CN112988529A - Method and system for predicting database system performance based on machine learning - Google Patents
Method and system for predicting database system performance based on machine learning Download PDFInfo
- Publication number
- CN112988529A CN112988529A CN201911296302.7A CN201911296302A CN112988529A CN 112988529 A CN112988529 A CN 112988529A CN 201911296302 A CN201911296302 A CN 201911296302A CN 112988529 A CN112988529 A CN 112988529A
- Authority
- CN
- China
- Prior art keywords
- database system
- database
- evaluation value
- vector
- system performance
- 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
- 238000000034 method Methods 0.000 title claims abstract description 36
- 238000010801 machine learning Methods 0.000 title claims abstract description 20
- 239000013598 vector Substances 0.000 claims abstract description 72
- 238000011156 evaluation Methods 0.000 claims abstract description 56
- 238000012549 training Methods 0.000 claims abstract description 21
- 238000012423 maintenance Methods 0.000 claims abstract description 20
- 230000004044 response Effects 0.000 claims description 10
- 230000008859 change Effects 0.000 claims description 7
- 230000006870 function Effects 0.000 claims description 7
- 230000008569 process Effects 0.000 claims description 5
- 238000011478 gradient descent method Methods 0.000 claims description 3
- 238000012545 processing Methods 0.000 description 5
- 238000010586 diagram Methods 0.000 description 4
- 238000005516 engineering process Methods 0.000 description 4
- 238000005070 sampling Methods 0.000 description 4
- 238000004422 calculation algorithm Methods 0.000 description 3
- 230000001960 triggered effect Effects 0.000 description 3
- 238000013480 data collection Methods 0.000 description 2
- 238000011161 development Methods 0.000 description 2
- 238000012544 monitoring process Methods 0.000 description 2
- 238000013515 script Methods 0.000 description 2
- 230000004075 alteration Effects 0.000 description 1
- 238000013459 approach Methods 0.000 description 1
- 230000009286 beneficial effect Effects 0.000 description 1
- 238000013461 design Methods 0.000 description 1
- 230000000694 effects Effects 0.000 description 1
- 238000007689 inspection Methods 0.000 description 1
- 238000012986 modification Methods 0.000 description 1
- 230000004048 modification Effects 0.000 description 1
- 230000000737 periodic effect Effects 0.000 description 1
- 238000011084 recovery Methods 0.000 description 1
Images
Classifications
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06F—ELECTRIC DIGITAL DATA PROCESSING
- G06F11/00—Error detection; Error correction; Monitoring
- G06F11/30—Monitoring
- G06F11/34—Recording or statistical evaluation of computer activity, e.g. of down time, of input/output operation ; Recording or statistical evaluation of user activity, e.g. usability assessment
- G06F11/3452—Performance evaluation by statistical analysis
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06N—COMPUTING ARRANGEMENTS BASED ON SPECIFIC COMPUTATIONAL MODELS
- G06N3/00—Computing arrangements based on biological models
- G06N3/02—Neural networks
- G06N3/04—Architecture, e.g. interconnection topology
- G06N3/045—Combinations of networks
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06N—COMPUTING ARRANGEMENTS BASED ON SPECIFIC COMPUTATIONAL MODELS
- G06N3/00—Computing arrangements based on biological models
- G06N3/02—Neural networks
- G06N3/08—Learning methods
- G06N3/084—Backpropagation, e.g. using gradient descent
Landscapes
- Engineering & Computer Science (AREA)
- Physics & Mathematics (AREA)
- Theoretical Computer Science (AREA)
- General Engineering & Computer Science (AREA)
- Life Sciences & Earth Sciences (AREA)
- General Physics & Mathematics (AREA)
- Computational Linguistics (AREA)
- Health & Medical Sciences (AREA)
- Data Mining & Analysis (AREA)
- Evolutionary Computation (AREA)
- General Health & Medical Sciences (AREA)
- Molecular Biology (AREA)
- Computing Systems (AREA)
- Biomedical Technology (AREA)
- Artificial Intelligence (AREA)
- Mathematical Physics (AREA)
- Software Systems (AREA)
- Biophysics (AREA)
- Bioinformatics & Cheminformatics (AREA)
- Bioinformatics & Computational Biology (AREA)
- Evolutionary Biology (AREA)
- Probability & Statistics with Applications (AREA)
- Computer Hardware Design (AREA)
- Quality & Reliability (AREA)
- Information Retrieval, Db Structures And Fs Structures Therefor (AREA)
- Debugging And Monitoring (AREA)
- Feedback Control In General (AREA)
- Testing And Monitoring For Control Systems (AREA)
Abstract
The invention provides a method and a system for predicting database system performance based on machine learning. The method comprises the following steps: collecting operation and maintenance data of a database and forming a multi-dimensional state vector of each time period; calculating a trend vector across a plurality of time periods based on the multi-dimensional state vector; acquiring the performance indexes of the database system corresponding to each time period; calculating a system evaluation value based on the acquired database system performance index; training a database system performance prediction model based on the trend vector and the system evaluation value; and inputting the trend vector formed in real time into the database system performance prediction model to output a predicted system evaluation value.
