CN109800139A - Server health degree analysis method, device, storage medium and electronic equipment - Google Patents
Server health degree analysis method, device, storage medium and electronic equipment Download PDFInfo
- Publication number
- CN109800139A CN109800139A CN201811554842.6A CN201811554842A CN109800139A CN 109800139 A CN109800139 A CN 109800139A CN 201811554842 A CN201811554842 A CN 201811554842A CN 109800139 A CN109800139 A CN 109800139A
- Authority
- CN
- China
- Prior art keywords
- server
- sample data
- index
- group
- obtains
- 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
Abstract
Purpose of this disclosure is to provide a kind of server health degree analysis method, device, storage medium and electronic equipments, to solve the problems, such as that server health analysis is not comprehensive enough in the related technology.Method includes: to obtain the multiple groups history index of server;Health status label is added according to warning information history index described in each group in the period of server history index described in generating each group, obtains sample data;Machine learning is carried out according to the sample data, obtains health degree assessment models, wherein the input of the health degree assessment models is the operating index of server, is exported to characterize the label of the server health status.
Description
Technical field
This disclosure relates to equipment O&M technical field, and in particular, to a kind of server health degree analysis method, device,
Storage medium and electronic equipment.
Background technique
With the fast development of IT technology, branch that the equipment that various industries field all has increasing need for IT O&M is well run
It holds.The different indexs of IT server can be monitored by monitor supervision platform, and issue alarm in the case where Indexes Abnormality
Information.
In the related technology, staff can rule of thumb come judge index indicate whether equipment be in operate normally shape
State.Further, it is also possible to server be judged by simple program whether normal operation, for example, in the CPU usage of server
Lower than 70%, memory usage is lower than 80%, in the case where currently running disk remaining space more abundance, determines service
What device was up.
Summary of the invention
Purpose of this disclosure is to provide a kind of server health degree analysis method, device, storage medium and electronic equipment, with
Solve the problems, such as that server health analysis is not comprehensive enough in the related technology.
To achieve the goals above, in a first aspect, the disclosure provides a kind of server health degree analysis method, comprising:
Obtain the multiple groups history index of server;
According to the warning information in the period of server history index described in generating each group to each group of institute
History index addition health status label is stated, sample data is obtained;
Machine learning is carried out according to the sample data, obtains health degree assessment models, wherein the health degree assesses mould
The input of type is the operating index of server, is exported to characterize the label of the server health status.
Optionally, described that machine learning is carried out according to the sample data, obtain health degree assessment models, comprising:
The sample data is sent to multiple nodes and carries out distributed storage;
It is trained, is obtained more using the sample data that neural network model stores the node in each described node
A target nerve network model;
The health degree assessment models include the multiple target nerve network model.
Optionally, the method also includes:
Obtain the operating index at the server current time;
The operating index is inputted into the multiple target nerve network model respectively, obtains multiple labeling results;
According to majority voting algorithm from the multiple labeling as a result, target labels classification results are determined, and according to institute
State the health status that target labels classification results determine the server.
Optionally, the node is Ignite node;And/or the neural network model is extreme learning machine ELM nerve
Network model.
Optionally, the warning information according in the period of server history index described in generating each group
The history index described in each group adds health status label, obtains sample data, comprising:
For any group of history index, the warning information in the period of this group of history index is generated according to the server
Grade, calculate the dbjective state score of the server, and according to the corresponding relationship of state score and health status label, will
Sample label of the corresponding target health status label of the dbjective state score as this group of history index.
Second aspect, the disclosure provide a kind of server health degree analysis device, comprising:
First obtains module, for obtaining the multiple groups history index of server;
Categorization module, for according to the alarm letter in the period of server history index described in generating each group
It ceases the history index described in each group and adds health status label, obtain sample data;
Machine learning module, for obtaining health degree assessment models according to sample data progress machine learning,
In, the input of the health degree assessment models is the operating index of server, is exported to characterize the server health status
Label.
Optionally, the machine learning module, is used for:
The sample data is sent to multiple nodes and carries out distributed storage;
It is trained, is obtained more using the sample data that neural network model stores the node in each described node
A target nerve network model;
The health degree assessment models include the multiple target nerve network model.
Optionally, described device further include:
Second obtains module, for obtaining the operating index at the server current time;
Categorization module obtains multiple for the operating index to be inputted the multiple target nerve network model respectively
Labeling result;
Determining module is used for according to majority voting algorithm from the multiple labeling as a result, determining target labels classification
As a result, and determining the health status of the server according to the target labels classification results.
Optionally, the node is Ignite node;And/or the neural network model is extreme learning machine ELM nerve
Network model.
Optionally, the warning information according in the period of server history index described in generating each group
The history index described in each group adds health status label, obtains sample data, comprising:
For any group of history index, the warning information in the period of this group of history index is generated according to the server
Grade, calculate the dbjective state score of the server, and according to the corresponding relationship of state score and health status label, will
Sample label of the corresponding target health status label of the dbjective state score as this group of history index.
The third aspect, the disclosure provide a kind of computer readable storage medium, are stored thereon with computer program, the program
The step of any one server health degree analysis method is realized when being executed by processor.
Fourth aspect, the disclosure provide a kind of electronic equipment, comprising: memory is stored thereon with computer program;Processing
Device, for executing the computer program in the memory, to realize any one server health degree analysis method
The step of.
Above-mentioned technical proposal can at least reach following technical effect:
By obtaining the multiple groups history index of server, further according to server history index described in generating each group
Period in warning information history index described in each group add health status label, sample data is obtained, in this way, logical
The statement server that the label evaluation for crossing warning information to construct can allow sample data clear and accurate is within each period
State.
