CN109474445A - A kind of distributed system root Fault Locating Method and device - Google Patents
A kind of distributed system root Fault Locating Method and device Download PDFInfo
- Publication number
- CN109474445A CN109474445A CN201710801677.9A CN201710801677A CN109474445A CN 109474445 A CN109474445 A CN 109474445A CN 201710801677 A CN201710801677 A CN 201710801677A CN 109474445 A CN109474445 A CN 109474445A
- Authority
- CN
- China
- Prior art keywords
- attribute
- test data
- complete subgraph
- test
- attribute value
- Prior art date
- Legal status (The legal status is an assumption and is not a legal conclusion. Google has not performed a legal analysis and makes no representation as to the accuracy of the status listed.)
- Granted
Links
Classifications
-
- H—ELECTRICITY
- H04—ELECTRIC COMMUNICATION TECHNIQUE
- H04L—TRANSMISSION OF DIGITAL INFORMATION, e.g. TELEGRAPHIC COMMUNICATION
- H04L41/00—Arrangements for maintenance, administration or management of data switching networks, e.g. of packet switching networks
- H04L41/06—Management of faults, events, alarms or notifications
- H04L41/0677—Localisation of faults
Abstract
The present invention relates to a kind of distributed system root Fault Locating Method and devices, which comprises obtains distributed system test data, and selects to be tested the not available test data of service in the test data；Trouble-shooting chart is constructed using the not available test data of service is tested in the test data, and obtains the complete subgraph in the trouble-shooting chart；Fault location is carried out according to the complete subgraph, technical solution provided by the invention, trouble-shooting chart is constructed using test data provided by the invention, Faults in Distributed Systems is rapidly and accurately diagnosed according to trouble-shooting chart, the puzzlement and economic loss for effectivelying prevent thrashing to generate.
Description
Technical field
The present invention relates to data mining technology fields, and in particular to a kind of distributed system root Fault Locating Method and dress
It sets.
Background technique
With the rapid development of the Internet technologies such as cloud computing and big data, the scale of large enterprise's distributed information system
It is more and more huger.For example, cut-off 2014, Amazon has built up 11 cloud region, 28 data central sets, and 2,000,000
Platform server.Cut-off 2014, Google has 1,000,000 servers in global data center.Once these are carried
It largely breaks down with the closely bound up operation system of economical production and social life, by network attack, it will be to society
Production and living bring greatly inconvenient and economic loss, even result in serious social security events.For example, in May, 2015
27, there is large area paralysis in the Alipay for possessing nearly 300,000,000 any active ues；Ctrip.com is small up to 12 on May 28th, 2015
Shi Wufa access, about 50,000,000 yuans of economic loss.Therefore, in extensive information system, event is fast and effeciently carried out
The automatic detection of barrier and diagnosis are all very important in research and practice.
The method of present analysis distributed system root failure include the method based on figure, the method based on expert system,
Analysis model, data-driven method.
Method for diagnosing faults based on graph theory mainly includes signed digraph (Signed directed graph, SDG) side
Method and fault tree analysis method.SDG is a kind of causal graphical model of description system, and node indicates event or variable, has
The causality between variable is indicated to side.When breaking down, causality between being changed according to node, and combine certain search
Rope strategy judges the reason of failure occurs.Fault tree be it is a kind of by fruit to because logic chart, it from the malfunction of system,
Fundamental cause, influence degree and probability of happening that determining failure occurs are made inferences step by step.
Knowledge base is established using the practical experience of domain expert based on the method for diagnosing faults of expert system, and is made inferences
Fault diagnosis is carried out with decision process.Expertise commonly uses deterministic IF-THEN Rule Expression.Since expertise has not
Certainty, fuzzy expert system introduces the concept of fuzzy membership, and is made inferences using fuzzy logic, can be located in very well
Manage the uncertainty in expertise.
System expected behavior is described using the accurate mathematical model of system based on the method for diagnosing faults of analytic modell analytical model, is passed through
It compares whether actual input and output amount unanimously carries out fault diagnosis, mainly includes method for estimating state, method for parameter estimation, base
In the method for equivalent space.Method based on state estimation mainly includes filtered method and observer method, is examined using hypothesis
It the methods of tests and to carry out fault diagnosis.Method based on parameter Estimation thinks that failure can cause the variation of model parameter, and passes through inspection
The Parameters variation surveyed in model carries out fault diagnosis.
The method for diagnosing faults of data-driven is not required to by being analyzed and processed carry out fault diagnosis to system operation data
Know system accurate Analysis model, is broadly divided into machine learning class method, multi-variate statistical analysis class method etc..Machine learning class side
Method is the historical data training machines such as neural network or support vector machines using system under normal and various fault conditions
Learning algorithm is used for fault diagnosis.Method based on multi-variate statistical analysis is to carry out failure using the correlation between multiple variables
Diagnosis, mainly there is PCA method.The sample matrix of variable is projected to lower dimensional space by PCA method, reflects the main of process variable
Variation, and fault diagnosis is carried out by given Testing index.