Description
Technical Field
The invention relates to the field of database operation and maintenance, in particular to a method and a system for database system performance prediction based on machine learning.
Background
With the continuous popularization of internet technology and the internet expansion of various service fields, an open LMAP architecture has gradually replaced an early IOE architecture. Compared with a commercial database, the open source database is more open and contained, and is more beneficial to quick iteration of services, wherein MySQL is outstanding as the open source database, has the characteristics of high performance, low cost, easiness in deployment and the like, gradually becomes a mainstream database solution, and is widely applied to various service fields.
The database is used as a core function of a business system, and is strongly related to business logic, and generally relates to multiple links such as system design, development, deployment, operation and maintenance, and the like. The main problems with this approach are: (1) depending on years of operation and maintenance experience of database experts (DBA), whether the system is stable and the fault recovery time are strongly related to the DBA technology level; (2) the existing database operation and maintenance technology has hysteresis, manual intervention processing is carried out only when performance is reduced or even faults occur, and the prediction capability of the performance of a database system is lacked; (3) due to the mutual cooperation of a plurality of teams related to database, development, operation and maintenance and the like, the fault processing speed is too slow.
For these current situations, some third-party tools or plug-in platforms are available in the market to improve the operation and maintenance efficiency of the database, such as a monitoring system, automated inspection and the like, and although the efficiency of finding problems is improved, the following limitations still exist: (1) the method can be processed only when a fault occurs or the performance index of the database system is reduced to reach a threshold value, cannot be predicted according to the trend, and has no prospect; (2) database systems are usually strongly related to business logic, and third-party tools are not suitable for large-scale complex systems with high relevance; (3) although the third-party tool has some unified basic monitoring, the operation and maintenance of the database for different service scenes still depends on the expert experience of the database and is customized by self-writing scripts and defining parameter weights, and the third-party tool has no objectivity and universality.
Disclosure of Invention
This summary is provided to introduce a selection of concepts in a simplified form that are further described below in the detailed description. This summary is not intended to identify key features or essential features of the claimed subject matter, nor is it intended to be used as an aid in determining the scope of the claimed subject matter.
In order to solve the limitation of the current database operation and maintenance technology, the invention provides a method which works in a system, can accurately judge the IO performance trend of a database during high-dimensional input and can predict the IO performance of the database for a period of time in the future. According to the method, database parameter data and system IO performance index data are collected, model training is carried out by adopting an SOFTMAX multi-classifier, and the IO performance of the database in the future time is predicted.
The method does not depend on the experience of personnel, all input data and model revisions are actual operation indexes of the database, the IO performance trend of the database can be objectively predicted, meanwhile, the method is stable and efficient, the machine learning algorithm can be automatically adjusted and optimized through continuous online learning, the prediction accuracy is improved, and the method has important reference significance for improving the operation and maintenance efficiency of the database.