Further, machine learning is carried out according to the sample data, health degree assessment models is obtained, due to machine learning
The feature training that multiple dimensions can be integrated obtains health degree assessment models, can find inner link between index automatically, make
The health degree that health degree assessment models are capable of more comprehensive Analysis server is obtained, clothes are then inputted in health degree assessment models
It is engaged in after the operating index of device, the label for more accurately characterizing the server health status can be obtained.
Other feature and advantage of the disclosure will the following detailed description will be given in the detailed implementation section.
Detailed description of the invention
Attached drawing is and to constitute part of specification for providing further understanding of the disclosure, with following tool
Body embodiment is used to explain the disclosure together, but does not constitute the limitation to the disclosure.In the accompanying drawings:
Fig. 1 is a kind of server health degree analysis method flow diagram shown according to an exemplary embodiment.
Fig. 2 is another server health degree analysis method flow diagram shown according to an exemplary embodiment.
Fig. 3 is another server health degree analysis method flow diagram shown according to an exemplary embodiment.
Fig. 4 is a kind of server health degree analysis device block diagram shown according to an exemplary embodiment.
Fig. 5 is a kind of electronic equipment block diagram shown according to an exemplary embodiment.
Fig. 6 is another electronic device block diagram shown according to an exemplary embodiment.
Specific embodiment
It is described in detail below in conjunction with specific embodiment of the attached drawing to the disclosure.It should be understood that this place is retouched
The specific embodiment stated is only used for describing and explaining the disclosure, is not limited to the disclosure.
In the related technology, some servers run smoothly from index, but are practically at the state of inferior health.Phase
Detection device operating status program is too simple in the technology of pass, lacks the ability for finding multiple index internal association problems, may
Fail risk present in timely discovering device operation.In this regard, the embodiment of the present disclosure proposes a kind of server health degree analysis side
Method, with the health degree of more comprehensive Analysis server.
Fig. 1 is a kind of server health degree analysis method flow diagram shown according to an exemplary embodiment.
The described method includes:
S11 obtains the multiple groups history index of server.
Exemplary, the history index of server can be divided into following type, be CPU usage (CPU), physics respectively
Memory usage (Memory), average load (Load) in 5 minutes, disk utilization rate (Disk) and virtual memory utilization rate
(Virtual Memory, Vir-Memory).Specifically, the multiple groups history index obtained can be the finger within the scope of certain time
Mark.Wherein, each group of index corresponds to the achievement data set of the above-mentioned type of a certain moment server.For example, being one shown in table 1
Multiple groups history index in a month.
Table 1
S12, according to the warning information in the period of server history index described in generating each group to each
The group history index adds health status label, obtains sample data.
In the specific implementation, Health Category can be divided into four grades, be respectively " health ", " inferior health ", " danger ",
" serious ".
Optionally, the warning information according in the period of server history index described in generating each group
The history index described in each group adds health status label, obtains sample data, comprising: is directed to any group of history index, root
The grade that the warning information in the period of this group of history index is generated according to the server calculates the target-like of the server
State score, and according to the corresponding relationship of state score and health status label, the corresponding target of the dbjective state score is good for
Sample label of the health state tag as this group of history index.
For example, history alarm is divided into 3 kinds of grades, the alarm that grade is 1 is scored 10 points, the alarm score 25 of grade 2,
The alarm score that grade is 3 is 50 points, and total score is 100 points.
The formula of the dbjective state score of calculation server may is that
The dbjective state score of server=100-10*n1-25*n2-50*n3;
Wherein, n1 is the number that grade 1 alerts, and n2 is the number that grade 2 alerts, and n3 is the number that grade 3 alerts.When
When dividing less than 0, meter is scored at 0.
The corresponding relationship of state score and health status label can be as shown in table 2:
Table 2
Health status label | State score |
Health | >=85 |
Inferior health | 65~85 |
It is dangerous | 45~65 |
Seriously | ≤ 45 |
According to multiple groups history index shown in table 1, the sample label of every group of obtained group history index is as shown in table 3:
Table 3
S13 carries out machine learning according to the sample data, obtains health degree assessment models, wherein the health degree is commented
The input for estimating model is the operating index of server, is exported to characterize the label of the server health status.
Above-mentioned technical proposal can at least reach following technical effect:
By obtaining the multiple groups history index of server, further according to server history index described in generating each group
Period in warning information history index described in each group add health status label, sample data is obtained, in this way, logical
The statement server that the label evaluation for crossing warning information to construct can allow sample data clear and accurate is within each period
State.
Further, machine learning is carried out according to the sample data, health degree assessment models is obtained, due to machine learning
The feature training that multiple dimensions can be integrated obtains health degree assessment models, can find inner link between index automatically, make
The health degree that health degree assessment models are capable of more comprehensive Analysis server is obtained, clothes are then inputted in health degree assessment models
It is engaged in after the operating index of device, the label for more accurately characterizing the server health status can be obtained.
Fig. 2 is a kind of server health degree analysis method flow diagram shown according to an exemplary embodiment.
The described method includes:
S21 obtains the multiple groups history index of server.
Exemplary, the history index of server can be divided into following type, be CPU usage (CPU), physics respectively
Memory usage (Memory), average load (Load) in 5 minutes, disk utilization rate (Disk) and virtual memory utilization rate
(Virtual Memory, Vir-Memory).Specifically, the multiple groups history index obtained can be the finger within the scope of certain time
Mark.Wherein, each group of index corresponds to the achievement data set of the above-mentioned type of a certain moment server.
