CN116827759A - Method and device for processing restarting instruction of converging current divider - Google Patents
Method and device for processing restarting instruction of converging current divider Download PDFInfo
- Publication number
- CN116827759A CN116827759A CN202311090831.8A CN202311090831A CN116827759A CN 116827759 A CN116827759 A CN 116827759A CN 202311090831 A CN202311090831 A CN 202311090831A CN 116827759 A CN116827759 A CN 116827759A
- Authority
- CN
- China
- Prior art keywords
- time sequence
- splitter
- server
- server load
- load
- 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
- 238000012545 processing Methods 0.000 title claims abstract description 54
- 238000000034 method Methods 0.000 title claims abstract description 41
- 230000002159 abnormal effect Effects 0.000 claims abstract description 36
- 238000001514 detection method Methods 0.000 claims abstract description 19
- 230000004044 response Effects 0.000 claims abstract description 17
- 238000012790 confirmation Methods 0.000 claims abstract description 7
- 230000005856 abnormality Effects 0.000 claims abstract description 6
- 239000013598 vector Substances 0.000 claims description 94
- 238000012549 training Methods 0.000 claims description 43
- 239000011159 matrix material Substances 0.000 claims description 28
- 238000013527 convolutional neural network Methods 0.000 claims description 16
- 238000003672 processing method Methods 0.000 claims description 8
- 238000012300 Sequence Analysis Methods 0.000 claims description 7
- 230000007704 transition Effects 0.000 claims description 4
- 238000004458 analytical method Methods 0.000 claims description 3
- 230000006870 function Effects 0.000 description 11
- 238000010586 diagram Methods 0.000 description 10
- 238000009826 distribution Methods 0.000 description 8
- 230000002776 aggregation Effects 0.000 description 7
- 238000004220 aggregation Methods 0.000 description 7
- 230000008569 process Effects 0.000 description 7
- 238000012544 monitoring process Methods 0.000 description 4
- 230000000694 effects Effects 0.000 description 3
- 238000013528 artificial neural network Methods 0.000 description 2
- 230000008859 change Effects 0.000 description 2
- 239000000284 extract Substances 0.000 description 2
- 238000000605 extraction Methods 0.000 description 2
- 230000036541 health Effects 0.000 description 2
- 238000007477 logistic regression Methods 0.000 description 2
- 238000013507 mapping Methods 0.000 description 2
- 230000007246 mechanism Effects 0.000 description 2
- 238000003062 neural network model Methods 0.000 description 2
- 238000011176 pooling Methods 0.000 description 2
- 206010019233 Headaches Diseases 0.000 description 1
- 230000002411 adverse Effects 0.000 description 1
- 239000003795 chemical substances by application Substances 0.000 description 1
- 238000004891 communication Methods 0.000 description 1
- 238000004590 computer program Methods 0.000 description 1
- 238000005094 computer simulation Methods 0.000 description 1
- 238000007405 data analysis Methods 0.000 description 1
- 238000005516 engineering process Methods 0.000 description 1
- 230000002708 enhancing effect Effects 0.000 description 1
- 231100000869 headache Toxicity 0.000 description 1
- 230000006872 improvement Effects 0.000 description 1
- 230000003993 interaction Effects 0.000 description 1
- 238000012804 iterative process Methods 0.000 description 1
- 238000012423 maintenance Methods 0.000 description 1
- 238000012986 modification Methods 0.000 description 1
- 230000004048 modification Effects 0.000 description 1
- 238000003058 natural language processing Methods 0.000 description 1
- 238000005457 optimization Methods 0.000 description 1
- 238000002360 preparation method Methods 0.000 description 1
- 230000000306 recurrent effect Effects 0.000 description 1
- 230000003252 repetitive effect Effects 0.000 description 1
- 230000002441 reversible effect Effects 0.000 description 1
- 238000013519 translation Methods 0.000 description 1
Classifications
-
- H—ELECTRICITY
- H04—ELECTRIC COMMUNICATION TECHNIQUE
- H04L—TRANSMISSION OF DIGITAL INFORMATION, e.g. TELEGRAPHIC COMMUNICATION
- H04L43/00—Arrangements for monitoring or testing data switching networks
- H04L43/08—Monitoring or testing based on specific metrics, e.g. QoS, energy consumption or environmental parameters
- H04L43/0805—Monitoring or testing based on specific metrics, e.g. QoS, energy consumption or environmental parameters by checking availability
- H04L43/0817—Monitoring or testing based on specific metrics, e.g. QoS, energy consumption or environmental parameters by checking availability by checking functioning
-
- H—ELECTRICITY
- H04—ELECTRIC COMMUNICATION TECHNIQUE
- H04L—TRANSMISSION OF DIGITAL INFORMATION, e.g. TELEGRAPHIC COMMUNICATION
- H04L67/00—Network arrangements or protocols for supporting network services or applications
- H04L67/01—Protocols
- H04L67/10—Protocols in which an application is distributed across nodes in the network
- H04L67/1001—Protocols in which an application is distributed across nodes in the network for accessing one among a plurality of replicated servers
- H04L67/1004—Server selection for load balancing
- H04L67/1008—Server selection for load balancing based on parameters of servers, e.g. available memory or workload
-
- H—ELECTRICITY
- H04—ELECTRIC COMMUNICATION TECHNIQUE
- H04L—TRANSMISSION OF DIGITAL INFORMATION, e.g. TELEGRAPHIC COMMUNICATION
- H04L67/00—Network arrangements or protocols for supporting network services or applications
- H04L67/01—Protocols
- H04L67/10—Protocols in which an application is distributed across nodes in the network