There are still significant limitations for existing distributed information system method for diagnosing faults, mainly have following two aspect former
Cause: it is on the one hand, not intuitive enough for the Faults in Distributed Systems diagnostic method of usability testing failure, in practice, due to distribution
Formula information system becomes increasingly complex, and fault point is difficult to be detected comprehensively, therefore is difficult accurately to diagnose fault, and needs more
Intuitive analysis and diagnosis tool；On the other hand, existing method for diagnosing faults is difficult to solve real-time, unsupervised fault diagnosis
Problem.The method of graph theory, the method for expert system and the method based on data-driven need system structure knowledge or training data,
Method based on parsing is difficult to complex system modeling, and these data or knowledge are difficult to obtain in practice, meanwhile, these methods
Also without the information using system business usability testing, and these information have important value for effective fault diagnosis.
Summary of the invention
The present invention provides a kind of distributed system root Fault Locating Method and device, and the purpose is to be provided using the present invention
Test data construct trouble-shooting chart, Faults in Distributed Systems is rapidly and accurately diagnosed according to trouble-shooting chart, is effectively prevent
The puzzlement and economic loss that thrashing generates.
The purpose of the present invention is adopt the following technical solutions realization:
A kind of distributed system root Fault Locating Method, it is improved in that including:
Distributed system test data is obtained, and selects to be tested the not available test data of service in the test data；
Trouble-shooting chart is constructed using the not available test data of service is tested in the test data, and obtains the event
Hinder the complete subgraph in diagnostic graph；
Fault location is carried out according to the complete subgraph.
Preferably, the distributed system test data, comprising: test condition attribute and its corresponding test result category
Property, the test condition attribute include: external attribute and its corresponding built-in attribute, the external attribute include: test address,
Operator and tested service, the corresponding built-in attribute of the external attribute include: network device state, operation system state and answer
With program state, wherein the corresponding built-in attribute of the external attribute is obtained according to the external attribute using data flow tracing
It takes.
Further, the not available test data of service is tested in the selection test data, comprising:
If the attribute value of the corresponding test result attribute of the test condition attribute of the test data be it is unavailable, select
The test data.
It is preferably, described to construct trouble-shooting chart using the not available test data of service tested in the test data,
And obtain the complete subgraph in the trouble-shooting chart, comprising:
The not available test data of service is tested using in the test data as node, if the test number between node
The corresponding attribute value of attribute being tested in the not available test data of service in is identical, then by the corresponding category of the attribute between node
Property value connection, obtain trouble-shooting chart；
The connection figure between the corresponding single attribute value of single attribute of the trouble-shooting chart interior joint is extracted as complete
Subgraph.
It is preferably, described that fault location is carried out according to the complete subgraph, comprising:
The reliability of the complete subgraph is verified, the complete subgraph for meeting reliability requirement is obtained；
Fault location is carried out according to the number of nodes of the complete subgraph.
Further, the reliability of the verifying complete subgraph, obtains the complete subgraph for meeting reliability requirement, packet
It includes:
F inspection is carried out to k-th of complete subgraph, determines the test value F of k-th of complete subgraph as the following formulak:
Fk=(SSAk/fSSAk)/(SSEk/fSSEk)
In above formula, SSAkFor quadratic sum in the group of k-th of complete subgraph, fSSAkFor SSAkFreedom degree, SSEkIt is k-th
The sum of squares between groups of complete subgraph, fSSEkFor SSEkFreedom degree；
If the test value F of k-th of complete subgraphkGreater than F test value threshold value, then k-th of complete subgraph meets
Reliability requirement.
Further, quadratic sum SSA in the group of k-th of complete subgraph is determined as the following formulak:
The sum of squares between groups SSE of k-th of complete subgraph is determined as the following formulak:
nkFor the number of the identical test data of attribute value of attribute corresponding with k-th of complete subgraph,For with k-th
The attribute value of the test result of the identical test data of attribute value of the corresponding attribute of complete subgraph is not available occupation rate,
Attribute value for test result in test data is not available occupation rate,For attribute corresponding with k-th of complete subgraph
The number of the different test data of attribute value,For the different test of the attribute value of attribute corresponding from k-th of complete subgraph
The attribute value of the test result of data is not available occupation rate, xkMiFor the attribute value of attribute corresponding with k-th of complete subgraph
The attribute value coefficient of the test result of identical test data,For the attribute value of attribute corresponding with k-th of complete subgraph
The attribute value coefficient of the test result of different test datas, it is right when the attribute value of the test result of test data is available
The attribute value coefficient of the test result for the test data answered is 0, when the attribute value of the test result of test data is unavailable,
The attribute value coefficient of the test result of corresponding test data is 1.
