying drawings.The present embodiment is with technical solution of the present invention
Premised on implemented, the detailed implementation method and specific operation process are given, but protection scope of the present invention is not limited to
Following embodiments.
Embodiment
As shown in Figure 1, a kind of system for cloud computing volume forecasting device based on timing statistical sectional, comprising:
Flow statistical module 1 is deployed in each server node of cloud computing, each virtual machine in statistical server node
Transmitting-receiving flow changes with time, and obtains the accurate transformation period sequence of each virtual machine network bandwidth usage at any time；
Sequence segment module 2, for current-time sequence, using sliding window mechanism, by the sequence of unsegmented in window
It is segmented, so that every section of resource occupation amount statistical property difference is not less than given threshold, then window sliding to current time
Point continues to be segmented the sequence of unsegmented；
Sequence statistic module 3, based on the time slice learnt as a result, carrying out the network based on segmentation for time series
Bandwidth usage characteristic statistics obtain the holding time in heterogeneous networks bandwidth section of each virtual machine in its corresponding timeslice
Ratio；
Piecewise prediction module 4 describes each tenant, each virtual machine according to the network flow statistic feature in timeslice
State transfer characteristic, its hiding Markov model shifted using situation state is constructed, and according to this model prediction
Duration, state tag and the corresponding network flow statistic probability distribution of next time slice；
Confidence interval detection module 5 carries out real-time statistical check to prediction result, and gained comparison result feed back input is extremely
In piecewise prediction module 4, meanwhile, the assessment of the confidence interval based on existing model is provided for current prediction result.
A kind of system for cloud computing method for predicting based on timing statistical sectional, including following below scheme:
Traffic statistics, the flow of virtual machine changes with time in statistical server node, obtains current-time sequence；
The sequence of unsegmented in window is carried out current-time sequence using sliding window mechanism by sequence segment
Segmentation makes every section of resource occupation amount statistical property difference be not less than given threshold, then window sliding to current point in time, after
It is continuous that the sequence of unsegmented is segmented；Specifically, using top-down sequence segment method, specifically: first sequence is divided into
Two sections, keep the statistical property difference between two sections maximum, then to every section of progress recursive segmentation, the statistics between two sections is special
Sex differernce reaches setting range, and it is bright the time series of lasting variation can be divided into bandwidth occupancy statistical nature difference in real time in this way
Aobvious different phase can accurately find influence stage and material time point of the practical tenant's behavior to network flow；
Sequence statistic carries out the network bandwidth occupancy characteristic statistics based on segmentation, obtains each to the sequence being segmented
The holding time ratio in heterogeneous networks bandwidth section of the virtual machine within its corresponding period；
Piecewise prediction describes the state transfer spy of each user, each virtual machine according to the statistical nature in the period
Property, the hiding Markov model of building state transfer carries out the prediction of next period according to established model, predicts
Content includes duration and resource occupation amount.
As shown in Fig. 2, time Series Processing process provided by the present embodiment mainly includes the following steps: with method
1. for the time series of input, using sliding window mechanism, by the Sequence being segmented in sliding window from
It is rejected in window, window forward slip is read in into time series data；
2., using top-down window segmentation algorithm, being by window cutting for specific sliding window sometime
Two sections, so that the traffic statistics property difference between two sections is maximum.To every section of progress recurrence cutting after cutting, between two sections
Until flow difference characteristic is less than given threshold value
3. after the completion of cutting, in addition to final stage time series, before all split time sequences filings enter statistics
Information storage；Time window slides into current time sequence of points, re-starts step 2.
4. for the timeslice statistical information filed, settling time piece state transition model, to facilitate long-term forecast.
Claims (6)
1. a kind of system for cloud computing method for predicting based on timing statistical sectional, which is characterized in that in system for cloud computing
Virtual machine carry out at times resource occupation amount statistics, Markov state model is established according to statistical result, according to model
Carry out the predicting network flow of subsequent period.
2. a kind of system for cloud computing method for predicting based on timing statistical sectional according to claim 1, feature
It is, the resource occupation amount includes flow and/or bandwidth usage.
