CN107292388A - A kind of Forecasting Methodology and system of the hot spot data based on neutral net - Google Patents
A kind of Forecasting Methodology and system of the hot spot data based on neutral net Download PDFInfo
- Publication number
- CN107292388A CN107292388A CN201710500159.3A CN201710500159A CN107292388A CN 107292388 A CN107292388 A CN 107292388A CN 201710500159 A CN201710500159 A CN 201710500159A CN 107292388 A CN107292388 A CN 107292388A
- Authority
- CN
- China
- Prior art keywords
- data
- preset time
- network model
- hot spot
- neural network
- 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
- G06N—COMPUTING ARRANGEMENTS BASED ON SPECIFIC COMPUTATIONAL MODELS
- G06N3/00—Computing arrangements based on biological models
- G06N3/02—Neural networks
- G06N3/08—Learning methods
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06F—ELECTRIC DIGITAL DATA PROCESSING
- G06F12/00—Accessing, addressing or allocating within memory systems or architectures
- G06F12/02—Addressing or allocation; Relocation
- G06F12/08—Addressing or allocation; Relocation in hierarchically structured memory systems, e.g. virtual memory systems
- G06F12/0802—Addressing of a memory level in which the access to the desired data or data block requires associative addressing means, e.g. caches
- G06F12/0877—Cache access modes
Landscapes
- Engineering & Computer Science (AREA)
- Theoretical Computer Science (AREA)
- Physics & Mathematics (AREA)
- General Physics & Mathematics (AREA)
- General Engineering & Computer Science (AREA)
- General Health & Medical Sciences (AREA)
- Computing Systems (AREA)
- Computational Linguistics (AREA)
- Data Mining & Analysis (AREA)
- Evolutionary Computation (AREA)
- Biomedical Technology (AREA)
- Molecular Biology (AREA)
- Biophysics (AREA)
- Artificial Intelligence (AREA)
- Life Sciences & Earth Sciences (AREA)
- Mathematical Physics (AREA)
- Software Systems (AREA)
- Health & Medical Sciences (AREA)
- Management, Administration, Business Operations System, And Electronic Commerce (AREA)
- Feedback Control In General (AREA)
Abstract
A kind of Forecasting Methodology for hot spot data based on neutral net that the application is provided, history access record of the user in preset time period is obtained first is used as monitoring data collection, then integrated using the monitoring data as training data, multilayer neural network model is trained, until the accuracy rate of the multilayer neural network model prediction hot spot data is more than threshold value, target multi layer neural network model is obtained;Multiple data finally currently stored to storage system are traveled through, based on the target multi layer neural network model, and whether judge the currently stored each data of the storage system successively is hot spot data in next preset time unit.History access record of the invention according to user, the data that future, user may access effectively are predicted, the hot spot data of future time unit is loaded in advance, and unload the non-thermal point data of future time unit in caching in advance, user experience is improved, systematic function is improved.
Description
Technical field
The present invention relates to technical field of data storage, more particularly to a kind of hot spot data based on neutral net
Forecasting Methodology and system.
Background technology
Caching loading and hot spot data statistics all have hysteresis quality in modern memory systems, and typically all data are accessed
Caching can just once or be repeatedly loaded into afterwards, just will appreciate that this is data focus, can just carry out accelerated loading, influence
The overall performance and Consumer's Experience of system.
There are some old hot spot datas not discharge caching slowly in addition, these old hot spot datas are due in warp of past period
It is often accessed so be loaded into caching, but in the current and future period, these old hot spot datas it is no longer often accessed but
It is to remain stored in caching, causes caching crowded, new hot spot data can not be added to caching, influences carrying for systematic function
Rise.
The content of the invention
In view of this, the invention provides a kind of Forecasting Methodology of hot spot data based on neutral net and device, it is based on
Multilayer neural network model presets the hot spot data of future time, improves systematic function and the usage experience of user.
In order to realize foregoing invention purpose, the specific technical scheme that the present invention is provided is as follows:
A kind of Forecasting Methodology of the hot spot data based on neutral net, including:
Monitoring data collection is obtained, the monitoring data collection have recorded in preset time period in units of preset time unit to be used
The access data at family;
Integrated using the monitoring data as training data, multilayer neural network model is trained, until multilayer god
The accuracy rate for predicting hot spot data through network model is more than threshold value, obtains target multi layer neural network model;
The multiple data currently stored to storage system are traveled through, based on the target multi layer neural network model, according to
Whether the secondary each data for judging that the storage system is currently stored are hot spot data in next preset time unit.