It is further, described that fault location is carried out according to the number of nodes of the complete subgraph, comprising:
Obtain the maximum complete subgraph C of complete subgraph interior joint number for meeting reliability requirement1, and by C1Corresponding attribute
Attribute value be added failure collection in；
Compare x-th of the complete subgraph C met in the complete subgraph of reliability requirementxNode and the C1Node
Coincidence factor, if the coincidence factor is greater than relevance threshold, by CxThe attribute value of corresponding attribute is added in the failure collection,
In, x ∈ [1, X], X are the complete subgraph sum for meeting reliability requirement；
Using the attribute value in the failure collection as failure cause.
A kind of distributed system root fault locator, it is improved in that described device includes:
Acquiring unit, for obtaining distributed system test data, and select in the test data be tested service can not
Test data；
Construction unit, for constructing fault diagnosis using the not available test data of service tested in the test data
Figure, and obtain the complete subgraph in the trouble-shooting chart；
Failure location unit, for carrying out fault location according to the complete subgraph.
Beneficial effects of the present invention:
Technical solution provided by the invention services not available test number using being tested in test data provided by the invention
According to building trouble-shooting chart, and the complete subgraph in the trouble-shooting chart is obtained, event is finally carried out according to the complete subgraph
Barrier positioning, is described and is analyzed to test sample using more relational graphs, can more intuitively showed in current system and fail
Incidence relation between record, relative to relatively conventional weighted graph, more relational graphs can preferably reflect in fault diagnosis
The directly exact incidence relation of failure node realizes while will have in failure by carrying out fault location to the relationship cluster of mistake
The attribute and attribute value of pass cluster and the cluster of failure logging (record), and attribute value has been able to explicitly embody in many cases
Root failure cause, the result after clustering in remaining situation can substantially position root location of fault；
Further, during fault location, the reliability of complete subgraph is determined using F inspection, is filtered out
The attribute value of certain attributes is since value is less, the situation of biggish complete subgraph easy to form, can will really with failure original
Because relevant complete subgraph screens.
Detailed description of the invention
Fig. 1 is a kind of flow chart of distributed system root Fault Locating Method of the present invention；
Fig. 2 is the structural schematic diagram for inventing a kind of distributed system root fault locator.
Specific embodiment
It elaborates with reference to the accompanying drawing to a specific embodiment of the invention.
In order to make the object, technical scheme and advantages of the embodiment of the invention clearer, below in conjunction with the embodiment of the present invention
In attached drawing, technical scheme in the embodiment of the invention is clearly and completely described, it is clear that described embodiment is
A part of the embodiment of the present invention, instead of all the embodiments.Based on the embodiments of the present invention, those of ordinary skill in the art
All other embodiment obtained without making creative work, shall fall within the protection scope of the present invention.
A kind of distributed system root Fault Locating Method provided by the invention, as shown in Figure 1, comprising:
101. obtaining distributed system test data, and select to be tested the not available test number of service in the test data
According to；
102. constructing trouble-shooting chart using the not available test data of service is tested in the test data, and obtain institute
State the complete subgraph in trouble-shooting chart；
103. carrying out fault location according to the complete subgraph.
Wherein, the distributed system test data, comprising: test condition attribute and its corresponding test result attribute,
The test condition attribute includes: external attribute and its corresponding built-in attribute, and the external attribute includes: test address, fortune
Seek quotient and tested service, the corresponding built-in attribute of the external attribute includes: network device state, operation system state and application
Program state, wherein the corresponding built-in attribute of the external attribute is obtained according to the external attribute using data flow tracing
It takes.
The test result attribute is the use of tested service of the simulation using distributed system as a result, if the tested service
It can use, then the attribute value of the test result attribute is available, if the tested service is unavailable, the test result attribute
Attribute value is unavailable.
Further, in the step 101, the tool that the not available test data of service is tested in the test data is selected
Body process includes:
If the attribute value of the corresponding test result attribute of the test condition attribute of the test data be it is unavailable, select
The test data.