S22, according to the warning information in the period of server history index described in generating each group to each
The group history index adds health status label, obtains sample data.
In the specific implementation, Health Category can be divided into four grades, be respectively " health ", " inferior health ", " danger ",
" serious ".
Optionally, the warning information according in the period of server history index described in generating each group
The history index described in each group adds health status label, obtains sample data, comprising: is directed to any group of history index, root
The grade that the warning information in the period of this group of history index is generated according to the server calculates the target-like of the server
State score, and according to the corresponding relationship of state score and health status label, the corresponding target of the dbjective state score is good for
Sample label of the health state tag as this group of history index.
For example, history alarm is divided into 3 kinds of grades, the alarm that grade is 1 is scored 10 points, the alarm score 25 of grade 2,
The alarm score that grade is 3 is 50 points, and total score is 100 points.
The formula of the dbjective state score of calculation server may is that
The dbjective state score of server=100-10*n1-25*n2-50*n3;
Wherein, n1 is the number that grade 1 alerts, and n2 is the number that grade 2 alerts, and n3 is the number that grade 3 alerts.When
When dividing less than 0, meter is scored at 0.
S23 handles the sample data by resampling, obtains multiple sample data training subsets.
Resampling, which refers to the process of, goes out another kind of picture dot information according to the message interpolation of a kind of picture dot.
Such as sample data obtained above has 5 dimensions, corresponding history index is divided into 5 dimension types, per one-dimensional
Degree is N number of according to having, then sample data is the matrix X of (N*5).Further, by putting back to uniform sampling to above-mentioned sample
Data carry out resampling processing, obtain multiple and different sample data training subset Q.Wherein, each sample data training subset Q
Can be (M*5) matrix, due to the data of the every dimension of Q matrix be put to the proof according to X in every N number of data resampling of dimension obtain
Out, the data amount check M of the every dimension of Q matrix should be slightly less than N.
By resampling, multiple sample data training subsets for having otherness can be generated, are based on multiple sample datas
Training subset can be configured with the target nerve network model of multiple othernesses in the next steps.Multiple sample data training
Otherness between the target nerve network model for multiple othernesses that the otherness of collection can largely guarantee.
The number of obtained sample data training subset can be according to of subsequent obtained target nerve network model
Number is consistent.
In the present embodiment, health degree analysis is substantially a classification problem, therefore sample data is the classification of 5 dimensions
Data, wherein 5 column of front are different variation, last sample label column are equivalent to category.Correspondingly, in subsequent machine
Learn in obtained health degree assessment models, the operating index of input is the data of 5 dimensions.
Further, machine learning is carried out according to the sample data, obtains health degree assessment models, wherein is described strong
The input of Kang Du assessment models is the operating index of server, is exported to characterize the label of the server health status.
The sample data training subset is sent to multiple nodes and carries out distributed storage by S24.
S25 is trained using the sample data that neural network model stores the node in each described node, is obtained
To multiple target nerve network models.
Wherein, the health degree assessment models include the multiple target nerve network model.
S26 obtains the operating index at the server current time.
The operating index is inputted the multiple target nerve network model respectively, obtains multiple labeling knots by S27
Fruit.
S28, according to majority voting algorithm from the multiple labeling as a result, determining target labels classification results, and root
The health status of the server is determined according to the target labels classification results.
Majority voting algorithm can be divided into the progress of two steps, and the first step finds the most labeling result of frequency of occurrence
(majority element, most elements);Second step, if the number that majority element occurs is greater than total first prime number
The half of amount, being returned to majority element is the target labels classification results.
For more intuitive displaying above-mentioned technical proposal, exemplary illustration is done to above-mentioned technical proposal with Fig. 3.
As shown in figure 3, being monitored first to server, Server history monitoring data is obtained, these history monitor number
It is the multiple groups history index that may include above-mentioned server in.
Further, data processing is carried out to multiple groups history index, and carries out the division of health degree, gone through described in each group
History index adds health status label, obtains sample data.
Further, resampling is carried out to sample data, obtains T sample data training subset (data subset 1, data
Subset 2 ... data subset T).Specifically, resampling can be carried out using bootstrap (bootstrap) methods of sampling.
(as shown in dotted outline in FIG.) hereafter is carried out in distributed node to the processing of data.
Further, then each data subset is assigned to multiple nodes and carries out distributed storage.Optionally, the node
For Ignite node.
Apache Ignite is a kind of memory platform, it has the characteristics that high-performance, integrated and distributed, can be with
Executive Office's reason and calculating business in the case where big data quantity in real time.With traditional storage based on disk or flash memory
Technology is compared, and data processing is more excellent.In addition, insertion type is developed, occupancy resource is small, calculates solving big data parallel-type
It is more advantageous when problem.
Ignite distributed computing is provided based on the memory platform, this distributed computing can pass through parallel form
Enhance the performance of data processing, reduces delay, promote linear extended capability.Ignite distributed computing is provided in clustered node
Perhaps task can be executed or be closed in the form of one distributed by the method that number of different types calculate being run in a cluster group
Packet.
In above-mentioned optional embodiment, Ignite distributed computing and ability to communicate can use, by the way that data are sub
Collection is assigned to multiple Ignite Node distribution formula storages, then carries out weak collector training in each Ignite node, works as training
It completes result synchronous transfer obtaining final result to unified node.Ignite distributed computing characteristic is introduced, is effectively subtracted
Lack analysis time, improves prediction timeliness.
Further, sample data is trained using neural network model in Ignite node.