It is preferred that, in each data for judging that the storage system is currently stored successively in next preset time unit
After whether being hot spot data, methods described also includes:
According to predicting the outcome for the target multi layer neural network model, each next preset time unit is judged successively
Whether hot spot data has been loaded into caching;
If it is not, the hot spot data that next preset time unit is not loaded into caching is loaded into caching.
It is preferred that, in each data for judging that the storage system is currently stored successively in next preset time unit
After whether being hot spot data, methods described also includes:
According to predicting the outcome for the target multi layer neural network model, judge whether the data in caching are next successively
The hot spot data of preset time unit;
If it is not, to not being that the data of next preset time unit are unloaded in caching.
It is preferred that, the acquisition monitoring data collection, including:
In the preset time period, user accesses data in each preset time unit is obtained respectively and is deposited described
Position and visitation frequency in storage system;
According to the Data Position and visitation frequency in each preset time unit, monitoring data collection is generated.
It is preferred that, the multiple data currently stored to storage system are traveled through, based on target multi layer nerve
Whether network model, it is focus number in next preset time unit that the currently stored each data of the storage system are judged successively
According to, including:
According to traversal order, the time parameter of the currently stored each data of storage system and location parameter are inputted successively
The target multi layer neural network model, the time parameter of each data is next preset time unit, each number
According to location parameter be relevant position within the storage system;
Obtain the output result of the target multi layer neural network model corresponding with each data, the output
As a result to access or not accessing, when output result is accesses, the corresponding data are hot spot data.
A kind of forecasting system of the hot spot data based on neutral net, including:
Acquiring unit, for obtaining monitoring data collection, the monitoring data collection have recorded in units of preset time unit
The access data of user in preset time period;
Training unit, for being integrated using the monitoring data as training data, is trained, directly to multilayer neural network model
Accuracy rate to the multilayer neural network model prediction hot spot data is more than threshold value, obtains target multi layer neural network model;
Predicting unit, is traveled through for multiple data currently stored to storage system, based on target multi layer god
Whether through network model, it is focus in next preset time unit that the currently stored each data of the storage system are judged successively
Data.
It is preferred that, the system also includes:
First judging unit, for predicting the outcome according to the target multi layer neural network model, judges each successively
Whether the hot spot data of next preset time unit has been loaded into caching;If it is not, next preset time unit is not loaded into slow
The hot spot data deposited is loaded into caching.
It is preferred that, the system also includes:
Second judging unit, for predicting the outcome according to the target multi layer neural network model, judges caching successively
In data whether be next preset time unit hot spot data;
If it is not, to not being that the data of next preset time unit are unloaded in caching.
It is preferred that, the acquiring unit includes:
First obtains subelement, in the preset time period, obtaining respectively in each preset time unit
User accesses data position within the storage system and visitation frequency;
Subelement is generated, for according to the Data Position and visitation frequency in each preset time unit, generation prison
Superintend and direct data set.
It is preferred that, the predicting unit includes:
Subelement is inputted, for according to traversal order, successively joining the time of the currently stored each data of storage system
Number and location parameter input the target multi layer neural network model, and the time parameter of each data is next preset time
Unit, the location parameter of each data is relevant position within the storage system;
Second obtains subelement, for obtaining the target multi layer neural network model corresponding with each data
Output result, the output result is accesses or not accessed, and when output result is accesses, the corresponding data are focus
Data.
Relative to prior art, beneficial effects of the present invention are as follows:
A kind of Forecasting Methodology for hot spot data based on neutral net that the present invention is provided, obtains user when default first
Between history access record in section as monitoring data collection, then integrated using the monitoring data as training data, to multilayer nerve
Network model is trained, and until the accuracy rate of the multilayer neural network model prediction hot spot data is more than threshold value, obtains mesh
Mark multilayer neural network model;Multiple data finally currently stored to storage system are traveled through, based on the target multi layer
Whether neural network model, it is heat in next preset time unit that the currently stored each data of the storage system are judged successively
Point data.The present invention effectively predicts the data that future, user may access according to the history access record of user, adds in advance
The hot spot data of future time unit is carried, when making user's actual access hot spot data, the hot spot data is already loaded into slow
Deposit, quickly to the user feedback user data to be accessed, improve user experience.Meanwhile, under unloading in advance in caching
The non-thermal point data of one chronomere, improves the utilization ratio of caching, improves systematic function.