According to one aspect of the invention, a method for database system performance prediction based on machine learning is provided, the method comprising:
collecting database operation and maintenance data to form a multi-dimensional state vector of each time period;
calculating a trend vector across a plurality of time periods based on the multi-dimensional state vector;
acquiring the performance indexes of the database system corresponding to each time period;
calculating a system evaluation value based on the acquired database system performance index;
training a database system performance prediction model based on the trend vector and the system evaluation value; and
and inputting the trend vector formed in real time into the database system performance prediction model to output a predicted system evaluation value.
According to one embodiment of the invention, calculating a trend vector over a plurality of time periods based on the multi-dimensional state vector further comprises:
and splicing the state change vector of each pair of adjacent time periods of the plurality of involved time periods and the multidimensional state vector of the last time period of the plurality of involved time periods to form involved trend vectors, wherein the state change vectors of the adjacent time periods are the difference between the multidimensional state vector of the next time period and the multidimensional state vector of the previous time period.
According to a further embodiment of the present invention, calculating the system evaluation value based on the obtained database system performance index further comprises:
comparing the obtained database system performance index with a corresponding reference value; and
the system evaluation value is determined as one of a plurality of equal-place values set in advance based on the comparison result.
According to a further embodiment of the present invention, the database system performance prediction model is based on a SOFTMAX classifier, and training the database system performance prediction model based on the trend vector and the system evaluation value further comprises:
and respectively taking a multidimensional trend vector and a system evaluation value corresponding to the same time period as the input and the output of the SOFTMAX classifier.
According to a further embodiment of the invention, the training process is performed online in the following manner:
inputting the trend vector into the SOFTMAX classifier;
the SOFTMAX classifier outputs a prediction result based on the current weight;
performing backward propagation on the real system evaluation value corresponding to the period of the input trend vector, and deriving a loss function;
minimizing the loss function at the fastest rate by a gradient descent method;
updating the weights of the SOFTMAX classifier with derivatives; and
and repeating the above steps.
According to a further embodiment of the invention, the method further comprises: and triggering preset early warning conditions in response to the predicted system evaluation value output by the database system performance prediction model, and sending out early warning.
According to a further embodiment of the invention, the method further comprises: and triggering a preset alarm condition in response to the calculated system evaluation value, and sending an alarm.
According to another aspect of the present invention, there is provided a system for database system performance prediction based on machine learning, the system comprising:
a data acquisition module configured to:
collecting database operation and maintenance data to form a multi-dimensional state vector of each time period; and;
calculating a trend vector across a plurality of time periods based on the multi-dimensional state vector;
a database system performance evaluation module configured to:
acquiring the performance indexes of the database system corresponding to each time period; and
calculating a system evaluation value based on the acquired database system performance index;
a machine learning module configured to train a database system performance prediction model based on the trend vector and the system evaluation value;
a database system performance prediction module configured to provide a predicted system evaluation value using a trained database system performance prediction model; and
an alert module configured to:
triggering preset early warning conditions in response to the predicted system evaluation value output by the database system performance prediction model, and sending out early warning; and
and triggering a preset alarm condition in response to the calculated system evaluation value, and sending an alarm.
According to one embodiment of the invention, the database system is a MySQL database system.
According to a further embodiment of the present invention, the database system performance indicators include database IO performance indicators.
Compared with the scheme in the prior art, the database system performance prediction method provided by the invention at least has the following advantages:
(1) has objectivity. The invention works in the database system, the acquired data come from the database system, and the personal experience of the database operation and maintenance personnel is not depended on;
(2) it is predictive. The method can analyze and actively process the performance trend of the database system, and has great effect on effectively avoiding and processing the database fault;
(3) has universality. Because the database system environment and performance requirements of different services are different, compared with a manually set unified quality judgment model, the prediction function can be automatically adapted under various working environments by using a machine learning algorithm.