Such as: distributed system test data, including test condition attribute and its corresponding test result attribute are obtained,
In, test condition attribute includes: external test data and internal monitoring data, and data are associated.External test data
Attribute include test address, operator and tested service；Internal monitoring data include network device state, operation system state
And Application Status；Enable xi=[xi1,xi2,…,xiM] indicate complete usability testing record, wherein xikIndicate i-th of survey
K-th of attribute value of trial record, M indicate attribute number；The data set of all test records is X={ xi, i=1 ..., | X |,
Wherein | X | indicate the record number in X, test data set example is as shown in table 1:
1 test data set example of table
Wherein, the data in table 1 represent 16 original test records, i.e., | X |=16, and every initial data has 7 surveys
Try conditional attribute, comprising: test ground, operator, tested service, system region, the network equipment, server and CPU occupation rate, 1
A test result attribute, attribute value are 1 or 0, when tested service is available, attribute value 0, when tested service is unavailable
When, attribute value 1；
It obtains and is tested in the test data after the not available test data of service, quilt in the test data need to be utilized
It surveys and services not available test data building trouble-shooting chart, and obtain the complete subgraph in the trouble-shooting chart, therefore, institute
State step 102, comprising:
The not available test data of service is tested using in the test data as node, if the test number between node
The corresponding attribute value of attribute being tested in the not available test data of service in is identical, then by the corresponding category of the attribute between node
Property value connection, obtain trouble-shooting chart；
The connection figure between the corresponding single attribute value of single attribute of the trouble-shooting chart interior joint is extracted as complete
Subgraph.
For example, in test data as shown in Table 1, will be tested in the test data the not available test data of service as
Node, including the 1st, 3,4 and 5 article of test data indicate 4 nodes, first attribute test in node attribute have 4
Value, each value form the complete subgraph that a number of nodes is 1, and for second attribute, node has 1 value, are formed
The complete subgraph that one number of nodes is 4, for third attribute, node has 2 values, formed a number of nodes for 3 it is complete
The complete subgraph that subgraph and a number of nodes are 1, for the 4th attribute, node has 1 value, and forming a number of nodes is 4
Complete subgraph, for the 5th attribute, node has 1 value, formed a number of nodes be 4 complete subgraph, for the 6th
A attribute, node have 2 values, form the complete subgraph that a number of nodes is 3 and the complete subgraph that a number of nodes is 1, right
In the 7th attribute, node has 3 values, forms the complete subgraph that a number of nodes is 2 and the complete son that two number of nodes are 1
Figure.Therefore, practical for the set of records ends of lasting failure in the example to share 14 complete subgraphs.
Further, fault location, therefore the step 103 need to be carried out according to the complete subgraph, comprising:
The reliability of the complete subgraph is verified, the complete subgraph for meeting reliability requirement is obtained；
Fault location is carried out according to the number of nodes of the complete subgraph.
Specifically, the reliability of the verifying complete subgraph, obtains the complete subgraph for meeting reliability requirement, packet
It includes:
F inspection is carried out to k-th of complete subgraph, determines the test value F of k-th of complete subgraph as the following formulak:
Fk=(SSAk/fSSAk)/(SSEk/fSSEk)
In above formula, SSAkFor quadratic sum in the group of k-th of complete subgraph, fSSAkFor SSAkFreedom degree, SSEkIt is k-th
The sum of squares between groups of complete subgraph, fSSEkFor SSEkFreedom degree；
Wherein, fSSEk=test data sum-fSSAk- 1, fSSAk=1；
If the test value F of k-th of complete subgraphkGreater than F test value threshold value, then k-th of complete subgraph meets
Reliability requirement.
Wherein, quadratic sum SSA in the group of k-th of complete subgraph is determined as the following formulak:
The sum of squares between groups SSE of k-th of complete subgraph is determined as the following formulak:
nkFor the number of the identical test data of attribute value of attribute corresponding with k-th of complete subgraph,For with kth
The attribute value of the test result of the identical test data of attribute value of the corresponding attribute of a complete subgraph is not available occupation rate,Attribute value for test result in test data is not available occupation rate,For attribute corresponding with k-th of complete subgraph
The different test data of attribute value number,For the different survey of the attribute value of attribute corresponding from k-th of complete subgraph
The attribute value for trying the test result of data is not available occupation rate, xkMiFor the attribute of attribute corresponding with k-th of complete subgraph
It is worth the attribute value coefficient of the test result of identical test data,For the attribute of attribute corresponding with k-th of complete subgraph
It is worth the attribute value coefficient of the test result of different test datas, when the attribute value of the test result of test data is available,
The attribute value coefficient of the test result of corresponding test data is 0, when the attribute value of the test result of test data is unavailable,
The attribute value coefficient of the test result of its corresponding test data is 1.
For example, second attribute value is that the record number of connection is in raw data set X in test data as shown in Table 1
nk=100,SSA=2.4067, SSE=25.4933,
Therefore F=14.9, and it is 20 that F, which examines threshold value, therefore does not have significant difference.