Brief description of the drawings
In order to illustrate more clearly about the embodiment of the present invention or technical scheme of the prior art, below will be to embodiment or existing
There is the accompanying drawing used required in technology description to be briefly described, it should be apparent that, drawings in the following description are only this
The embodiment of invention, for those of ordinary skill in the art, on the premise of not paying creative work, can also basis
The accompanying drawing of offer obtains other accompanying drawings.
Fig. 1 is a kind of Forecasting Methodology flow chart of the hot spot data based on neutral net disclosed in the embodiment of the present invention;
Fig. 2 is the Forecasting Methodology flow chart of another hot spot data based on neutral net disclosed in the embodiment of the present invention;
Fig. 3 is a kind of forecasting system structural representation of the hot spot data based on neutral net disclosed in the embodiment of the present invention
Figure.
Embodiment
Below in conjunction with the accompanying drawing in the embodiment of the present invention, the technical scheme in the embodiment of the present invention is carried out clear, complete
Site preparation is described, it is clear that described embodiment is only a part of embodiment of the invention, rather than whole embodiments.It is based on
Embodiment in the present invention, it is every other that those of ordinary skill in the art are obtained under the premise of creative work is not made
Embodiment, belongs to the scope of protection of the invention.
Referring to Fig. 1, present embodiment discloses a kind of Forecasting Methodology of the hot spot data based on neutral net, specifically including
Following steps:
S101:Obtain monitoring data collection;
The monitoring data collection have recorded the access data of user in preset time period in units of preset time unit.
The preset time period is certain time before current time, can be one week, one month etc..
Preset time unit is the data storage unit of monitoring data collection, can be the chronomeres such as minute.
It is preferred that, S101 specific implementation procedure is as follows:
In the preset time period, user accesses data in each preset time unit is obtained respectively and is deposited described
Position and visitation frequency in storage system;
According to the Data Position and visitation frequency in each preset time unit, monitoring data collection is generated.
It is understood that the monitoring data collection contains the institute of each preset time unit in the preset time period
There are the data being accessed by the user.
It should be noted that the corresponding element of time of the monitoring data intensive data includes:Year, month, day, week, when,
Minute, second, whether festivals or holidays.If the preset time period is one week, year, the moon day need not can be inputted, if the preset time list
Position be minute, then the second may not necessarily input, can so reduce amount of calculation.
The corresponding position of the monitoring data intensive data is the data in some disk array cabinet, certain block hard disk, some LBA
(Logical Block Address, logical block addresses) position, and the volume information and host main frames letter that the data belong to
Breath.
S102:Integrated using the monitoring data as training data, multilayer neural network model is trained, until described many
The accuracy rate of layer Neural Network model predictive hot spot data is more than threshold value, obtains target multi layer neural network model;
Multilayer neural network model is trained, mainly the adjustment to multilayer neural network Model Weight.
In training process, the input data of multilayer neural network model is time parameter and the position of monitoring data intensive data
Put parameter, i.e. the corresponding access time of monitoring data intensive data (in units of preset time unit) and within the storage system
Position, the output data of multilayer neural network model is accesses and not accessed.
It is by judging the output result of multilayer neural network model with the actual access situation of monitoring data intensive data
It is no consistent, it is possible to determine that the predictablity rate of multilayer neural network model.That is, when to multilayer neural network mode input supervise number
According to intensive data S time parameter and location parameter when, the multilayer neural network model output result for access, if actual feelings
Data S is accessed in the corresponding preset time unit of time parameter in condition, then the multilayer neural network model prediction is accurate,
Conversely, the then multilayer neural network model prediction mistake.
When the accuracy rate of the multilayer neural network model prediction hot spot data is more than threshold value, stop to multilayer god
Training through network model, obtains target multi layer neural network model.The threshold value can be set in advance according to actual needs
It is fixed.
It should also be noted that, when current goal multilayer neural network model can not accurately to future time section focus
, it is necessary to reacquire new monitoring data collection when data are predicted, multilayer neural network model is trained, obtains new
Target multi layer neural network model.It is understood that the different no target multi layer neutral nets of monitoring data collection correspondence
Model.
S103:The multiple data currently stored to storage system are traveled through, based on the target multi layer neutral net mould
Whether type, it is hot spot data in next preset time unit that the currently stored each data of the storage system are judged successively.