(4) The online learning capability is provided. The IO performance of the current MySQL database is used as a module for machine learning through system evaluation, and the module can be used for training, so that a machine learning algorithm is continuously and automatically adjusted and optimized, and the prediction accuracy is finally continuously improved.
These and other features and advantages will become apparent upon reading the following detailed description and upon reference to the accompanying drawings. It is to be understood that both the foregoing general description and the following detailed description are explanatory only and are not restrictive of aspects as claimed.
Drawings
So that the manner in which the above recited features of the present invention can be understood in detail, a more particular description of the invention, briefly summarized above, may be had by reference to embodiments, some of which are illustrated in the appended drawings. It is to be noted, however, that the appended drawings illustrate only some typical aspects of this invention and are therefore not to be considered limiting of its scope, for the description may admit to other equally effective aspects.
FIG. 1 is an exemplary architecture diagram of a database system performance prediction system, according to one embodiment of the invention.
FIG. 2 is an overall flow diagram of a method for machine learning-based prediction of the IO performance of a MySQL database according to one embodiment of the invention.
Detailed Description
The present invention will be described in detail below with reference to the attached drawings, and the features of the present invention will be further apparent from the following detailed description.
FIG. 1 is an exemplary architecture diagram of a database system performance prediction system, according to one embodiment of the invention. As shown in fig. 1, the database system performance prediction system of the present invention may include: a data collection module 102, a database system performance evaluation module 104, a machine learning module 106, a database system performance prediction module 108, and an alarm module 110. As mentioned above, MySQL is the mainstream open source database, the application range is the widest, a mature and general scheme has been formed for the indexes of MySQL database in the industry, most of other open source relational databases in the market are branches of MySQL evolution, and the parameter indexes have general value. Thus, the database system herein may be a MySQL database system, however, those skilled in the art will appreciate that the database system performance prediction of the present invention is not limited to MySQL databases. Further, database system performance may include various database system performance indicators such as database system IO performance instructions. For ease of explanation, the principles and specific operation of the present invention will be described below directly using the MySQL database and database IO system.
FIG. 2 is an overall flow diagram of a method for machine learning-based prediction of the IO performance of a MySQL database according to one embodiment of the invention.
First, in step 1, database operation and maintenance data is collected. More specifically, as one example, collecting MySQL database operation data may include collecting database operational parameters for which IO performance is affected, sampling by a particular script at fixed time intervals (e.g., 5 minutes), and forming a multidimensional state vector S. The collected operation and maintenance data may include, but is not limited to: the method comprises the following parameters of processing request number in unit Time (TPS), query request number in unit time (QPS), thread connection number, thread hit rate, MyISAM cache hit rate, INNODB cache hit rate, temporary table use condition, binlog cache use condition, JOIN operation information, slow query, table scanning times, index utilization rate, lock waiting condition and the like. Correspondingly, the multidimensional state vector S ═ number of processing requests per unit Time (TPS), number of query requests per unit time (QPS), number of thread connections, thread hit rate, MyISAM cache hit rate, inodb cache hit rate, temporary table usage, binlog cache usage, JOIN operation information, slow lookup, number of table scans, index utilization, lock wait. As a preferred embodiment, under a specific period T (the period T is larger than the sampling time interval, and the unit is also assumed to be minutes), the average of the sampling results performed multiple times (i.e. T/5 times) in the period T can be calculated, and the average is used as the state vector s (T) in the period T.
In step 2, the multi-dimensional state vector S for each cycle is stored and a trend vector across cycles is calculated. For example, by 3 adjacent cycles, a trend vector over 3 cycles can be computed:
1) assuming that the T1 period is adjacent in time to the T2 period and T2 is after T1, the state change vector of T1 to T2 is defined as V12=S(t2)-S(t1). Similarly, the state change vectors of T2-T3 are V23=S(t3)-S(t2)。
2) By concatenating the involved state change vector and the latest state vector, trend vectors of T1, T2, T3 are obtained, which may be defined as { V }12,V23,S3}。
And 3, acquiring one or more performance index data of the database system. Taking IO performance as an example, the IO performance indicator data may include, but is not limited to: I/O number per second ([ r/s w/s ]), throughput per second ([ rkb/s wkb/s ]), I/O wait queue length (avgq-sz), wait time (await), service time (svctm), disk active time percentage (% util), and the like. Similarly, if the period is T and the sampling interval is 30 minutes, T/30 system index acquisitions can be performed in the period. Similarly, the average of the indexes acquired for the multiple times can be used as the performance index data of the database system in the period T.