In the example of fig. 3, neural network model is ELM (extreme learning machine) network.T target nerve is obtained after training
Network model, that is, trained ELM network (ELM1, ELM2 ... ELMT).It is defeated in each trained ELM network
Enter/weight is exported, number of nodes is implied, biasing and activation primitive are fixed.
ELM (extreme learning machine) network has cracking training speed.The analysis result of multiple ELM networks may have centainly
Otherness, and integrated study is just needing this otherness just.Have otherness between the ELM that each fulcrum is respectively trained,
The different ELM networks that can be trained have different emphasis, thus the new data of more comprehensive analysis input.This
Outside, integrated learning approach can promote the accuracy and generalization ability of ELM network analysis result.
Further, to the operating index at each Ignite node incoming service device current time, that is, newest number
According to.The operating index at these current times is also 5 dimensions, using training complete ELM network carry out analysis obtain analysis as a result,
It is total obtain T classification results (classification results 1, classification results 2 ... classification results T).
Further, the T result obtained in above-mentioned steps is screened using majority voting method, obtains final knot
Fruit.Handled using majority voting method integrated multiple EML networks as a result, it is possible to provides reasonable response to the data newly inputted,
Enhance generalization ability, promotes the accuracy of result.
The present embodiment obtains multiple ELM networks using integrated learning approach.The essence of ELM is a kind of neural network,
Realization principle can be indicated by following formula:
N number of training sample ui=[ui1,ui2,...,uip]T∈RpIt is input vector;
ti=[ti1,ti2,…,tiq]T∈RqIt is output vector, m is the number of hidden nodes, and b is to imply bigoted, and g (x) is activation
Function, wj=[wj1,wj2,…,wjp]TIt is input quantity weight, βj=[βj1,βj2,...,βjq]TIt is output quantity weight.
The form of above formula matrix multiple of being write as is Y=Η β;Wherein Y is the output of desired network, and H is known as the state square of ELM
Battle array, expansion are as follows:
The error of multiple ELM networks is done to analyze below.Assuming that the collection of integrated multiple ELM (extreme learning machine) network
It is equivalent to the resultful simple average of institute at result, then can obtain following formula:
Wherein, n is the number for participating in integrated ELM network, yiThe output of any ELM network is represented,What representative integrated
Output.Assuming that y representative sample reality output, yiIndicate the predicted value of i-th of ELM network, εiIt is error term, then can obtains:
yi=y+ εi;
The expectation of the mean square error of so i-th ELM network are as follows: E [{ yi-y}2]=E [εi 2];N ELM network it is square
Average error are as follows:
The mean square error for the multiple ELM networks that can must be integrated it is expected are as follows:
Assuming that error term εiBetween it is uncorrelated and have zero-mean, i.e. E [εiεj]=0, E [εi]=0;So then have:
It can be seen that the mean square error of integrated multiple ELM networks is significantly lower than the mean error of single extreme learning machine.
It is worth noting that the error in actual implementation between multiple ELM networks may be highly relevant, integrated is multiple
The mean square error of ELM network is slightly above the 1/n of mean error, but the error of integrated extreme learning machine be less than mean error still at
It is vertical.
The embodiment of the present disclosure at least can be realized following technical effect:
1. adding health status label in the history index described in each group, refining classification using health degree is indicated, is helped
In precise expression server current state;
2. pair multiple Artificial neural network ensembles learn to form strong learner and enhance the Generalization Capability of method forecast analysis;
3. utilizing Ignite distributed communication parallel computation, the training effectiveness of ensemble machine learning is promoted;
4. can be more efficient analyze server current state, help user to make accurate judgment O&M situation.
Fig. 4 is a kind of server health degree analysis device block diagram shown according to an exemplary embodiment.The device includes:
First obtains module 410, for obtaining the multiple groups history index of server;
Categorization module 420, for according to the announcement in the period of server history index described in generating each group
Alert information history index described in each group adds health status label, obtains sample data;
Machine learning module 430, for obtaining health degree assessment models according to sample data progress machine learning,
Wherein, the input of the health degree assessment models is the operating index of server, is exported to characterize the server health status
Label.
Above-mentioned technical proposal can at least reach following technical effect:
By obtaining the multiple groups history index of server, further according to server history index described in generating each group
Period in warning information history index described in each group add health status label, sample data is obtained, in this way, logical
The statement server that the label evaluation for crossing warning information to construct can allow sample data clear and accurate is within each period
State.
Further, machine learning is carried out according to the sample data, health degree assessment models is obtained, due to machine learning
The feature training that multiple dimensions can be integrated obtains health degree assessment models, can find inner link between index automatically, make
The health degree that health degree assessment models are capable of more comprehensive Analysis server is obtained, clothes are then inputted in health degree assessment models
It is engaged in after the operating index of device, the label for more accurately characterizing the server health status can be obtained.
Optionally, the machine learning module, is used for:
The sample data is sent to multiple nodes and carries out distributed storage;
It is trained, is obtained more using the sample data that neural network model stores the node in each described node
A target nerve network model;
The health degree assessment models include the multiple target nerve network model.
Optionally, described device further include:
Second obtains module, for obtaining the operating index at the server current time;
Categorization module obtains multiple for the operating index to be inputted the multiple target nerve network model respectively
Labeling result;
Determining module is used for according to majority voting algorithm from the multiple labeling as a result, determining target labels classification
As a result, and determining the health status of the server according to the target labels classification results.
Optionally, the node is Ignite node;And/or the neural network model is extreme learning machine ELM nerve
Network model.