It is preferred that, S103 specific executive mode is:
According to traversal order, the time parameter of the currently stored each data of storage system and location parameter are inputted successively
The target multi layer neural network model, the time parameter of each data is next preset time unit, each number
According to location parameter be relevant position within the storage system;
Obtain the output result of the target multi layer neural network model corresponding with each data, the output
As a result to access or not accessing, when output result is accesses, the corresponding data are hot spot data.
It should be noted that the present embodiment determines the target multi layer neural network model output result for the data accessed
Justice is hot spot data.
The traversal order can be any one traversal order, certainly, for optimization angle, can be pre-designed traversal
Sequentially, to improve traversal speed.
As a kind of preferred real-time proposals of the present embodiment, referring to Fig. 2, after S103, methods described also includes:
S104:According to predicting the outcome for the target multi layer neural network model, each next preset time is judged successively
Whether the hot spot data of unit has been loaded into caching;If it is not, next preset time unit not to be loaded into the hot spot data of caching
It is loaded into caching.
It should be noted that need not add again if the hot spot data of next preset time unit is already loaded into caching
Carry.
S105:According to predicting the outcome for the target multi layer neural network model, judge whether is data in caching successively
For the hot spot data of next preset time unit;If it is not, to not being that the data of next preset time unit are unloaded in caching.
It is understood that when the hot spot data that the data in caching are next preset time unit, then need not unload
Data.
A kind of Forecasting Methodology for hot spot data based on neutral net that the present embodiment is provided, obtains user default first
Then history access record in period is integrated as training data as monitoring data collection using the monitoring data, to multilayer god
It is trained, until the accuracy rate of the multilayer neural network model prediction hot spot data is more than threshold value, obtains through network model
Target multi layer neural network model;Multiple data finally currently stored to storage system are traveled through, many based on the target
Layer neural network model, judge successively the currently stored each data of the storage system next preset time unit whether be
Hot spot data.The present embodiment is effectively predicted the data that future, user may access, carried according to the history access record of user
The hot spot data of preceding loading future time unit, when making user's actual access hot spot data, the hot spot data has been loaded
To caching, quickly to the user feedback user data to be accessed, user experience is improved.Meanwhile, unloading in advance is cached
The non-thermal point data of middle future time unit, improves the utilization ratio of caching, improves systematic function.
Based on the Forecasting Methodology of the hot spot data based on neutral net disclosed in above-described embodiment, referring to Fig. 3, this implementation
Example correspondence discloses a kind of forecasting system of the hot spot data based on neutral net, including:
Acquiring unit 101, for obtaining monitoring data collection, the monitoring data collection is recorded in units of preset time unit
The access data of user in preset time period;
It is preferred that, the acquiring unit 101 includes:
First obtains subelement, in the preset time period, obtaining respectively in each preset time unit
User accesses data position within the storage system and visitation frequency;
Subelement is generated, for according to the Data Position and visitation frequency in each preset time unit, generation prison
Superintend and direct data set.
Training unit 102, for being integrated using the monitoring data as training data, is instructed to multilayer neural network model
Practice, until the accuracy rate of the multilayer neural network model prediction hot spot data is more than threshold value, obtain target multi layer neutral net
Model;
Predicting unit 103, is traveled through for multiple data currently stored to storage system, based on the target multi layer
Whether neural network model, it is heat in next preset time unit that the currently stored each data of the storage system are judged successively
Point data.
It is preferred that, the predicting unit 103 includes:
Subelement is inputted, for according to traversal order, successively joining the time of the currently stored each data of storage system
Number and location parameter input the target multi layer neural network model, and the time parameter of each data is next preset time
Unit, the location parameter of each data is relevant position within the storage system;
Second obtains subelement, for obtaining the target multi layer neural network model corresponding with each data
Output result, the output result is accesses or not accessed, and when output result is accesses, the corresponding data are focus
Data.
It is preferred that, the system also includes:
First judging unit, for predicting the outcome according to the target multi layer neural network model, judges each successively
Whether the hot spot data of next preset time unit has been loaded into caching;If it is not, next preset time unit is not loaded into slow
The hot spot data deposited is loaded into caching.
Second judging unit, for predicting the outcome according to the target multi layer neural network model, judges caching successively
In data whether be next preset time unit hot spot data;If it is not, to not being next preset time unit in caching
Data are unloaded.