- H04L67/1001—Protocols in which an application is distributed across nodes in the network for accessing one among a plurality of replicated servers
- H04L67/1029—Protocols in which an application is distributed across nodes in the network for accessing one among a plurality of replicated servers using data related to the state of servers by a load balancer
-
- H—ELECTRICITY
- H04—ELECTRIC COMMUNICATION TECHNIQUE
- H04L—TRANSMISSION OF DIGITAL INFORMATION, e.g. TELEGRAPHIC COMMUNICATION
- H04L67/00—Network arrangements or protocols for supporting network services or applications
- H04L67/01—Protocols
- H04L67/10—Protocols in which an application is distributed across nodes in the network
- H04L67/1001—Protocols in which an application is distributed across nodes in the network for accessing one among a plurality of replicated servers
- H04L67/1034—Reaction to server failures by a load balancer
Landscapes
- Engineering & Computer Science (AREA)
- Computer Networks & Wireless Communication (AREA)
- Signal Processing (AREA)
- Environmental & Geological Engineering (AREA)
- Computer Hardware Design (AREA)
- General Engineering & Computer Science (AREA)
- Debugging And Monitoring (AREA)
Abstract
The application discloses a method and a device for processing a restarting instruction of a converging shunt. Firstly, in response to receiving a restarting instruction of a converging flow divider, carrying out abnormality detection on the working state of the converging flow divider, then, in response to the working state of the converging flow divider being abnormal, executing the restarting instruction of the converging flow divider, and then, in response to the working state of the converging flow divider being normal, generating a confirmation prompt. In this way, intelligent anomaly detection of the operating state of the sink splitter can be performed based on load values of a plurality of servers connected to the sink splitter at a plurality of predetermined time points within a predetermined period of time.
Description
Technical Field
The present disclosure relates to the field of convergence splitters, and more particularly, to a method and apparatus for processing a restart instruction of a convergence splitter.
Background
A convergence splitter is a network device commonly used for network load balancing and redundancy backup. If the aggregate splitter fails or performs abnormally, such as memory leaks, resource conflicts, software errors, or network problems, a reboot may attempt to fix the problem. In addition, when a change is made to the configuration of the converging splitter, a reboot is sometimes required to effect the change. That is, the restart instructions of the converging splitter may address some issues or perform system maintenance to ensure its proper operation and performance optimization.
However, a restart may result in a brief network disruption, which may result in a loss of data or configuration if the configuration of the device is backed up prior to the restart operation. That is, if the user operates by mistake to generate a restart instruction, this may cause a more headache problem.
Thus, an optimized restart instruction processing scheme for a converging splitter is desired.
Disclosure of Invention
In view of this, the disclosure provides a method and an apparatus for processing a restart instruction of a converging splitter, which can perform intelligent anomaly detection on an operating state of the converging splitter based on load values of a plurality of servers connected to the converging splitter at a plurality of predetermined time points in a predetermined period.
According to an aspect of the present disclosure, there is provided a restart instruction processing method of a sink splitter, including:
in response to receiving a restarting instruction of the converging flow divider, detecting the working state of the converging flow divider abnormally;
responding to the abnormal working state of the converging current divider, and executing a restarting instruction of the converging current divider; and
and generating a confirmation prompt in response to the working state of the converging current divider being normal.
According to another aspect of the present disclosure, there is provided a restart instruction processing apparatus of a sink splitter, which operates in the foregoing method.
According to the embodiment of the disclosure, firstly, in response to receiving a restart instruction of a converging flow divider, abnormality detection is performed on the working state of the converging flow divider, then, in response to the working state of the converging flow divider being abnormal, the restart instruction of the converging flow divider is executed, and then, in response to the working state of the converging flow divider being normal, a confirmation prompt is generated. In this way, intelligent anomaly detection of the operating state of the sink splitter can be performed based on load values of a plurality of servers connected to the sink splitter at a plurality of predetermined time points within a predetermined period of time.
Other features and aspects of the present disclosure will become apparent from the following detailed description of exemplary embodiments, which proceeds with reference to the accompanying drawings.
Drawings
The accompanying drawings, which are incorporated in and constitute a part of this specification, illustrate exemplary embodiments, features and aspects of the present disclosure and together with the description, serve to explain the principles of the disclosure.