Optionally, the warning information according in the period of server history index described in generating each group
The history index described in each group adds health status label, obtains sample data, comprising:
For any group of history index, the warning information in the period of this group of history index is generated according to the server
Grade, calculate the dbjective state score of the server, and according to the corresponding relationship of state score and health status label, will
Sample label of the corresponding target health status label of the dbjective state score as this group of history index.
About the device in above-described embodiment, wherein modules execute the concrete mode of operation in related this method
Embodiment in be described in detail, no detailed explanation will be given here.
The embodiment of the present disclosure provides a kind of computer readable storage medium, is stored thereon with computer program, the program quilt
The step of any one server health degree analysis method is realized when processor executes.
The embodiment of the present disclosure provides a kind of electronic equipment, comprising: memory is stored thereon with computer program;Processor,
For executing the computer program in the memory, to realize the step of any one server health degree analysis method
Suddenly.
Fig. 5 is the block diagram of a kind of electronic equipment 500 shown according to an exemplary embodiment.The electronic equipment can be mentioned
For the health status detection platform for server.
As shown in figure 5, the electronic equipment 500 may include: processor 501, memory 502.The electronic equipment 500 may be used also
To include multimedia component 503, one or more of input/output (I/O) interface 504 and communication component 505.
Wherein, processor 501 is used to control the integrated operation of the electronic equipment 500, to complete above-mentioned server health
Spend all or part of the steps in analysis method.Memory 502 is for storing various types of data to support to set in the electronics
Standby 500 operation, these data for example may include any application or method for operating on the electronic equipment 500
Instruction and the relevant data of application program, for example, warning information, index classification data etc.;Furthermore it is also possible that connection
It is personal data, the message of transmitting-receiving, picture, audio, video etc..The memory 502 can be by any kind of volatibility or non-
Volatile storage devices or their combination are realized, such as static random access memory (Static Random Access
Memory, abbreviation SRAM), electrically erasable programmable read-only memory (Electrically Erasable Programmable
Read-Only Memory, abbreviation EEPROM), Erasable Programmable Read Only Memory EPROM (Erasable Programmable
Read-Only Memory, abbreviation EPROM), programmable read only memory (Programmable Read-Only Memory, letter
Claim PROM), read-only memory (Read-Only Memory, abbreviation ROM), magnetic memory, flash memory, disk or CD.
Multimedia component 503 may include screen and audio component.Wherein screen for example can be touch screen, and audio component is for exporting
And/or input audio signal.For example, audio component may include a microphone, microphone is for receiving external audio signal.
The received audio signal can be further stored in memory 502 or be sent by communication component 505.Audio component also wraps
At least one loudspeaker is included, output audio signal is used for.I/O interface 504 provides between processor 501 and other interface modules
Interface, other above-mentioned interface modules can be keyboard, mouse, button etc..These buttons can be virtual push button or entity is pressed
Button.Communication component 505 is for carrying out wired or wireless communication between the electronic equipment 500 and other equipment.Wireless communication, such as
Wi-Fi, bluetooth, near-field communication (Near Field Communication, abbreviation NFC), 2G, 3G, 4G, NB-IOT, eMTC or
Other 5G etc. or they one or more of combination, it is not limited here.Therefore the corresponding communication component 505 can
To include: Wi-Fi module, bluetooth module, NFC module etc..
In one exemplary embodiment, electronic equipment 500 can be by one or more application specific integrated circuit
(Application Specific Integrated Circuit, abbreviation ASIC), digital signal processor (Digital
Signal Processor, abbreviation DSP), digital signal processing appts (Digital Signal Processing Device,
Abbreviation DSPD), programmable logic device (Programmable Logic Device, abbreviation PLD), field programmable gate array
(Field Programmable Gate Array, abbreviation FPGA), controller, microcontroller, microprocessor or other electronics member
Part is realized, for executing above-mentioned server health degree analysis method.
In a further exemplary embodiment, a kind of computer readable storage medium including program instruction is additionally provided, it should
The step of above-mentioned server health degree analysis method is realized when program instruction is executed by processor.For example, this is computer-readable
Storage medium can be the above-mentioned memory 502 including program instruction, and above procedure instruction can be by the processor of electronic equipment 500
501 execute to complete above-mentioned server health degree analysis method.
Fig. 6 is the block diagram of a kind of electronic equipment 600 shown according to an exemplary embodiment.For example, electronic equipment 600 can
To be provided as a server.Referring to Fig. 6, electronic equipment 600 includes processor 622, and quantity can be one or more, with
And memory 632, for storing the computer program that can be executed by processor 622.The computer program stored in memory 632
May include it is one or more each correspond to one group of instruction module.In addition, processor 622 can be configured as
The computer program is executed, to execute above-mentioned server health degree analysis method.
In addition, electronic equipment 600 can also include power supply module 626 and communication component 650, which can be with
It is configured as executing the power management of electronic equipment 600, which, which can be configured as, realizes electronic equipment 600
Communication, for example, wired or wireless communication.In addition, the electronic equipment 600 can also include input/output (I/O) interface 658.Electricity
Sub- equipment 600 can be operated based on the operating system for being stored in memory 632, such as Windows ServerTM, Mac OS
XTM, UnixTM, LinuxTM etc..
In a further exemplary embodiment, a kind of computer readable storage medium including program instruction is additionally provided, it should
The step of above-mentioned server health degree analysis method is realized when program instruction is executed by processor.For example, this is computer-readable
Storage medium can be the above-mentioned memory 632 including program instruction, and above procedure instruction can be by the processor of electronic equipment 600
622 execute to complete above-mentioned server health degree analysis method.