A kind of forecasting system for hot spot data based on neutral net that the present embodiment is provided, obtains user default first
Then history access record in period is integrated as training data as monitoring data collection using the monitoring data, to multilayer god
It is trained, until the accuracy rate of the multilayer neural network model prediction hot spot data is more than threshold value, obtains through network model
Target multi layer neural network model;Multiple data finally currently stored to storage system are traveled through, many based on the target
Layer neural network model, judge successively the currently stored each data of the storage system next preset time unit whether be
Hot spot data.The present embodiment is effectively predicted the data that future, user may access, carried according to the history access record of user
The hot spot data of preceding loading future time unit, when making user's actual access hot spot data, the hot spot data has been loaded
To caching, quickly to the user feedback user data to be accessed, user experience is improved.Meanwhile, unloading in advance is cached
The non-thermal point data of middle future time unit, improves the utilization ratio of caching, improves systematic function.
The foregoing description of the disclosed embodiments, enables professional and technical personnel in the field to realize or using the present invention.
A variety of modifications to these embodiments will be apparent for those skilled in the art, as defined herein
General Principle can be realized in other embodiments without departing from the spirit or scope of the present invention.Therefore, it is of the invention
The embodiments shown herein is not intended to be limited to, and is to fit to and principles disclosed herein and features of novelty phase one
The most wide scope caused.
Claims (10)
1. a kind of Forecasting Methodology of the hot spot data based on neutral net, it is characterised in that including:
Monitoring data collection is obtained, the monitoring data collection have recorded user in preset time period in units of preset time unit
Access data;
Integrated using the monitoring data as training data, multilayer neural network model is trained, until the multilayer nerve net
The accuracy rate of network model prediction hot spot data is more than threshold value, obtains target multi layer neural network model;
The multiple data currently stored to storage system are traveled through, and based on the target multi layer neural network model, are sentenced successively
Whether the currently stored each data of the storage system of breaking are hot spot data in next preset time unit.
2. according to the method described in claim 1, it is characterised in that judge that the storage system is currently stored successively described
Each data are after whether next preset time unit is hot spot data, and methods described also includes:
According to predicting the outcome for the target multi layer neural network model, the focus of each next preset time unit is judged successively
Whether data have been loaded into caching;
If it is not, the hot spot data that next preset time unit is not loaded into caching is loaded into caching.
3. according to the method described in claim 1, it is characterised in that judge that the storage system is currently stored successively described
Each data are after whether next preset time unit is hot spot data, and methods described also includes:
Whether according to predicting the outcome for the target multi layer neural network model, it is next default that the data in caching are judged successively
The hot spot data of chronomere;
If it is not, to not being that the data of next preset time unit are unloaded in caching.
4. according to the method described in claim 1, it is characterised in that the acquisition monitoring data collection, including:
In the preset time period, user accesses data in each preset time unit is obtained respectively is in the storage
Position and visitation frequency in system;
According to the Data Position and visitation frequency in each preset time unit, monitoring data collection is generated.
5. according to the method described in claim 1, it is characterised in that the multiple data currently stored to storage system are carried out
Traversal, based on the target multi layer neural network model, judges the currently stored each data of the storage system under successively
Whether one preset time unit is hot spot data, including:
According to traversal order, the time parameter of the currently stored each data of storage system and location parameter are inputted successively described
Target multi layer neural network model, the time parameter of each data is next preset time unit, each data
Location parameter is relevant position within the storage system;
Obtain the output result of the target multi layer neural network model corresponding with each data, the output result
To access or not accessing, when output result is accesses, the corresponding data are hot spot data.
6. a kind of forecasting system of the hot spot data based on neutral net, it is characterised in that including:
Acquiring unit, for obtaining monitoring data collection, the monitoring data collection have recorded default in units of preset time unit
The access data of user in period;
Training unit, for being integrated using the monitoring data as training data, is trained, Zhi Daosuo to multilayer neural network model
The accuracy rate for stating multilayer neural network model prediction hot spot data is more than threshold value, obtains target multi layer neural network model;
Predicting unit, is traveled through for multiple data currently stored to storage system, based on the target multi layer nerve net
Whether network model, it is focus number in next preset time unit that the currently stored each data of the storage system are judged successively
According to.
7. system according to claim 6, it is characterised in that the system also includes:
First judging unit, for predicting the outcome according to the target multi layer neural network model, judges each next successively
Whether the hot spot data of preset time unit has been loaded into caching;If it is not, next preset time unit is not loaded into caching
Hot spot data is loaded into caching.