Fig. 1 shows a flowchart of a restart instruction processing method of a sink splitter according to an embodiment of the present disclosure.
Fig. 2 shows a flowchart of substep S110 of a restart instruction processing method of a sink splitter according to an embodiment of the present disclosure.
Fig. 3 shows an architectural diagram of substep S110 of a method of processing a restart instruction of a sink splitter according to an embodiment of the disclosure.
Fig. 4 shows a flowchart of training steps further included in a method of processing a restart instruction of a sink splitter according to an embodiment of the present disclosure.
Fig. 5 illustrates a block diagram of a restart instruction processing system of a sink splitter according to an embodiment of the disclosure.
Fig. 6 illustrates an application scenario diagram of a restart instruction processing method of a sink splitter according to an embodiment of the disclosure.
Detailed Description
The following description of the embodiments of the present disclosure will be made clearly and fully with reference to the accompanying drawings, in which it is apparent that the embodiments described are only some, but not all embodiments of the disclosure. All other embodiments, which can be made by one of ordinary skill in the art without undue burden based on the embodiments of the present disclosure, are also within the scope of the present disclosure.
As used in this disclosure and in the claims, the terms "a," "an," "the," and/or "the" are not specific to a singular, but may include a plurality, unless the context clearly dictates otherwise. In general, the terms "comprises" and "comprising" merely indicate that the steps and elements are explicitly identified, and they do not constitute an exclusive list, as other steps or elements may be included in a method or apparatus.
Various exemplary embodiments, features and aspects of the disclosure will be described in detail below with reference to the drawings. In the drawings, like reference numbers indicate identical or functionally similar elements. Although various aspects of the embodiments are illustrated in the accompanying drawings, the drawings are not necessarily drawn to scale unless specifically indicated.
In addition, numerous specific details are set forth in the following detailed description in order to provide a better understanding of the present disclosure. It will be understood by those skilled in the art that the present disclosure may be practiced without some of these specific details. In some instances, methods, means, elements, and circuits well known to those skilled in the art have not been described in detail in order not to obscure the present disclosure.
A convergence splitter is a network device for implementing load balancing and redundancy backup that is capable of distributing network traffic across multiple servers to improve system availability and performance. Convergence splitters are commonly used in large network environments, such as data centers or enterprise internal networks. The aggregation splitter can distribute network requests to different servers according to the load conditions of the servers so as to realize load balancing, thus avoiding overload of a certain server and improving the performance and availability of the whole system. The convergence diverter can be configured with a plurality of servers as backup, and when the main server fails, the flow is automatically switched to the backup server, so that the continuity and the reliability of the system are ensured. The convergence splitter can perform regular health checks on the servers to ensure proper operation of the servers, and if a certain server fails or is unavailable, the convergence splitter can remove traffic from the server to avoid sending requests to unavailable servers. The convergence diverter provides a management interface, can be configured and managed, and an administrator can adjust a load balancing algorithm, add or delete servers, configure health checks and the like according to requirements.
In view of the foregoing technical problems, the technical idea of the present disclosure is to, after receiving a restart instruction of a converging splitter, take into account that the restart instruction may be generated due to a user's misoperation, monitor an operating state of the converging splitter before executing the restart instruction of the converging splitter to determine whether the operating state is normal, if so, execute the restart instruction of the converging splitter, and if so, generate a prompt for determining whether to execute the restart instruction of the converging splitter.
Specifically, the present disclosure provides a method for processing a restart instruction of a sink splitter, and fig. 1 shows a flowchart of a method for processing a restart instruction of a sink splitter according to an embodiment of the present disclosure. As shown in fig. 1, a method for processing a restart instruction of a sink splitter according to an embodiment of the disclosure includes the steps of: s110, in response to receiving a restarting instruction of the converging flow divider, performing anomaly detection on the working state of the converging flow divider; s120, responding to the abnormal working state of the converging current divider, and executing a restarting instruction of the converging current divider; and S130, generating a confirmation prompt in response to the working state of the converging current divider being normal.
In particular, the S1 step is an important step in the technical solution of the present disclosure. It should be understood that the accuracy of the anomaly detection of the working state of the converging splitter in step S110 is critical to the reliability of the whole scheme. If the anomaly detection is inaccurate, it may result in an erroneous judgment of the device state, thereby performing an unnecessary restart operation or ignoring a case where a restart is actually required.
In order to accurately detect the abnormality of the operating state of the converging splitter, the present disclosure contemplates intelligent abnormality detection of the operating state of the converging splitter based on load values of a plurality of servers connected to the converging splitter at a plurality of predetermined time points within a predetermined period of time.