The preferred embodiment of the disclosure is described in detail in conjunction with attached drawing above, still, the disclosure is not limited to above-mentioned reality
The detail in mode is applied, in the range of the technology design of the disclosure, a variety of letters can be carried out to the technical solution of the disclosure
Monotropic type, these simple variants belong to the protection scope of the disclosure.
It is further to note that specific technical features described in the above specific embodiments, in not lance
In the case where shield, can be combined in any appropriate way, in order to avoid unnecessary repetition, the disclosure to it is various can
No further explanation will be given for the combination of energy.
In addition, any combination can also be carried out between a variety of different embodiments of the disclosure, as long as it is without prejudice to originally
Disclosed thought equally should be considered as disclosure disclosure of that.
Claims (10)
1. a kind of server health degree analysis method characterized by comprising
Obtain the multiple groups history index of server;
It is gone through described in each group according to the warning information in the period of server history index described in generating each group
History index adds health status label, obtains sample data;
Machine learning is carried out according to the sample data, obtains health degree assessment models, wherein the health degree assessment models
Input is the operating index of server, is exported to characterize the label of the server health status.
2. being obtained the method according to claim 1, wherein described carry out machine learning according to the sample data
To health degree assessment models, comprising:
The sample data is sent to multiple nodes and carries out distributed storage;
It is trained in each described node using the sample data that neural network model stores the node, obtains multiple mesh
Mark neural network model;
The health degree assessment models include the multiple target nerve network model.
3. according to the method described in claim 2, it is characterized in that, the method also includes:
Obtain the operating index at the server current time;
The operating index is inputted into the multiple target nerve network model respectively, obtains multiple labeling results;
According to majority voting algorithm from the multiple labeling as a result, target labels classification results are determined, and according to the mesh
Mark labeling result determines the health status of the server.
4. according to the method in claim 2 or 3, which is characterized in that the node is Ignite node;And/or the mind
It is extreme learning machine ELM neural network model through network model.
5. according to the method in any one of claims 1 to 3, which is characterized in that described to be generated according to the server
Warning information history index described in each group in the period of history index described in each group adds health status label, obtains
To sample data, comprising:
For any group of history index, generated according to the server warning information in the period of this group of history index etc.
Grade calculates the dbjective state score of the server, and according to the corresponding relationship of state score and health status label, will be described
Sample label of the corresponding target health status label of dbjective state score as this group of history index.
6. a kind of server health degree analysis device characterized by comprising
First obtains module, for obtaining the multiple groups history index of server;
Categorization module, for according to the warning information pair in the period of server history index described in generating each group
History index described in each group adds health status label, obtains sample data;
Machine learning module obtains health degree assessment models, wherein institute for carrying out machine learning according to the sample data
The input for stating health degree assessment models is the operating index of server, is exported to characterize the label of the server health status.
7. device according to claim 6, which is characterized in that the machine learning module is used for:
The sample data is sent to multiple nodes and carries out distributed storage;
It is trained in each described node using the sample data that neural network model stores the node, obtains multiple mesh
Mark neural network model;
The health degree assessment models include the multiple target nerve network model.
8. device according to claim 7, which is characterized in that described device further include:
Second obtains module, for obtaining the operating index at the server current time;
Categorization module obtains multiple labels for the operating index to be inputted the multiple target nerve network model respectively
Classification results;
Determining module, for according to majority voting algorithm from the multiple labeling as a result, determine target labels classification results,
And the health status of the server is determined according to the target labels classification results.
9. a kind of computer readable storage medium, is stored thereon with computer program, which is characterized in that the program is held by processor
The step of any one of claim 1-5 the method is realized when row.
10. a kind of electronic equipment characterized by comprising
Memory is stored thereon with computer program;
Processor, for executing the computer program in the memory, to realize described in any one of claim 1-5
The step of method.
Priority Applications (1)
Application Number | Priority Date | Filing Date | Title |
---|---|---|---|
CN201811554842.6A CN109800139A (en) | 2018-12-18 | 2018-12-18 | Server health degree analysis method, device, storage medium and electronic equipment |