8. system according to claim 6, it is characterised in that the system also includes:
Second judging unit, for predicting the outcome according to the target multi layer neural network model, judges in caching successively
Data whether be next preset time unit hot spot data;If it is not, to be not in caching next preset time unit data
Unloaded.
9. system according to claim 6, it is characterised in that the acquiring unit includes:
First obtains subelement, in the preset time period, user in each preset time unit to be obtained respectively
Access data position within the storage system and visitation frequency;
Subelement is generated, for according to the Data Position and visitation frequency in each preset time unit, generation supervision number
According to collection.
10. system according to claim 6, it is characterised in that the predicting unit includes:
Input subelement, for according to traversal order, successively by the time parameter of the currently stored each data of storage system and
Location parameter inputs the target multi layer neural network model, and the time parameter of each data is next preset time list
Position, the location parameter of each data is relevant position within the storage system;
Second obtains subelement, for obtaining the defeated of the target multi layer neural network model corresponding with each data
Go out result, the output result is accesses or do not accessed, and when output result is accesses, the corresponding data are focus number
According to.
Priority Applications (1)
Application Number | Priority Date | Filing Date | Title |
---|---|---|---|
CN201710500159.3A CN107292388A (en) | 2017-06-27 | 2017-06-27 | A kind of Forecasting Methodology and system of the hot spot data based on neutral net |
Applications Claiming Priority (1)
Application Number | Priority Date | Filing Date | Title |
---|---|---|---|
CN201710500159.3A CN107292388A (en) | 2017-06-27 | 2017-06-27 | A kind of Forecasting Methodology and system of the hot spot data based on neutral net |
Publications (1)
Publication Number | Publication Date |
---|---|
CN107292388A true CN107292388A (en) | 2017-10-24 |
Family
ID=60098230
Family Applications (1)
Application Number | Title | Priority Date | Filing Date |
---|---|---|---|
CN201710500159.3A Pending CN107292388A (en) | 2017-06-27 | 2017-06-27 | A kind of Forecasting Methodology and system of the hot spot data based on neutral net |
Country Status (1)
Country | Link |
---|---|
CN (1) | CN107292388A (en) |
Cited By (7)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
CN107728952A (en) * | 2017-10-31 | 2018-02-23 | 郑州云海信息技术有限公司 | A kind of prediction type data migration method and system |
CN108182471A (en) * | 2018-01-24 | 2018-06-19 | 上海岳芯电子科技有限公司 | A kind of convolutional neural networks reasoning accelerator and method |
CN108446340A (en) * | 2018-03-02 | 2018-08-24 | 哈尔滨工业大学(威海) | A kind of user's hot spot data access prediction technique towards mass small documents |
CN108462605A (en) * | 2018-02-06 | 2018-08-28 | 国家电网公司 | A kind of prediction technique and device of data |
CN112734126A (en) * | 2021-01-18 | 2021-04-30 | 武汉烽火技术服务有限公司 | Hot spot prediction method, device and equipment and readable storage medium |
CN113076339A (en) * | 2021-03-18 | 2021-07-06 | 北京沃东天骏信息技术有限公司 | Data caching method, device, equipment and storage medium |
CN117270794A (en) * | 2023-11-22 | 2023-12-22 | 成都大成均图科技有限公司 | Redis-based data storage method, medium and device |
Citations (5)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
CN103092867A (en) * | 2011-11-03 | 2013-05-08 | 中国移动通信集团甘肃有限公司 | Method and system for managing data, and data analyzing device |
CN103218416A (en) * | 2013-03-27 | 2013-07-24 | 华为技术有限公司 | Method, device and system for loading database |
CN104134103A (en) * | 2014-07-30 | 2014-11-05 | 中国石油天然气股份有限公司 | Method for predicating energy consumption of hot oil pipeline through corrected BP neural network model |
CN105635968A (en) * | 2016-01-18 | 2016-06-01 | 广州杰赛科技股份有限公司 | Hotspot area identification method based on time unit and predication method and device |
CN106709068A (en) * | 2017-01-22 | 2017-05-24 | 郑州云海信息技术有限公司 | Hotspot data identification method and device |
-
2017
- 2017-06-27 CN CN201710500159.3A patent/CN107292388A/en active Pending
Patent Citations (5)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