3. a kind of system for cloud computing method for predicting based on timing statistical sectional according to claim 1, feature
It is, including following below scheme:
Traffic statistics, the flow of virtual machine changes with time in statistical server node, obtains current-time sequence；
The sequence of unsegmented in window is segmented current-time sequence using sliding window mechanism by sequence segment,
Make every section of resource occupation amount statistical property difference not less than given threshold, then window sliding to current point in time, continues pair
The sequence of unsegmented is segmented；
Sequence statistic carries out the network bandwidth occupancy characteristic statistics based on segmentation to the sequence being segmented, and obtains each virtual
The holding time ratio in heterogeneous networks bandwidth section of the machine within its corresponding period；
Piecewise prediction describes the state transfer characteristic of each user, each virtual machine, structure according to the statistical nature in the period
The hiding Markov model for building state transfer carries out the prediction of next period, predictive content packet according to established model
Include duration and resource occupation amount.
4. a kind of system for cloud computing method for predicting based on timing statistical sectional according to claim 3, feature
It is, in the sequence segment, using top-down sequence segment method, specifically: sequence is first divided into two sections, makes two sections
Between statistical property difference it is maximum, then to every section of progress recursive segmentation, the statistical property difference between two sections reaches
Setting range.
5. a kind of system for cloud computing volume forecasting device based on timing statistical sectional characterized by comprising
Flow statistical module, the flow of virtual machine changes with time in statistical server node, obtains current-time sequence；
The sequence of unsegmented in window is carried out current-time sequence using sliding window mechanism by sequence segment module
Segmentation makes every section of resource occupation amount statistical property difference be not less than given threshold, then window sliding to current point in time, after
It is continuous that the sequence of unsegmented is segmented；
Sequence statistic module carries out the network bandwidth occupancy characteristic statistics based on segmentation, obtains each to the sequence being segmented
The holding time ratio in heterogeneous networks bandwidth section of the virtual machine within its corresponding period；
Piecewise prediction module describes the state transfer spy of each user, each virtual machine according to the statistical nature in the period
Property, the hiding Markov model of building state transfer carries out the prediction of next period according to established model, predicts
Content includes duration and resource occupation amount.
6. a kind of system for cloud computing volume forecasting device based on timing statistical sectional according to claim 5, feature
It is, further includes confidence interval detection module, real-time statistical check, gained comparison result feed back input is carried out to prediction result
Into piecewise prediction module.
Priority Applications (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN201711286862.5A CN109905255A (en) | 2017-12-07 | 2017-12-07 | A kind of system for cloud computing method for predicting and device based on timing statistical sectional |
Applications Claiming Priority (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN201711286862.5A CN109905255A (en) | 2017-12-07 | 2017-12-07 | A kind of system for cloud computing method for predicting and device based on timing statistical sectional |
Publications (1)
| Publication Number | Publication Date |
|---|---|
| CN109905255A true CN109905255A (en) | 2019-06-18 |
Family
ID=66939441
Family Applications (1)
| Application Number | Title | Priority Date | Filing Date |
|---|---|---|---|
| CN201711286862.5A Pending CN109905255A (en) | 2017-12-07 | 2017-12-07 | A kind of system for cloud computing method for predicting and device based on timing statistical sectional |
Country Status (1)
| Country | Link |
|---|---|
| CN (1) | CN109905255A (en) |
Cited By (5)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN110659681A (en) * | 2019-09-17 | 2020-01-07 | 上海仪电（集团）有限公司中央研究院 | Time sequence data prediction system and method based on pattern recognition |
| CN113538026A (en) * | 2020-04-15 | 2021-10-22 | 北京京东振世信息技术有限公司 | Traffic calculation method and device |
| CN113673822A (en) * | 2021-07-15 | 2021-11-19 | 微梦创科网络科技（中国）有限公司 | Elastic scheduling method and system |
| CN114070757A (en) * | 2021-11-15 | 2022-02-18 | 南方电网数字电网研究院有限公司 | Data traffic change monitoring method for cloud computing management control platform |
| CN115037642A (en) * | 2022-03-30 | 2022-09-09 | 武汉烽火技术服务有限公司 | Method and device for identifying flow bottleneck |