Applications Claiming Priority (1)
Application Number | Priority Date | Filing Date | Title |
---|---|---|---|
CN201811554842.6A CN109800139A (en) | 2018-12-18 | 2018-12-18 | Server health degree analysis method, device, storage medium and electronic equipment |
Publications (1)
Publication Number | Publication Date |
---|---|
CN109800139A true CN109800139A (en) | 2019-05-24 |
Family
ID=66557136
Family Applications (1)
Application Number | Title | Priority Date | Filing Date |
---|---|---|---|
CN201811554842.6A Pending CN109800139A (en) | 2018-12-18 | 2018-12-18 | Server health degree analysis method, device, storage medium and electronic equipment |
Country Status (1)
Country | Link |
---|---|
CN (1) | CN109800139A (en) |
Cited By (16)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
CN110572297A (en) * | 2019-08-09 | 2019-12-13 | 网宿科技股份有限公司 | Network performance evaluation method, server and storage medium |
CN110750784A (en) * | 2019-09-24 | 2020-02-04 | 深圳市智莱科技股份有限公司 | Safety prevention and control method and system for automatic vending equipment |
CN110865925A (en) * | 2019-11-12 | 2020-03-06 | 拉扎斯网络科技(上海)有限公司 | Information processing method, information processing device, electronic equipment and computer readable storage medium |
CN111107158A (en) * | 2019-12-26 | 2020-05-05 | 远景智能国际私人投资有限公司 | Alarm method, device, equipment and medium for Internet of things equipment cluster |
CN111897706A (en) * | 2020-07-15 | 2020-11-06 | 中国工商银行股份有限公司 | Server performance prediction method, device, computer system and medium |
CN111949498A (en) * | 2020-08-10 | 2020-11-17 | 中国工商银行股份有限公司 | Application server abnormity prediction method and system |
CN112085344A (en) * | 2020-08-17 | 2020-12-15 | 远光软件股份有限公司 | Equipment health analysis method and device and electronic equipment |
CN112393931A (en) * | 2019-08-13 | 2021-02-23 | 北京国双科技有限公司 | Detection method, detection device, electronic equipment and computer readable medium |
CN112799911A (en) * | 2021-01-27 | 2021-05-14 | 苏州浪潮智能科技有限公司 | Node health state detection method, device, equipment and storage medium |
CN112817832A (en) * | 2021-01-19 | 2021-05-18 | 网易(杭州)网络有限公司 | Method, device and equipment for acquiring health state of game server and storage medium |
CN113110981A (en) * | 2021-03-26 | 2021-07-13 | 北京中大科慧科技发展有限公司 | Air conditioner room health energy efficiency detection method for data center |
WO2021184554A1 (en) * | 2020-03-18 | 2021-09-23 | 平安科技(深圳)有限公司 | Database exception monitoring method and device, computer device, and storage medium |
CN113496330A (en) * | 2020-04-01 | 2021-10-12 | 逸驾智能科技有限公司 | Method, apparatus, and medium for assessing quality of service of a station |
CN113762347A (en) * | 2021-08-06 | 2021-12-07 | 佳都科技集团股份有限公司 | Method and device for evaluating health degree of sliding door body |
CN115793990A (en) * | 2023-02-06 | 2023-03-14 | 天翼云科技有限公司 | Memory health state determination method and device, electronic equipment and storage medium |
CN116975567A (en) * | 2023-07-28 | 2023-10-31 | 上海优立检测技术股份有限公司 | Method, system, equipment and storage medium for testing radiation interference resistance of server |
Citations (8)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
CN102033965A (en) * | 2011-01-17 | 2011-04-27 | 安徽海汇金融投资集团有限公司 | Method and system for classifying data based on classification model |
CN106886481A (en) * | 2017-02-28 | 2017-06-23 | 深圳市华傲数据技术有限公司 | A kind of system health degree static analysis Forecasting Methodology and device |
CN106953766A (en) * | 2017-03-31 | 2017-07-14 | 北京奇艺世纪科技有限公司 | A kind of alarm method and device |
US20170286624A1 (en) * | 2016-03-31 | 2017-10-05 | Alibaba Group Holding Limited | Methods, Systems, and Devices for Evaluating a Health Condition of an Internet User |
CN107346433A (en) * | 2016-05-06 | 2017-11-14 | 华为技术有限公司 | A kind of text data sorting technique and server |
CN107943677A (en) * | 2017-10-13 | 2018-04-20 | 东软集团股份有限公司 | Application performance monitoring method, device, readable storage medium storing program for executing and electronic equipment |
CN108304935A (en) * | 2017-05-09 | 2018-07-20 | 腾讯科技(深圳)有限公司 | Machine learning model training method, device and computer equipment |
CN108710555A (en) * | 2018-05-23 | 2018-10-26 | 郑州云海信息技术有限公司 | A kind of server error diagnosis method based on supervised learning |
-
2018
- 2018-12-18 CN CN201811554842.6A patent/CN109800139A/en active Pending
Patent Citations (8)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
CN102033965A (en) * | 2011-01-17 | 2011-04-27 | 安徽海汇金融投资集团有限公司 | Method and system for classifying data based on classification model |
US20170286624A1 (en) * | 2016-03-31 | 2017-10-05 | Alibaba Group Holding Limited | Methods, Systems, and Devices for Evaluating a Health Condition of an Internet User |
CN107346433A (en) * | 2016-05-06 | 2017-11-14 | 华为技术有限公司 | A kind of text data sorting technique and server |
CN106886481A (en) * | 2017-02-28 | 2017-06-23 | 深圳市华傲数据技术有限公司 | A kind of system health degree static analysis Forecasting Methodology and device |
CN106953766A (en) * | 2017-03-31 | 2017-07-14 | 北京奇艺世纪科技有限公司 | A kind of alarm method and device |
CN108304935A (en) * | 2017-05-09 | 2018-07-20 | 腾讯科技(深圳)有限公司 | Machine learning model training method, device and computer equipment |
CN107943677A (en) * | 2017-10-13 | 2018-04-20 | 东软集团股份有限公司 | Application performance monitoring method, device, readable storage medium storing program for executing and electronic equipment |
CN108710555A (en) * | 2018-05-23 | 2018-10-26 | 郑州云海信息技术有限公司 | A kind of server error diagnosis method based on supervised learning |
Non-Patent Citations (1)
Title |
---|
郭躬德等: "《近邻分类方法及其应用 上册》", 31 December 2013 * |
Cited By (23)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
CN110572297B (en) * | 2019-08-09 | 2021-07-06 | 网宿科技股份有限公司 | Network performance evaluation method, server and storage medium |