Fig. 2 shows a flowchart of substep S110 of a restart instruction processing method of a sink splitter according to an embodiment of the present disclosure. Fig. 3 shows an architectural diagram of substep S110 of a method of processing a restart instruction of a sink splitter according to an embodiment of the disclosure. As shown in fig. 2 and 3, according to a method for processing a restart instruction of a sink splitter in an embodiment of the disclosure, in response to receiving a restart instruction of the sink splitter, performing anomaly detection on a working state of the sink splitter, including: s111, load values of a plurality of servers connected with the converging splitter at a plurality of preset time points in a preset time period are obtained; s112, carrying out time sequence analysis on load values of a plurality of predetermined time points of the servers in a predetermined time period to obtain a server load time sequence association feature vector; and S113, determining whether the working state of the converging current divider is abnormal or not based on the server load time sequence association characteristic vector.
Based on this, in the technical solution of the present disclosure, first, load values of a plurality of servers connected to the sink splitter at a plurality of predetermined time points within a predetermined period of time are acquired. It should be appreciated that the load value may reflect the operational state of the server. Normally, the load value of the server should fluctuate within a certain range, and when the load value exceeds the normal range, it may indicate that an abnormal situation occurs in the server. By collecting the load values of a plurality of servers, more comprehensive information can be obtained to judge whether the working state of the converging current divider is abnormal. That is, if the load values of a plurality of servers are abnormal at the same time, it is likely that the sink splitter itself has a problem. In contrast, if only the load values of the individual servers are abnormal, which may be caused by a failure of a single server, there is no need to restart the aggregate splitter.
It should be understood that the following method may be used to obtain load values of a plurality of servers connected to the sink splitter at a plurality of predetermined time points within a predetermined period of time: SNMP (Simple Network Management Protocol) is a protocol for managing and monitoring network devices, and by using SNMP, various information including a load value can be obtained from a server, and load information of the server can be obtained by querying an SNMP agent on the server; 2. the load value of the server can be obtained by calling the corresponding API, and in general, a management interface of the server can provide some methods for obtaining load information; 3. the monitoring tool can be used for monitoring the load condition of the server in real time and recording the load condition. It should be noted that, obtaining the load value of the server requires the server to support the corresponding monitoring function and to have the corresponding authority to obtain the information, and in addition, it is required to establish a correct connection between the aggregation splitter and the server and ensure that the network communication is normal.
And then, carrying out time sequence analysis on load values of a plurality of predetermined time points of the servers in a predetermined time period to obtain a server load time sequence association characteristic vector. Here, considering that there is a certain correlation between loads of multiple servers, for example, in the case of load balancing, multiple servers may commonly bear the load of a user request, and when the load of a certain server is too high, the system may forward a part of the request to other servers with lower loads, so as to achieve load balancing, and thus, the loads of the multiple servers may affect each other to a certain extent. In the disclosed solution, it is desirable to capture such a correlation pattern distribution to facilitate subsequent anomaly detection.
Correspondingly, performing time sequence analysis on load values of a plurality of predetermined time points of the plurality of servers in a predetermined time period to obtain a server load time sequence association feature vector, including: respectively carrying out structuring processing on load values of a plurality of preset time points of each server in a preset time period to obtain a plurality of server load time sequence input vectors; the load time sequence input vectors of the servers are respectively passed through a load time sequence feature extractor based on a one-dimensional convolutional neural network model to obtain load time sequence feature vectors of the servers; and extracting the correlation pattern features among the plurality of server load time sequence feature vectors to obtain the server load time sequence correlation feature vectors. It should be appreciated that a one-dimensional convolutional neural network (1D CNN) is a neural network model that processes one-dimensional sequence data, such as time series data, to better capture local patterns and timing features in the data than a conventional fully-connected neural network. In the server load time sequence analysis, the one-dimensional convolutional neural network can be used as a load time sequence feature extractor, and the feature vector of each server load time sequence can be extracted by transmitting load time sequence input vectors of a plurality of servers in a preset time period to a one-dimensional convolutional neural network model. The one-dimensional convolutional neural network model learns local patterns and timing information in an input sequence by using a one-dimensional convolutional layer and a pooling layer. The convolution layer extracts features on the input sequence by sliding a window of fixed size, and then the pooling layer downsamples the extracted features, reducing the dimensionality of the features. This preserves important timing characteristics and reduces computational complexity. By using a one-dimensional convolutional neural network model as the load timing feature extractor, complex load timing data can be converted into a more representative and compact feature vector representation. These feature vectors can be used for later correlation pattern analysis to help identify the time-series correlation features of the server load and thus determine if the operational state of the converging splitter is abnormal.