And 4, obtaining a system evaluation value by comparing the acquired one or more performance index data with a reference value. For example, the obtained performance index data may be compared with a system hardware factory theory IO performance index to obtain a system evaluation value. Taking the IO performance indicator data mentioned above as an example,
1) the evaluation characteristic value of the periodic system is obtained by the following formula 1:
(t) max (K ([ r/sw/s ], K ([ rkb/swkb/s ], K (avgq-sz), K (await), K (svctm),% util ] formula 1
The expression 1 indicates that the performance percentage of each performance index data can be calculated separately, and then the maximum value of the performance percentages of all the index data is taken as the system evaluation value.
2) The classification result fitness (t) of the system evaluation value in this cycle is obtained by equation 2:
equation 2 above is merely an example of a system evaluation value classification. The five specific classification results of the system evaluation values "excellent", "good", "medium", and "poor" each represent one classification of the system evaluation, and may be replaced with 1, 2, 3, 4, and 5 or any other index value. Those skilled in the art will understand that different classification names, classification numbers and classification standards can be set according to actual needs.
And 5, training a database system performance early warning model based on the trend vector obtained in the step 2 and the system evaluation value obtained in the step 4. More specifically, the obtained trend vector of the corresponding cycle and the system evaluation value are made into training data to perform machine learning training.
As an example, the trend vector and the corresponding system merit values may be trained as input and output, respectively, to the SOFTMAX classifier.
Assuming a set of four periods T1, T2, T3, T4, the model predicts the system rating of T4 with the states of T1, T2, T3: the trend vector is then { V12,V23,S3A real system evaluation value associated with the trend vector is FITNESS (4), and the two data form a set of training data; similarly, assuming another set of four periods T2, T3, T4, T5, the model predicts the system merit value of T5 with the states of T2, T3, T4: the input trend vector is { V23,V34,S4The real system associated with this trend vector is evaluated as FITNESS (5), which also constitutes a set of training data.
The machine learning training process can be performed either off-line or on-line. Off-line training can collect a large amount of training data to train the SOFTMAX classifier, and each weight of the SOFTMAX classifier can be obtained.
According to one example of the present invention, the SOFTMAX classifier is trained in an online training manner. More specifically, for example, when the period T3 is completed, V may be obtained23And S3Then the input trend vector can be expressed as { V }12,V23,S3These trend vectors may be input to the SOFTMAX classifier, which outputs the prediction result FITNESS (4), i.e., the prediction for T4 cycles. It can be understood that the FITNESS (4) is a predicted value, and at the initial stage of training, due to the insufficient amount of training data, the weights of the SOFTMAX classifier may be default weights, and there is a large deviation between the predicted value and the true value. However, at the end of the period T4, the system can obtain a true system rating. At this time, the classifier can be further trained by associating the feature vector and the T4 cycle truth evaluation, which is as follows: with true evaluation (classification) back-propagation, the loss function is derived (gradient), then minimized at the fastest rate by the gradient descent method, and then the existing weights are updated with the derivatives. By repeatedly circulating the steps, the prediction accuracy of the classifier is higher and higher until the preset accuracy requirement is met, and at the moment, the database system performance prediction model is trained. As will be appreciated by those skilled in the art, the template is trained onlineIn practice, the model training process of the invention can be continued all the time, and the prediction accuracy of the classifier can be continuously improved.
And 6, based on the trained database system performance prediction model, the system can predict the performance of the database system based on the database operation and maintenance data collected in real time.