CN110572297A (en) * | 2019-08-09 | 2019-12-13 | 网宿科技股份有限公司 | Network performance evaluation method, server and storage medium |
CN112393931B (en) * | 2019-08-13 | 2022-04-19 | 北京国双科技有限公司 | Detection method, detection device, electronic equipment and computer readable medium |
CN112393931A (en) * | 2019-08-13 | 2021-02-23 | 北京国双科技有限公司 | Detection method, detection device, electronic equipment and computer readable medium |
CN110750784A (en) * | 2019-09-24 | 2020-02-04 | 深圳市智莱科技股份有限公司 | Safety prevention and control method and system for automatic vending equipment |
CN110750784B (en) * | 2019-09-24 | 2023-10-03 | 深圳市智莱科技股份有限公司 | Security prevention and control method and system for automatic vending equipment |
CN110865925A (en) * | 2019-11-12 | 2020-03-06 | 拉扎斯网络科技(上海)有限公司 | Information processing method, information processing device, electronic equipment and computer readable storage medium |
CN111107158A (en) * | 2019-12-26 | 2020-05-05 | 远景智能国际私人投资有限公司 | Alarm method, device, equipment and medium for Internet of things equipment cluster |
CN111107158B (en) * | 2019-12-26 | 2023-02-17 | 远景智能国际私人投资有限公司 | Alarm method, device, equipment and medium for Internet of things equipment cluster |
WO2021184554A1 (en) * | 2020-03-18 | 2021-09-23 | 平安科技(深圳)有限公司 | Database exception monitoring method and device, computer device, and storage medium |
CN113496330A (en) * | 2020-04-01 | 2021-10-12 | 逸驾智能科技有限公司 | Method, apparatus, and medium for assessing quality of service of a station |
CN111897706A (en) * | 2020-07-15 | 2020-11-06 | 中国工商银行股份有限公司 | Server performance prediction method, device, computer system and medium |
CN111949498A (en) * | 2020-08-10 | 2020-11-17 | 中国工商银行股份有限公司 | Application server abnormity prediction method and system |
CN112085344A (en) * | 2020-08-17 | 2020-12-15 | 远光软件股份有限公司 | Equipment health analysis method and device and electronic equipment |
CN112817832A (en) * | 2021-01-19 | 2021-05-18 | 网易(杭州)网络有限公司 | Method, device and equipment for acquiring health state of game server and storage medium |
CN112799911A (en) * | 2021-01-27 | 2021-05-14 | 苏州浪潮智能科技有限公司 | Node health state detection method, device, equipment and storage medium |
CN113110981A (en) * | 2021-03-26 | 2021-07-13 | 北京中大科慧科技发展有限公司 | Air conditioner room health energy efficiency detection method for data center |
CN113110981B (en) * | 2021-03-26 | 2024-04-09 | 北京中大科慧科技发展有限公司 | Air conditioner room health energy efficiency detection method for data center |
CN113762347A (en) * | 2021-08-06 | 2021-12-07 | 佳都科技集团股份有限公司 | Method and device for evaluating health degree of sliding door body |
CN113762347B (en) * | 2021-08-06 | 2023-08-08 | 佳都科技集团股份有限公司 | Sliding door body health degree assessment method and device |
CN115793990A (en) * | 2023-02-06 | 2023-03-14 | 天翼云科技有限公司 | Memory health state determination method and device, electronic equipment and storage medium |
CN116975567A (en) * | 2023-07-28 | 2023-10-31 | 上海优立检测技术股份有限公司 | Method, system, equipment and storage medium for testing radiation interference resistance of server |
CN116975567B (en) * | 2023-07-28 | 2024-03-15 | 上海优立检测技术股份有限公司 | Method, system, equipment and storage medium for testing radiation interference resistance of server |
Similar Documents
Publication | Publication Date | Title |
---|---|---|
CN109800139A (en) | Server health degree analysis method, device, storage medium and electronic equipment | |
Kocaguneli et al. | Transfer learning in effort estimation | |
Jurgens et al. | Geolocation prediction in twitter using social networks: A critical analysis and review of current practice | |
Corizzo et al. | Anomaly detection and repair for accurate predictions in geo-distributed big data | |
TWI696124B (en) | Model integration method and device | |
TWI742040B (en) | Benchmark test method and device for supervised learning algorithm in distributed environment | |
CN101841435B (en) | Method, apparatus and system for detecting abnormality of DNS (domain name system) query flow | |
Kim et al. | Cloudinsight: Utilizing a council of experts to predict future cloud application workloads | |
CN106528951A (en) | Life prediction and safety warning methods and apparatuses for power battery | |
US20140164364A1 (en) | System and method for event-driven prioritization | |
Tang et al. | A distributed and scalable approach to semi-intrusive load monitoring | |
CN103678346A (en) | Man-machine recognition method and system | |
JP2017021772A (en) | Copula-theory based feature selection | |
CN106656662A (en) | Method and system for determining abnormal bandwidth, and electronic device | |
Tian et al. | Short-term wind speed forecasting based on autoregressive moving average with echo state network compensation | |
Zhuge et al. | Sensitivity analysis of integrated activity-based model: Using MATSim as an example | |
CN106059829A (en) | Hidden markov-based network utilization ratio sensing method | |
Sladescu et al. | Event aware workload prediction: A study using auction events | |
Cepeda et al. | Data‐mining‐based approach for predicting the power system post‐contingency dynamic vulnerability status | |
Nunes et al. | The elimination-selection based algorithm for efficient resource discovery in Internet of Things environments | |
Violos et al. | Predicting visitor distribution for large events in smart cities | |
Wang et al. | CPB: a classification-based approach for burst time prediction in cascades | |
JP6680663B2 (en) | Information processing apparatus, information processing method, prediction model generation apparatus, prediction model generation method, and program | |
Lichtenwalter et al. | Vertex collocation profiles: theory, computation, and results | |
Pramesti et al. | Autoscaling Based on Response Time Prediction for Microservice Application in Kubernetes |
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: 20190524 |
|
RJ01 | Rejection of invention patent application after publication |