The method for respectively structuring the load values of a plurality of preset time points of each server in a preset time period to obtain a plurality of server load time sequence input vectors comprises the following steps: and respectively arranging load values of a plurality of preset time points of each server in a preset time period into input vectors according to a time dimension to obtain a plurality of server load time sequence input vectors. It should be appreciated that the transducer model (Transformer Model) is a neural network model based on self-attention mechanisms, originally used for natural language processing tasks such as machine translation and language generation, which is excellent in sequential data processing and has been successfully applied to other fields including time-sequential data analysis. In server load time series correlation feature extraction, the converter model may be used as a server load correlation pattern feature extractor. By inputting a plurality of server load timing feature vectors into the converter model, correlation pattern features between these feature vectors can be learned. The transducer model captures the association information between different positions in the input sequence through a self-attention mechanism. It can consider the dependency between all input positions at the same time, unlike conventional recurrent neural networks which require sequential processing of sequences, which enables the converter model to better capture the correlation pattern between server load timing feature vectors. By using the converter model as a server load correlation pattern feature extractor, higher-level correlation features can be extracted to help identify timing correlations between server loads. These correlation features can be used to generate server load time sequence correlation feature vectors, and further used to determine whether the working state of the converging current divider is abnormal.
Extracting the correlation pattern features among the plurality of server load time sequence feature vectors to obtain the server load time sequence correlation feature vectors comprises the following steps: and passing the plurality of server load time sequence feature vectors through a server load correlation mode feature extractor based on a converter model to obtain the server load time sequence correlation feature vectors.
In a specific example of the disclosure, load values of the servers at a plurality of predetermined time points in a predetermined time period are firstly arranged as input vectors according to a time dimension respectively to obtain a plurality of server load time sequence input vectors; then, the plurality of server load time sequence input vectors respectively pass through a load time sequence feature extractor based on a one-dimensional convolutional neural network model to obtain a plurality of server load time sequence feature vectors; and then the server load time sequence feature vectors pass through a server load association mode feature extractor based on a converter model to obtain the server load time sequence association feature vectors.
And then, the server load time sequence association feature vector passes through a classifier to obtain a classification result, wherein the classification result is used for indicating whether the working state of the converging current divider is abnormal or not.
Correspondingly, based on the server load time sequence association feature vector, determining whether the working state of the converging current divider is abnormal comprises the following steps: and the server load time sequence association feature vector passes through a classifier to obtain a classification result, wherein the classification result is used for indicating whether the working state of the converging current divider is abnormal or not.
That is, in the technical solution of the present disclosure, the label of the classifier includes an abnormal working state (a first label) of the converging shunt, and an abnormal working state (a second label) of the converging shunt, where the classifier determines, through a soft maximum function, to which classification label the server load time sequence association feature vector belongs. It should be noted that the first tag p1 and the second tag p2 do not include a manually set concept, and in fact, during the training process, the computer model does not have a concept of "whether the working state of the converging splitter is abnormal", which is just that there are two kinds of classification tags and the probability that the output feature is under the two kinds of classification tags, that is, the sum of p1 and p2 is one. Therefore, the classification result of whether the working state of the converging current divider is abnormal is actually converted into a classification probability distribution conforming to the classification rule by classifying the labels, and the physical meaning of the natural probability distribution of the labels is essentially used instead of the language text meaning of whether the working state of the converging current divider is abnormal.
It should be appreciated that the role of the classifier is to learn the classification rules and classifier using a given class, known training data, and then classify (or predict) the unknown data. Logistic regression (logistics), SVM, etc. are commonly used to solve the classification problem, and for multi-classification problems (multi-class classification), logistic regression or SVM can be used as well, but multiple bi-classifications are required to compose multiple classifications, but this is error-prone and inefficient, and the commonly used multi-classification method is the Softmax classification function.
Accordingly, in one possible implementation manner, the server load time sequence association feature vector is passed through a classifier to obtain a classification result, where the classification result is used to indicate whether the working state of the aggregation shunt is abnormal, and the method includes: performing full-connection coding on the server load time sequence association feature vector by using a full-connection layer of the classifier to obtain a coding classification feature vector; and inputting the coding classification feature vector into a Softmax classification function of the classifier to obtain the classification result.
Further, in the technical solution of the present disclosure, the method for processing a restart instruction of a converging splitter further includes a training step: and training the load time sequence feature extractor based on the one-dimensional convolutional neural network model, the server load association mode feature extractor based on the converter model and the classifier. It should be understood that the training step plays an important role in the method for processing the restart instruction of the converging current divider, and by training the load time sequence feature extractor, the server load association mode feature extractor and the classifier, the models can learn the distinction between the normal working state and the abnormal working state, so that whether the working state of the converging current divider is abnormal can be accurately judged. In particular, the purpose of the training step is to train these models by providing sample data of known operating conditions. For the load timing feature extractor, the load timing data under known normal and abnormal operating conditions can be used for training so that it can extract feature vectors from the timing data. For the server load correlation pattern feature extractor, the load time sequence correlation feature vectors under known normal and abnormal working states can be used for training, so that the correlation pattern features among the feature vectors can be extracted. Finally, for the classifier, sample data with labels can be used for training, so that whether the working state of the converging current divider is abnormal can be accurately judged according to the extracted feature vector. By training the models, the accuracy and reliability of the detection of the working state of the converging current divider can be improved, so that corresponding measures, such as restarting the current divider, can be timely taken when abnormal conditions occur, and the normal operation of the system can be ensured.