After judging complete subgraph reliability, it need to find out and service unavailable maximally related attribute and attribute value with tested, with
And node constitutes several complete subgraphs；Ideally, if vertex failure is as caused by identical event, with event phase
The attribute on the vertex of pass is by value having the same, therefore all vertex with same alike result value will form a complete subgraph；
Different failures forms different complete subgraphs, but the complete subgraph of certain different attributes may be since same fault causes
, it needs to be merged according to the correlation of subgraph, it is therefore, described fixed according to the number of nodes of complete subgraph progress failure
Position, comprising:
Obtain the maximum complete subgraph C of complete subgraph interior joint number for meeting reliability requirement1, and by C1Corresponding attribute
Attribute value be added failure collection in；
Such as: the descending sequence of the number of nodes of the complete subgraph in table 1 is that second attribute value is that (size is connection
4, F test values be 14.9), the 4th attribute value be L1 (size 4, F test value be 28.2), the 5th attribute value be N11 (greatly
It is small be 4, F test value be 119.8), third attribute value be that (size 3, for F test value for 3.6), the 6th attribute value is S11 to X
(size 3, F test value are that 55.0), the 7th attribute value is that low (size 2, F test value are 1.9) remaining eight attribute value
Come last (size 1), wherein meet the attribute of reliability requirement and attribute value be the 4th attribute value be L1, the 5th category
Property value is N11, the 6th attribute value is S11.
Compare x-th of the complete subgraph C met in the complete subgraph of reliability requirementxNode and the C1Node
Coincidence factor, if the coincidence factor is greater than relevance threshold, by CxThe attribute value of corresponding attribute is added in the failure collection,
In, x ∈ [1, X], X are the complete subgraph sum for meeting reliability requirement；
Using the attribute value in the failure collection as failure cause.
For example, when detecting first failure in the example of table 1, primary fault subgraph H1It is second attribute for connection structure
At complete subgraph, failure collection D1In only one attribute D1The number of nodes row of complete subgraph is compared in=[system region=L1]
In deputy attribute and attribute value, with H1Registration, if more than 70%, then D1In be included in the attribute and attribute value, i.e. D1=
[system region=L1, the network equipment=N11], compares the attribute and attribute value for coming third position, with H1Registration be greater than
70%, so D1In be included in the attribute and attribute value, i.e. D1=[system region=L1, the network equipment=N11, server=
S11], the complete subgraph of reliability, final D are met without other1=[system region=L1, the network equipment=N11, server
=S11].
Further, embodiment provided by the invention use in can also include excluding oriented failure cause, continue into
Row fault location, it is faulty until orienting, specifically: after step 103 navigates to failure cause, remove the failure cause
Test record repeats step 101 to step 103 in remaining test record set, meets D for example, removing in table 11Condition
Record, subsequent detection failure subgraph, failure logging (record) if it exists, then repeat step 101 to step 103 obtain failure cause collection
It closes, until not having failure logging (record), fault detection is finished.
The present invention also provides a kind of distributed system root fault locators, as shown in Fig. 2, described device includes:
Acquiring unit, for obtaining distributed system test data, and select in the test data be tested service can not
Test data；
Construction unit, for constructing fault diagnosis using the not available test data of service tested in the test data
Figure, and obtain the complete subgraph in the trouble-shooting chart；
Failure location unit, for carrying out fault location according to the complete subgraph.
Wherein, the distributed system test data, comprising: test condition attribute and its corresponding test result attribute,
The test condition attribute includes: external attribute and its corresponding built-in attribute, and the external attribute includes: test address, fortune
Seek quotient and tested service, the corresponding built-in attribute of the external attribute includes: network device state, operation system state and application
Program state, wherein the corresponding built-in attribute of the external attribute is obtained according to the external attribute using data flow tracing
It takes.
The acquiring unit, comprising:
Selecting module, if the attribute value of the corresponding test result attribute of test condition attribute for the test data is
It is unavailable, then select the test data.
The construction unit, comprising:
Module is constructed, for being tested the not available test data of service using in the test data as node, if node
Between the test data in the corresponding attribute value of attribute that is tested in the not available test data of service it is identical, then will be between node
The corresponding attribute value connection of the attribute, obtains trouble-shooting chart；
Extraction module, the company between the corresponding single attribute value of single attribute for extracting the trouble-shooting chart interior joint
Map interlinking is as complete subgraph.
The failure location unit, comprising:
Module is obtained, for verifying the reliability of the complete subgraph, obtains the complete subgraph for meeting reliability requirement；
Fault location module, for carrying out fault location according to the number of nodes of the complete subgraph.
Wherein, F inspection is carried out to k-th of complete subgraph, determines the test value F of k-th of complete subgraph as the following formulak:
Fk=(SSAk/fSSAk)/(SSEk/fSSEk)
In above formula, SSAkFor quadratic sum in the group of k-th of complete subgraph, fSSAkFor SSAkFreedom degree, SSEkIt is k-th
The sum of squares between groups of complete subgraph, fSSEkFor SSEkFreedom degree；
If the test value F of k-th of complete subgraphkGreater than F test value threshold value, then k-th of complete subgraph meets
Reliability requirement.