Optionally, in step 7, if the prediction result triggers an early warning condition, the system may issue an early warning. For example, an early warning threshold may be preset, and according to different early warning conditions, if a predicted system evaluation value output by the database system performance prediction model exceeds, reaches, or is lower than the early warning threshold, early warning may be triggered.
Alternatively, in step 8, if the real-time system evaluation value result obtained in step 4 triggers an alarm condition, the system may directly perform an alarm. For example, an alarm threshold may be preset, and depending on the alarm condition, if the system evaluation value calculated in real time exceeds, reaches, or falls below the alarm threshold, an alarm may be triggered. As an example, the alarm triggering condition and the pre-warning triggering condition may be set to be the same or different according to needs, for example, the pre-warning triggering condition may be set to be tighter with respect to the alarm triggering condition, so that a certain margin is left between the pre-warning and the warning for the system to have time to avoid the alarm condition being triggered by adjusting the operation and maintenance operation after receiving the pre-warning.
Returning to FIG. 1, the data collection module 102 may be configured to collect database operation and maintenance data to form a multi-dimensional state vector for each time period, and to calculate a trend vector across multiple time periods based on the multi-dimensional state vector.
The database system performance evaluation module 104 may be configured to obtain database system performance indicators for respective time periods, and to calculate a system evaluation value based on the obtained database system performance indicators.
The machine learning module 106 may be configured to train a database system performance prediction model based on the trend vectors and the system evaluation values.
The database system performance prediction module 108 may be configured to provide a predicted system rating value using a trained database system performance prediction model.
The alert module 110 may be configured to receive and monitor real-time system rating values from the database system performance rating module and predictive system rating values from the database system performance prediction module output. Triggering preset early warning conditions in response to the predicted system evaluation value output by the database system performance prediction model, and sending out early warning; and triggering a preset alarm condition in response to the calculated system evaluation value, and sending an alarm.
What has been described above includes examples of aspects of the claimed subject matter. It is, of course, not possible to describe every conceivable combination of components or methodologies for purposes of describing the claimed subject matter, but one of ordinary skill in the art may recognize that many further combinations and permutations of the claimed subject matter are possible. Accordingly, the disclosed subject matter is intended to embrace all such alterations, modifications and variations that fall within the spirit and scope of the appended claims.
Claims (10)
1. A method for database system performance prediction based on machine learning, the method comprising:
collecting database operation and maintenance data to form a multi-dimensional state vector of each time period;
calculating a trend vector across a plurality of time periods based on the multi-dimensional state vector;
acquiring the performance indexes of the database system corresponding to each time period;
calculating a system evaluation value based on the acquired database system performance index;
training a database system performance prediction model based on the trend vector and the system evaluation value; and
and inputting the trend vector formed in real time into the database system performance prediction model to output a predicted system evaluation value.
2. The method of claim 1, wherein calculating a trend vector across a plurality of time periods based on the multi-dimensional state vector further comprises:
and splicing the state change vector of each pair of adjacent time periods of the plurality of involved time periods and the multidimensional state vector of the last time period of the plurality of involved time periods to form involved trend vectors, wherein the state change vectors of the adjacent time periods are the difference between the multidimensional state vector of the next time period and the multidimensional state vector of the previous time period.
3. The method of claim 1, wherein calculating a system merit value based on the obtained database system performance indicators further comprises:
comparing the obtained database system performance index with a corresponding reference value; and
the system evaluation value is determined as one of a plurality of equal-place values set in advance based on the comparison result.
4. The method of claim 1, wherein the database system performance prediction model is based on a SOFTMAX classifier, and training a database system performance prediction model based on the trend vector and the system merit value further comprises:
and respectively taking a multidimensional trend vector and a system evaluation value corresponding to the same time period as the input and the output of the SOFTMAX classifier.
5. The method of claim 4, wherein the training process is performed online in the following manner:
inputting the trend vector into the SOFTMAX classifier;
the SOFTMAX classifier outputs a prediction result based on the current weight;
performing backward propagation on the real system evaluation value corresponding to the period of the input trend vector, and deriving a loss function;
minimizing the loss function at the fastest rate by a gradient descent method;
updating the weights of the SOFTMAX classifier with derivatives; and
and repeating the above steps.