As shown in fig. 4, the training step includes: s210, training data are acquired, wherein the training data comprise training load values of a plurality of servers connected with the converging current divider at a plurality of preset time points in a preset time period and true values of whether the working state of the converging current divider is abnormal or not; s220, training load values of a plurality of preset time points of each server in a preset time period are respectively arranged into input vectors according to a time dimension to obtain a plurality of training server load time sequence input vectors; s230, enabling the plurality of training server load time sequence input vectors to respectively pass through the load time sequence feature extractor based on the one-dimensional convolutional neural network model so as to obtain a plurality of training server load time sequence feature vectors; s240, enabling the plurality of training server load time sequence feature vectors to pass through the server load association mode feature extractor based on the converter model to obtain training server load time sequence association feature vectors; s250, the training server load time sequence association feature vector passes through a classifier to obtain a classification loss function value; and S260, training the load time sequence feature extractor based on the one-dimensional convolutional neural network model, the server load association mode feature extractor based on the converter model and the classifier based on the classification loss function value and through back propagation of gradient descent, wherein in each round of iteration of the training, the weight matrix of the classifier is subjected to external boundary constraint based on reference annotation.
Further, in the technical solution of the present disclosure, because there is inconsistency in time sequence distribution of a plurality of predetermined time points in a predetermined time period of load values of each server, after local time sequence correlation feature extraction is performed by a one-dimensional convolutional neural network model, there is a feature distribution difference caused by a difference in source time sequence distribution modes between the plurality of server load time sequence feature vectors, so that geometric monotonicity of a high-dimensional feature manifold represented by the feature distribution difference is poor, even if global context correlation is performed on feature distributions of each of the plurality of server load time sequence feature vectors by a converter model, the obtained server load time sequence correlation feature vectors also cause inter-domain offset of class probability mapping of the server load time sequence correlation feature vectors in a weight matrix iteration process of a classifier, and further the weight matrix is based on domain divergence of the server load time sequence correlation feature vectors, thereby affecting training effect of the model and accuracy of classification results of the server load time sequence correlation feature vectors obtained by the trained model. Based on the method, the external boundary constraint of the weight matrix based on the reference annotation is carried out in the training process of the server load time sequence association feature vector passing through the classifier.
In one possible implementation, performing an external boundary constraint based on a reference annotation on a weight matrix of the classifier includes:
performing external boundary constraint based on reference annotation on the weight matrix of the classifier by using the following external boundary constraint formula to obtain an optimized weight matrix;
wherein the external boundary constraint formula is:
wherein , and />The weight matrix of the last iteration and the current iteration are respectively adopted, wherein, during the first iteration, different initialization strategies are adopted to set +.> and />(e.g.)>Set as a unitary matrix->Set as the diagonal matrix of the mean value of the feature vector to be classified),>is the server load time sequence associated characteristic vector in the form of column vector>Is a first transition matrix, < >>Is a second transition matrix, ">Representing matrix multiplication +.>Representing matrix addition, ++>Representing the transpose of the vector>Is the optimized weight matrix.
Here, by associating feature vectors at the server load timingThe iterative association expression in the weight space is used as the external association boundary constraint of the weight matrix iteration, so that the time sequence association characteristic vector of the server load in the weight space iteration process is reduced under the condition that the previous weight matrix is used as the reference annotation (benchmark annotation) in the iteration process>Is used as an anchor point, thereby carrying out the directional mismatching (oriented mismatch) of the weight matrix relative to the server load time sequence association characteristic vector in the iterative process>Inter-domain offset compensation of class probability mapping of (c) and further enhancing the weight matrix based on saidServer load time sequence association feature vector>The domain fitting aggregation of the server load time sequence association feature vector is carried out to improve the training effect of the model and the accuracy of the classification result of the server load time sequence association feature vector obtained by the trained model.
In summary, according to the method for processing the restart instruction of the converging splitter according to the embodiment of the disclosure, intelligent anomaly detection on the working state of the converging splitter may be performed based on load values of a plurality of servers connected to the converging splitter at a plurality of predetermined time points within a predetermined time period.
Further, the technical scheme of the disclosure also provides a restart instruction processing device of the convergence shunt, and the restart instruction processing device of the convergence shunt operates by the method. It should be understood that the restart instruction processing device of the converging splitter is used for processing the restart instruction of the converging splitter. The device has the following characteristics: the automatic processing and restarting instruction processing device can automatically receive and process restarting instructions without manual intervention, so that the operation efficiency can be improved, and the manual errors can be reduced; the intelligent judgment, the restart instruction processing device can intelligently judge when to execute the restart operation, and can make a decision based on the working state of the converging current divider and the abnormal detection result, so that the restart operation is ensured to be carried out when necessary, and unnecessary restart is avoided; the safe and reliable restarting instruction processing device adopts safety measures to ensure the reliability and stability of restarting operation, and can carry out necessary checking and preparation work before the restarting operation is executed to ensure that the system is not adversely affected; the log record and the restart instruction processing device can record the execution condition and the result of the restart operation, including information such as restart time, operators and the like. Thus, the subsequent fault detection and analysis can be facilitated. The restarting instruction processing device of the converging current divider can provide automatic, intelligent, safe and reliable restarting instruction processing functions, and improves the reliability and stability of the system.