Quadratic sum SSA in the group of k-th of complete subgraph is determined as the following formulak:
The sum of squares between groups SSE of k-th of complete subgraph is determined as the following formulak:
nkFor the number of the identical test data of attribute value of attribute corresponding with k-th of complete subgraph,For with k-th
The attribute value of the test result of the identical test data of attribute value of the corresponding attribute of complete subgraph is not available occupation rate,
Attribute value for test result in test data is not available occupation rate,For attribute corresponding with k-th of complete subgraph
The number of the different test data of attribute value,For the different test of the attribute value of attribute corresponding from k-th of complete subgraph
The attribute value of the test result of data is not available occupation rate, xkMiFor the attribute value of attribute corresponding with k-th of complete subgraph
The attribute value coefficient of the test result of identical test data,For the attribute value of attribute corresponding with k-th of complete subgraph
The attribute value coefficient of the test result of different test datas, it is right when the attribute value of the test result of test data is available
The attribute value coefficient of the test result for the test data answered is 0, when the attribute value of the test result of test data is unavailable,
The attribute value coefficient of the test result of corresponding test data is 1.
The fault location module, is also used to:
Obtain the maximum complete subgraph C of complete subgraph interior joint number for meeting reliability requirement1, and by C1Corresponding attribute
Attribute value be added failure collection in；
Compare x-th of the complete subgraph C met in the complete subgraph of reliability requirementxNode and the C1Node
Coincidence factor, if the coincidence factor is greater than relevance threshold, by CxThe attribute value of corresponding attribute is added in the failure collection,
In, x ∈ [1, X], X are the complete subgraph sum for meeting reliability requirement；
Using the attribute value in the failure collection as failure cause.
It should be understood by those skilled in the art that, embodiments herein can provide as method, system or computer program
Product.Therefore, complete hardware embodiment, complete software embodiment or reality combining software and hardware aspects can be used in the application
Apply the form of example.Moreover, it wherein includes the computer of computer usable program code that the application, which can be used in one or more,
The computer program implemented in usable storage medium (including but not limited to magnetic disk storage, CD-ROM, optical memory etc.) produces
The form of product.
The application is referring to method, the process of equipment (system) and computer program product according to the embodiment of the present application
Figure and/or block diagram describe.It should be understood that every one stream in flowchart and/or the block diagram can be realized by computer program instructions
The combination of process and/or box in journey and/or box and flowchart and/or the block diagram.It can provide these computer programs
Instruct the processor of general purpose computer, special purpose computer, Embedded Processor or other programmable data processing devices to produce
A raw machine, so that being generated by the instruction that computer or the processor of other programmable data processing devices execute for real
The device for the function of being specified in present one or more flows of the flowchart and/or one or more blocks of the block diagram.
These computer program instructions, which may also be stored in, is able to guide computer or other programmable data processing devices with spy
Determine in the computer-readable memory that mode works, so that it includes referring to that instruction stored in the computer readable memory, which generates,
Enable the manufacture of device, the command device realize in one box of one or more flows of the flowchart and/or block diagram or
The function of being specified in multiple boxes.
These computer program instructions also can be loaded onto a computer or other programmable data processing device, so that counting
Series of operation steps are executed on calculation machine or other programmable devices to generate computer implemented processing, thus in computer or
The instruction executed on other programmable devices is provided for realizing in one or more flows of the flowchart and/or block diagram one
The step of function of being specified in a box or multiple boxes.
Finally it should be noted that: the above embodiments are merely illustrative of the technical scheme of the present invention and are not intended to be limiting thereof, to the greatest extent
Invention is explained in detail referring to above-described embodiment for pipe, it should be understood by those ordinary skilled in the art that: still
It can be with modifications or equivalent substitutions are made to specific embodiments of the invention, and without departing from any of spirit and scope of the invention
Modification or equivalent replacement, should all cover within the scope of the claims of the present invention.
Claims (9)
1. a kind of distributed system root Fault Locating Method, which is characterized in that the described method includes:
Distributed system test data is obtained, and selects to be tested the not available test data of service in the test data；
Trouble-shooting chart is constructed using the not available test data of service is tested in the test data, and obtains the failure and examines
Complete subgraph in disconnected figure；
Fault location is carried out according to the complete subgraph.
2. the method as described in claim 1, which is characterized in that the distributed system test data, comprising: test condition category
Property and its corresponding test result attribute, the test condition attribute include: external attribute and its corresponding built-in attribute, described
External attribute includes: test address, operator and tested service, and the corresponding built-in attribute of the external attribute includes: that network is set
Standby state, operation system state and Application Status, wherein the corresponding built-in attribute of the external attribute is according to the outside
Attribute is obtained using data flow tracing.