Citations (1)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN103036974A (en) * | 2012-12-13 | 2013-04-10 | 广东省电信规划设计院有限公司 | Cloud computing resource scheduling method and system based on hidden markov model |
-
2017
- 2017-12-07 CN CN201711286862.5A patent/CN109905255A/en active Pending
Patent Citations (1)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN103036974A (en) * | 2012-12-13 | 2013-04-10 | 广东省电信规划设计院有限公司 | Cloud computing resource scheduling method and system based on hidden markov model |
Non-Patent Citations (2)
| Title |
|---|
| 覃征等: "时间序列数据的稳健最优分割方法", 《西安交通大学学报》 * |
| 黄雄波: "非平稳时序数据的分段辨识及其递推算法", 《计算机系统应用》 * |
Cited By (9)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN110659681A (en) * | 2019-09-17 | 2020-01-07 | 上海仪电（集团）有限公司中央研究院 | Time sequence data prediction system and method based on pattern recognition |
| CN113538026A (en) * | 2020-04-15 | 2021-10-22 | 北京京东振世信息技术有限公司 | Traffic calculation method and device |
| CN113538026B (en) * | 2020-04-15 | 2023-11-03 | 北京京东振世信息技术有限公司 | Service amount calculation method and device |
| CN113673822A (en) * | 2021-07-15 | 2021-11-19 | 微梦创科网络科技（中国）有限公司 | Elastic scheduling method and system |
| CN113673822B (en) * | 2021-07-15 | 2023-08-11 | 微梦创科网络科技（中国）有限公司 | Elastic scheduling method and system |
| CN114070757A (en) * | 2021-11-15 | 2022-02-18 | 南方电网数字电网研究院有限公司 | Data traffic change monitoring method for cloud computing management control platform |
| CN114070757B (en) * | 2021-11-15 | 2023-08-08 | 南方电网数字电网研究院有限公司 | Data flow change monitoring method for cloud computing management control platform |
| CN115037642A (en) * | 2022-03-30 | 2022-09-09 | 武汉烽火技术服务有限公司 | Method and device for identifying flow bottleneck |
| CN115037642B (en) * | 2022-03-30 | 2023-11-21 | 武汉烽火技术服务有限公司 | Method and device for identifying flow bottleneck |
Similar Documents
| Publication | Publication Date | Title |
|---|---|---|
| CN109905255A (en) | A kind of system for cloud computing method for predicting and device based on timing statistical sectional | |
| CN112559007B (en) | Parameter updating method and device of multitask model and electronic equipment | |
| US11823058B2 (en) | Data valuation using reinforcement learning | |
| CN110633594A (en) | Target detection method and device | |
| US20220284067A1 (en) | Method for pushing information, electronic device | |
| US20220053335A1 (en) | Method for detecting an abnormal device, device and storage medium | |
| CN115941322B (en) | Attack detection method, device, equipment and storage medium based on artificial intelligence | |
| CN110363427A (en) | Model quality evaluation method and apparatus | |
| CN114449343A (en) | Video processing method, device, equipment and storage medium | |
| CN112784734A (en) | Video identification method and device, electronic equipment and storage medium | |
| CN113099475A (en) | Network quality detection method and device, electronic equipment and readable storage medium | |
| CN113408306A (en) | Translation method, training method, device, equipment and storage medium of classification model | |
| CN113627536A (en) | Model training method, video classification method, device, equipment and storage medium | |
| CN113159934A (en) | Method and system for predicting passenger flow of network, electronic equipment and storage medium | |
| CN107291805B (en) | Method and device for controlling information pushing | |
| CN114090601B (en) | Data screening method, device, equipment and storage medium | |
| US10972921B2 (en) | Wireless signal strength optimizer | |
| CN114581732A (en) | Image processing and model training method, device, equipment and storage medium | |
| CN113904943A (en) | Account detection method and device, electronic equipment and storage medium | |
| CN113205189A (en) | Prediction model training method, prediction method and prediction device | |
| CN113132479A (en) | Traffic switching method, traffic switching apparatus, traffic model generating method, traffic model generating apparatus, traffic model generating | |
| JP6680663B2 (en) | Information processing apparatus, information processing method, prediction model generation apparatus, prediction model generation method, and program | |
| CN116662509A (en) | Open domain question-answering implementation method, device and equipment of large-scale language model | |
| CN113874888A (en) | Information processing apparatus, generation method, and generation program | |
| CN115203873A (en) | Topological relation construction method, device, equipment and medium applied to power distribution network |
Legal Events
| Date | Code | Title | Description |
|---|---|---|---|
| PB01 | Publication | ||
| PB01 | Publication | ||
| SE01 | Entry into force of request for substantive examination | ||
| SE01 | Entry into force of request for substantive examination | ||
| RJ01 | Rejection of invention patent application after publication |
Application publication date: 20190618 |
|
| RJ01 | Rejection of invention patent application after publication |