Fig. 5 illustrates a block diagram of a restart instruction processing system 100 of a sink splitter according to an embodiment of the disclosure. As shown in fig. 5, a restart instruction processing system 100 of a sink splitter according to an embodiment of the disclosure includes: an anomaly detection module 110, configured to perform anomaly detection on a working state of the converging splitter in response to receiving a restart instruction of the converging splitter; a restart instruction execution module 120, configured to execute a restart instruction of the converging splitter in response to an abnormal working state of the converging splitter; and a confirmation prompt generation module 130, configured to generate a confirmation prompt in response to the working state of the converging splitter being normal.
In one possible implementation, the anomaly detection module 110 includes: a load value obtaining unit, configured to obtain load values of a plurality of servers connected to the aggregation splitter at a plurality of predetermined time points within a predetermined time period; the time sequence analysis unit is used for performing time sequence analysis on load values of a plurality of servers at a plurality of preset time points in a preset time period to obtain a server load time sequence association characteristic vector; and the working state abnormality judging unit is used for determining whether the working state of the converging current divider is abnormal or not based on the server load time sequence association characteristic vector.
Here, it will be understood by those skilled in the art that the specific functions and operations of the respective units and modules in the above-described sink-splitter restart instruction processing system 100 have been described in detail in the above description of the sink-splitter restart instruction processing method with reference to fig. 1 to 4, and thus, repetitive descriptions thereof will be omitted.
As described above, the restart instruction processing system 100 of the sink splitter according to the embodiment of the present disclosure may be implemented in various wireless terminals, such as a server or the like having a restart instruction processing algorithm of the sink splitter. In one possible implementation, the restart instruction processing system 100 of the convergence splitter according to embodiments of the present disclosure may be integrated into the wireless terminal as one software module and/or hardware module. For example, the restart instruction processing system 100 of the sink splitter may be a software module in the operating system of the wireless terminal, or may be an application developed for the wireless terminal; of course, the restart instruction processing system 100 of the convergence splitter may also be one of a plurality of hardware modules of the wireless terminal.
Alternatively, in another example, the aggregation splitter restart instruction processing system 100 and the wireless terminal may be separate devices, and the aggregation splitter restart instruction processing system 100 may be connected to the wireless terminal through a wired and/or wireless network and transmit the interaction information according to an agreed data format.
Fig. 6 illustrates an application scenario diagram of a restart instruction processing method of a sink splitter according to an embodiment of the disclosure. As shown in fig. 6, in this application scenario, first, load values (for example, D illustrated in fig. 6) at a plurality of predetermined time points in a predetermined period of time of a plurality of servers connected to the sink splitter are acquired, and then the load values at the plurality of predetermined time points in the predetermined period of time of the plurality of servers are input to a server where a restart instruction processing algorithm of the sink splitter is deployed (for example, S illustrated in fig. 6), wherein the server is capable of processing the load values at the plurality of predetermined time points in the predetermined period of time of the plurality of servers using the restart instruction processing algorithm of the sink splitter to obtain a classification result for indicating whether the operating state of the sink splitter is abnormal.
The flowcharts and block diagrams in the figures illustrate the architecture, functionality, and operation of possible implementations of systems, methods and computer program products according to various embodiments of the present disclosure. In this regard, each block in the flowchart or block diagrams may represent a module, segment, or portion of instructions, which comprises one or more executable instructions for implementing the specified logical function(s). In some alternative implementations, the functions noted in the block may occur out of the order noted in the figures. For example, two blocks shown in succession may, in fact, be executed substantially concurrently, or the blocks may sometimes be executed in the reverse order, depending upon the functionality involved. It will also be noted that each block of the block diagrams and/or flowchart illustration, and combinations of blocks in the block diagrams and/or flowchart illustration, can be implemented by special purpose hardware-based systems which perform the specified functions or acts, or combinations of special purpose hardware and computer instructions.
The foregoing description of the embodiments of the present disclosure has been presented for purposes of illustration and description, and is not intended to be exhaustive or limited to the embodiments disclosed. Many modifications and variations will be apparent to those of ordinary skill in the art without departing from the scope and spirit of the various embodiments described. The terminology used herein was chosen in order to best explain the principles of the embodiments, the practical application, or the improvement of technology in the marketplace, or to enable others of ordinary skill in the art to understand the embodiments disclosed herein.