3. method according to claim 2, which is characterized in that it is not available to be tested service in the selection test data
Test data, comprising:
If the attribute value of the corresponding test result attribute of the test condition attribute of the test data be it is unavailable, select the survey
Try data.
4. the method as described in claim 1, which is characterized in that described not available using service is tested in the test data
Test data constructs trouble-shooting chart, and obtains the complete subgraph in the trouble-shooting chart, comprising:
The not available test data of service is tested using in the test data as node, if in the test data between node
The corresponding attribute value of attribute in the tested not available test data of service is identical, then by the corresponding attribute value of the attribute between node
Connection obtains trouble-shooting chart；
The connection figure between the corresponding single attribute value of single attribute of the trouble-shooting chart interior joint is extracted as complete subgraph.
5. the method as described in claim 1, which is characterized in that described to carry out fault location according to the complete subgraph, comprising:
The reliability of the complete subgraph is verified, the complete subgraph for meeting reliability requirement is obtained；
Fault location is carried out according to the number of nodes of the complete subgraph.
6. method as claimed in claim 5, which is characterized in that the reliability of the verifying complete subgraph obtains and meets
The complete subgraph of reliability requirement, comprising:
F inspection is carried out to k-th of complete subgraph, determines the test value F of k-th of complete subgraph as the following formulak:
Fk=(SSAk/fSSAk)/(SSEk/fSSEk)
In above formula, SSAkFor quadratic sum in the group of k-th of complete subgraph, fSSAkFor SSAkFreedom degree, SSEkIt is completely sub for k-th
The sum of squares between groups of figure, fSSEkFor SSEkFreedom degree；
If the test value F of k-th of complete subgraphkGreater than F test value threshold value, then k-th of complete subgraph meets reliability
It is required that.
7. method as claimed in claim 6, which is characterized in that determine interior square of the group of k-th of complete subgraph as the following formula
And SSAk:
The sum of squares between groups SSE of k-th of complete subgraph is determined as the following formulak:
nkFor the number of the identical test data of attribute value of attribute corresponding with k-th of complete subgraph,For with k-th completely
The attribute value of the test result of the identical test data of attribute value of the corresponding attribute of subgraph is not available occupation rate,To survey
The attribute value for trying test result in data is not available occupation rate,For the attribute of attribute corresponding with k-th of complete subgraph
It is worth the number of different test datas,For the different test data of the attribute value of attribute corresponding from k-th of complete subgraph
Test result attribute value be not available occupation rate, xkMiAttribute value for attribute corresponding with k-th of complete subgraph is identical
Test data test result attribute value coefficient,Attribute value for attribute corresponding from k-th of complete subgraph is different
Test data test result attribute value coefficient, it is corresponding when the attribute value of the test result of test data is available
The attribute value coefficient of the test result of test data is 0, when the attribute value of the test result of test data is unavailable, is corresponded to
Test data test result attribute value coefficient be 1.
8. method as claimed in claim 5, which is characterized in that described fixed according to the number of nodes of complete subgraph progress failure
Position, comprising:
Obtain the maximum complete subgraph C of complete subgraph interior joint number for meeting reliability requirement1, and by C1The category of corresponding attribute
Property value be added failure collection in；
Compare x-th of the complete subgraph C met in the complete subgraph of reliability requirementxNode and the C1Node coincidence
Rate, if the coincidence factor is greater than relevance threshold, by CxThe attribute value of corresponding attribute is added in the failure collection, wherein x
∈ [1, X], X are the complete subgraph sum for meeting reliability requirement；
Using the attribute value in the failure collection as failure cause.
9. a kind of distributed system root fault locator, which is characterized in that described device includes:
Acquiring unit for obtaining distributed system test data, and selects tested service in the test data not available
Test data；
Construction unit, for constructing trouble-shooting chart using the not available test data of service tested in the test data, and
Obtain the complete subgraph in the trouble-shooting chart；
Failure location unit, for carrying out fault location according to the complete subgraph.