CN103092867A (en) * | 2011-11-03 | 2013-05-08 | 中国移动通信集团甘肃有限公司 | Method and system for managing data, and data analyzing device |
CN103218416A (en) * | 2013-03-27 | 2013-07-24 | 华为技术有限公司 | Method, device and system for loading database |
CN104134103A (en) * | 2014-07-30 | 2014-11-05 | 中国石油天然气股份有限公司 | Method for predicating energy consumption of hot oil pipeline through corrected BP neural network model |
CN105635968A (en) * | 2016-01-18 | 2016-06-01 | 广州杰赛科技股份有限公司 | Hotspot area identification method based on time unit and predication method and device |
CN106709068A (en) * | 2017-01-22 | 2017-05-24 | 郑州云海信息技术有限公司 | Hotspot data identification method and device |
Cited By (10)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
CN107728952A (en) * | 2017-10-31 | 2018-02-23 | 郑州云海信息技术有限公司 | A kind of prediction type data migration method and system |
CN108182471A (en) * | 2018-01-24 | 2018-06-19 | 上海岳芯电子科技有限公司 | A kind of convolutional neural networks reasoning accelerator and method |
CN108462605A (en) * | 2018-02-06 | 2018-08-28 | 国家电网公司 | A kind of prediction technique and device of data |
CN108462605B (en) * | 2018-02-06 | 2022-03-15 | 国家电网公司 | Data prediction method and device |
CN108446340A (en) * | 2018-03-02 | 2018-08-24 | 哈尔滨工业大学(威海) | A kind of user's hot spot data access prediction technique towards mass small documents |
CN108446340B (en) * | 2018-03-02 | 2019-11-05 | 哈尔滨工业大学(威海) | A kind of user's hot spot data access prediction technique towards mass small documents |
CN112734126A (en) * | 2021-01-18 | 2021-04-30 | 武汉烽火技术服务有限公司 | Hot spot prediction method, device and equipment and readable storage medium |
CN113076339A (en) * | 2021-03-18 | 2021-07-06 | 北京沃东天骏信息技术有限公司 | Data caching method, device, equipment and storage medium |
CN117270794A (en) * | 2023-11-22 | 2023-12-22 | 成都大成均图科技有限公司 | Redis-based data storage method, medium and device |
CN117270794B (en) * | 2023-11-22 | 2024-02-23 | 成都大成均图科技有限公司 | Redis-based data storage method, medium and device |
Similar Documents
Publication | Publication Date | Title |
---|---|---|
CN107292388A (en) | A kind of Forecasting Methodology and system of the hot spot data based on neutral net | |
CN106156941B (en) | A kind of user credit scoring optimization method and device | |
CN101989236B (en) | Method for realizing instruction buffer lock | |
CN109558287A (en) | A kind of solid-state disk service life prediction technique, device and system | |
CN104090852A (en) | Method and equipment for managing hybrid cache | |
CN105653591A (en) | Hierarchical storage and migration method of industrial real-time data | |
CN105808620B (en) | A kind of method and apparatus of hyperdisk file | |
CN106502926A (en) | A kind of internal memory monitoring method, internal storage access controller and SoC systems | |
CN103150245B (en) | Determine method and the storage controller of the access characteristics of data entity | |
CN103902260A (en) | Pre-fetch method of object file system | |
CN111917882B (en) | File caching method and device and electronic equipment | |
CN109542339B (en) | Data layered access method and device, multilayer storage equipment and storage medium | |
CN103425564A (en) | Use predication method of smartphone software | |
CN109413694B (en) | Small cell caching method and device based on content popularity prediction | |
CN104699415A (en) | Method and device for controlling writing in of solid hard disk | |
CN107402851A (en) | A kind of data recovery control method and device | |
CN104794004A (en) | Information pre-loading method | |
CN115730738A (en) | Process parameter determination method, device, computer equipment and storage medium | |
CN116578409A (en) | Method, system and medium for identifying and migrating memory hot page | |
CN110413228A (en) | A kind of mapping table management method, system and electronic equipment and storage medium | |
CN107704507A (en) | Data base processing method and equipment | |
CN107180118A (en) | A kind of file system cache data managing method and device | |
CN110535686A (en) | Anomalous event treating method and apparatus | |
CN106575870A (en) | System, method and apparatuses for determining parameter settings for a power generation system and a tangible computer readable medium | |
US11301376B2 (en) | Data storage device with wear range optimization |
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 | ||
RJ01 | Rejection of invention patent application after publication |
Application publication date: 20171024 |