6. The method of claim 1, wherein the method further comprises:
and triggering preset early warning conditions in response to the predicted system evaluation value output by the database system performance prediction model, and sending out early warning.
7. The method of claim 1, wherein the method further comprises:
and triggering a preset alarm condition in response to the calculated system evaluation value, and sending an alarm.
8. A system for database system performance prediction based on machine learning, the system comprising:
a data acquisition module configured to:
collecting database operation and maintenance data to form a multi-dimensional state vector of each time period; and
calculating a trend vector across a plurality of time periods based on the multi-dimensional state vector;
a database system performance evaluation module configured to:
acquiring the performance indexes of the database system corresponding to each time period; and
calculating a system evaluation value based on the acquired database system performance index;
a machine learning module configured to train a database system performance prediction model based on the trend vector and the system evaluation value;
a database system performance prediction module configured to provide a predicted system evaluation value using a trained database system performance prediction model; and
an alert module configured to:
triggering preset early warning conditions in response to the predicted system evaluation value output by the database system performance prediction model, and sending out early warning; and
and triggering a preset alarm condition in response to the calculated system evaluation value, and sending an alarm.
9. The system of claim 8, wherein the database system is a MySQL database system.
10. The system of claim 8, wherein the database system performance indicators comprise database IO performance indicators.
Priority Applications (1)
Application Number | Priority Date | Filing Date | Title |
---|---|---|---|
CN201911296302.7A CN112988529B (en) | 2019-12-16 | 2019-12-16 | Method and system for predicting database system performance based on machine learning |
Applications Claiming Priority (1)
Application Number | Priority Date | Filing Date | Title |
---|---|---|---|
CN201911296302.7A CN112988529B (en) | 2019-12-16 | 2019-12-16 | Method and system for predicting database system performance based on machine learning |
Publications (2)
Publication Number | Publication Date |
---|---|
CN112988529A true CN112988529A (en) | 2021-06-18 |
CN112988529B CN112988529B (en) | 2024-08-16 |
Family
ID=76343527
Family Applications (1)
Application Number | Title | Priority Date | Filing Date |
---|---|---|---|
CN201911296302.7A Active CN112988529B (en) | 2019-12-16 | 2019-12-16 | Method and system for predicting database system performance based on machine learning |
Country Status (1)
Country | Link |
---|---|
CN (1) | CN112988529B (en) |
Cited By (1)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
US20230229542A1 (en) * | 2022-01-14 | 2023-07-20 | Servicenow, Inc. | Failure Prediction in a Computing System Based on Machine Learning Applied to Alert Data |
Citations (5)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
JPS62189539A (en) * | 1986-02-17 | 1987-08-19 | Hitachi Ltd | System for evaluating performance of related data base managing system |
US20060074970A1 (en) * | 2004-09-22 | 2006-04-06 | Microsoft Corporation | Predicting database system performance |
CN102521604A (en) * | 2011-11-21 | 2012-06-27 | 上海交通大学 | Device and method for estimating performance degradation of equipment based on inspection system |
CN108388503A (en) * | 2018-02-13 | 2018-08-10 | 中体彩科技发展有限公司 | Data-base performance monitoring method, system, equipment and computer readable storage medium |
CN109271374A (en) * | 2018-10-19 | 2019-01-25 | 国网江苏省电力有限公司信息通信分公司 | A kind of database health scoring method and scoring system based on machine learning |
-
2019
- 2019-12-16 CN CN201911296302.7A patent/CN112988529B/en active Active
Patent Citations (5)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
JPS62189539A (en) * | 1986-02-17 | 1987-08-19 | Hitachi Ltd | System for evaluating performance of related data base managing system |
US20060074970A1 (en) * | 2004-09-22 | 2006-04-06 | Microsoft Corporation | Predicting database system performance |