Priority Applications (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN201710801677.9A CN109474445B (en) | 2017-09-07 | 2017-09-07 | Distributed system root fault positioning method and device |
Applications Claiming Priority (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN201710801677.9A CN109474445B (en) | 2017-09-07 | 2017-09-07 | Distributed system root fault positioning method and device |
Publications (2)
| Publication Number | Publication Date |
|---|---|
| CN109474445A true CN109474445A (en) | 2019-03-15 |
| CN109474445B CN109474445B (en) | 2022-08-19 |
Family
ID=65658061
Family Applications (1)
| Application Number | Title | Priority Date | Filing Date |
|---|---|---|---|
| CN201710801677.9A Active CN109474445B (en) | 2017-09-07 | 2017-09-07 | Distributed system root fault positioning method and device |
Country Status (1)
| Country | Link |
|---|---|
| CN (1) | CN109474445B (en) |
Citations (4)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN101042714A (en) * | 2007-04-29 | 2007-09-26 | 哈尔滨工业大学 | Compressing method for SOC testing data suitable for suitable for multi-scanning chain designing core |
| CN101330417A (en) * | 2008-07-24 | 2008-12-24 | 安徽大学 | Quotient space overlay model for calculating network shortest path and building method thereof |
| CN103209094A (en) * | 2013-03-11 | 2013-07-17 | 中国科学院信息工程研究所 | System and method for fault positioning on basis of events |
| CN103914064A (en) * | 2014-04-01 | 2014-07-09 | 浙江大学 | Industrial process fault diagnosis method based on multiple classifiers and D-S evidence fusion |
-
2017
- 2017-09-07 CN CN201710801677.9A patent/CN109474445B/en active Active
Patent Citations (4)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN101042714A (en) * | 2007-04-29 | 2007-09-26 | 哈尔滨工业大学 | Compressing method for SOC testing data suitable for suitable for multi-scanning chain designing core |
| CN101330417A (en) * | 2008-07-24 | 2008-12-24 | 安徽大学 | Quotient space overlay model for calculating network shortest path and building method thereof |
| CN103209094A (en) * | 2013-03-11 | 2013-07-17 | 中国科学院信息工程研究所 | System and method for fault positioning on basis of events |
| CN103914064A (en) * | 2014-04-01 | 2014-07-09 | 浙江大学 | Industrial process fault diagnosis method based on multiple classifiers and D-S evidence fusion |
Non-Patent Citations (2)
| Title |
|---|
| IRINA RISH: "Adaptive Diagnosis in Distributed Systems", 《IEEE TRANSACTIONS ON NEURAL NETWORKS,》 * |
| 李建新: "求最大完全子图的启发式着色算法", 《滁州学院学报》 * |
Also Published As
| Publication number | Publication date |
|---|---|
| CN109474445B (en) | 2022-08-19 |
Similar Documents
| Publication | Publication Date | Title |
|---|---|---|
| CN107291911A (en) | A kind of method for detecting abnormality and device | |
| CN109840157A (en) | Method, apparatus, electronic equipment and the storage medium of fault diagnosis | |
| CN106165345A (en) | Mark is for dissolving the failture evacuation option of network failure | |
| CN106776208B (en) | Fault Locating Method when a kind of running software | |
| Friederich et al. | Towards data-driven reliability modeling for cyber-physical production systems | |
| CN107168842A (en) | Adaptive sequential fault diagnosis method based on pmc model | |
| CN107291063A (en) | Diagnostic device and diagnostic method for the operation of monitoring technology facility | |
| CN113391943A (en) | Micro-service fault root cause positioning method and device based on cause and effect inference | |
| Lazarova-Molnar et al. | Data-driven fault tree modeling for reliability assessment of cyber-physical systems | |
| Honsel et al. | Mining software dependency networks for agent-based simulation of software evolution | |
| CN107966648B (en) | A kind of embedded failure diagnosis method based on correlation matrix | |
| Atzmueller et al. | Anomaly detection and structural analysis in industrial production environments | |
| CN110472866A (en) | A kind of work order quality inspection analysis method and device | |
| Pauwels et al. | Detecting and explaining drifts in yearly grant applications | |
| CN117221087A (en) | Alarm root cause positioning method, device and medium | |
| Dhiman et al. | A Clustered Approach to Analyze the Software Quality Using Software Defects | |
| CN105159826B (en) | A kind of method and apparatus of wrong sentence in positioning target program | |
| CN109889258B (en) | Optical network fault checking method and equipment | |
| CN106569944A (en) | Constraint-tree-based onboard software test data analysis method | |
| CN109474445A (en) | A kind of distributed system root Fault Locating Method and device | |
| CN110188040A (en) | A kind of software platform for software systems fault detection and health state evaluation | |
| CN106487592B (en) | A kind of Faults in Distributed Systems diagnostic method based on data cube | |
| CN113033845B (en) | Construction method and device for power transmission resource co-construction and sharing | |
| CN114385403A (en) | Distributed cooperative fault diagnosis method based on double-layer knowledge graph framework | |
| KR102217092B1 (en) | Method and apparatus for providing quality information of application |
Legal Events
| Date | Code | Title | Description |
|---|---|---|---|
| PB01 | Publication | ||
| PB01 | Publication | ||
| SE01 | Entry into force of request for substantive examination | ||
| SE01 | Entry into force of request for substantive examination | ||
| GR01 | Patent grant | ||
| GR01 | Patent grant |