CN102521604A (en) * | 2011-11-21 | 2012-06-27 | 上海交通大学 | Device and method for estimating performance degradation of equipment based on inspection system |
CN108388503A (en) * | 2018-02-13 | 2018-08-10 | 中体彩科技发展有限公司 | Data-base performance monitoring method, system, equipment and computer readable storage medium |
CN109271374A (en) * | 2018-10-19 | 2019-01-25 | 国网江苏省电力有限公司信息通信分公司 | A kind of database health scoring method and scoring system based on machine learning |
Cited By (2)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
US20230229542A1 (en) * | 2022-01-14 | 2023-07-20 | Servicenow, Inc. | Failure Prediction in a Computing System Based on Machine Learning Applied to Alert Data |
US11829233B2 (en) * | 2022-01-14 | 2023-11-28 | Servicenow, Inc. | Failure prediction in a computing system based on machine learning applied to alert data |
Also Published As
Publication number | Publication date |
---|---|
CN112988529B (en) | 2024-08-16 |
Similar Documents
Publication | Publication Date | Title |
---|---|---|
CN111222290B (en) | Multi-parameter feature fusion-based method for predicting residual service life of large-scale equipment | |
CN108320043B (en) | Power distribution network equipment state diagnosis and prediction method based on electric power big data | |
CN111275288A (en) | XGboost-based multi-dimensional data anomaly detection method and device | |
CN110335168B (en) | Method and system for optimizing power utilization information acquisition terminal fault prediction model based on GRU | |
CN109583520B (en) | State evaluation method of cloud model and genetic algorithm optimization support vector machine | |
CN105205326B (en) | A kind of power plant's Real-time Load on-line prediction method | |
CN113032238A (en) | Real-time root cause analysis method based on application knowledge graph | |
CN116028887B (en) | Analysis method of continuous industrial production data | |
CN115081585A (en) | Man-machine-object cooperation abnormal state detection method for reinforcing heterogeneous graph neural network | |
CN115358281A (en) | Machine learning-based cold and hot all-in-one machine monitoring control method and system | |
CN112529053A (en) | Short-term prediction method and system for time sequence data in server | |
CN115688016A (en) | Integrated method for intelligent operation and maintenance of large database | |
CN115576267A (en) | Hub machining dimension error correction method based on digital twinning | |
CN108305002A (en) | A kind of decreasing failure rate height recognition methods based on monitoring Weibull form parameter | |
CN110781206A (en) | Method for predicting whether electric energy meter in operation fails or not by learning meter-dismantling and returning failure characteristic rule | |
CN112988529A (en) | Method and system for predicting database system performance based on machine learning | |
CN113918433A (en) | Adaptive intelligent network equipment performance index abnormity detection device and method | |
CN116258467B (en) | Electric power construction management and control system | |
CN113505818A (en) | Aluminum melting furnace energy consumption abnormity diagnosis method, system and equipment with improved decision tree algorithm | |
CN109359742A (en) | A kind of generation method in subway subsystem preventive maintenance period | |
CN116664098A (en) | Abnormality detection method and system for photovoltaic power station | |
CN116579789A (en) | Power battery performance analysis-based secondary vehicle estimation method and system | |
CN111026624A (en) | Fault prediction method and device of power grid information system | |
CN115828744A (en) | White light LED fault on-line diagnosis and service life prediction method | |
CN115935814A (en) | Transformer fault prediction method based on ARIMA-SVM model |
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 | ||
TA01 | Transfer of patent application right | ||
TA01 | Transfer of patent application right |
Effective date of registration: 20220128 Address after: Room 1423, No. 1256 and 1258, Wanrong Road, Jing'an District, Shanghai 200072 Applicant after: Tianyi Digital Life Technology Co.,Ltd. Address before: 201702 3rd floor, 158 Shuanglian Road, Qingpu District, Shanghai Applicant before: Tianyi Smart Family Technology Co.,Ltd. |
|
GR01 | Patent grant | ||
GR01 | Patent grant |