Claims (9)
1. The restart instruction processing method of the converging splitter is characterized by comprising the following steps of:
in response to receiving a restarting instruction of the converging flow divider, detecting the working state of the converging flow divider abnormally;
responding to the abnormal working state of the converging current divider, and executing a restarting instruction of the converging current divider; and
and generating a confirmation prompt in response to the working state of the converging current divider being normal.
2. The method for processing the restart instruction of the sink splitter according to claim 1, wherein the abnormality detection of the working state of the sink splitter in response to receiving the restart instruction of the sink splitter comprises:
load values of a plurality of servers connected with the converging splitter at a plurality of preset time points in a preset time period are obtained;
carrying out time sequence analysis on load values of a plurality of preset time points of the servers in a preset time period to obtain a server load time sequence association feature vector; and
and determining whether the working state of the converging current divider is abnormal or not based on the server load time sequence association characteristic vector.
3. The method for processing the restart instruction of the sink splitter according to claim 2, wherein performing timing analysis on load values of the plurality of servers at a plurality of predetermined time points within a predetermined period of time to obtain a server load timing association feature vector comprises:
respectively carrying out structuring processing on load values of a plurality of preset time points of each server in a preset time period to obtain a plurality of server load time sequence input vectors;
the load time sequence input vectors of the servers are respectively passed through a load time sequence feature extractor based on a one-dimensional convolutional neural network model to obtain load time sequence feature vectors of the servers; and
and extracting the correlation pattern characteristics among the plurality of server load time sequence characteristic vectors to obtain the server load time sequence correlation characteristic vectors.
4. The method for processing the restart instruction of the sink splitter according to claim 3, wherein the structuring the load values of the respective servers at a plurality of predetermined time points within a predetermined time period to obtain a plurality of server load time sequence input vectors, respectively, includes:
and respectively arranging load values of a plurality of preset time points of each server in a preset time period into input vectors according to a time dimension to obtain a plurality of server load time sequence input vectors.
5. The method according to claim 4, wherein extracting correlation pattern features among the plurality of server load timing feature vectors to obtain server load timing correlation feature vectors comprises:
and passing the plurality of server load time sequence feature vectors through a server load correlation mode feature extractor based on a converter model to obtain the server load time sequence correlation feature vectors.
6. The method for processing the restart instruction of the sink splitter according to claim 5, wherein determining whether the operation state of the sink splitter is abnormal based on the server load time sequence association feature vector comprises:
and the server load time sequence association feature vector passes through a classifier to obtain a classification result, wherein the classification result is used for indicating whether the working state of the converging current divider is abnormal or not.
7. The method for processing the restart instruction of the sink splitter according to claim 6, further comprising the training step of: training the load time sequence feature extractor based on the one-dimensional convolutional neural network model, the server load association mode feature extractor based on the converter model and the classifier;
wherein the training step comprises:
acquiring training data, wherein the training data comprises training load values of a plurality of servers connected with the converging current divider at a plurality of preset time points in a preset time period and a true value of whether the working state of the converging current divider is abnormal or not;
training load values of a plurality of preset time points of each server in a preset time period are respectively arranged into input vectors according to a time dimension to obtain a plurality of training server load time sequence input vectors;
the load time sequence input vectors of the training servers are respectively passed through the load time sequence feature extractor based on the one-dimensional convolutional neural network model to obtain load time sequence feature vectors of the training servers;
passing the plurality of training server load time sequence feature vectors through the converter model-based server load correlation pattern feature extractor to obtain training server load time sequence correlation feature vectors;
the training server load time sequence association feature vector passes through a classifier to obtain a classification loss function value; and
training the one-dimensional convolutional neural network model-based load time sequence feature extractor, the converter model-based server load correlation mode feature extractor and the classifier based on the classification loss function values and through back propagation of gradient descent, wherein, in each iteration of the training, an external boundary constraint based on reference annotation is performed on a weight matrix of the classifier.
8. The method for processing the restart instruction of the sink splitter according to claim 7, wherein the step of performing the reference annotation-based external boundary constraint on the weight matrix of the classifier includes:
performing external boundary constraint based on reference annotation on the weight matrix of the classifier by using the following external boundary constraint formula to obtain an optimized weight matrix;
wherein the external boundary constraint formula is:
wherein , and />The weight matrix of last and current iteration, respectively,/->Is the server load time sequence associated feature vector, < >>Is a first transition matrix, < >>Is a second transition matrix, ">Representing matrix multiplication +.>Representing matrix addition, ++>Representing the transpose of the vector>Is the optimized weight matrix.
9. A restart instruction processing apparatus of a sink splitter, characterized in that the restart instruction processing apparatus of a sink splitter operates in the method according to claims 1 to 8.
Priority Applications (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN202311090831.8A CN116827759B (en) | 2023-08-29 | 2023-08-29 | Method and device for processing restarting instruction of converging current divider |
Applications Claiming Priority (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN202311090831.8A CN116827759B (en) | 2023-08-29 | 2023-08-29 | Method and device for processing restarting instruction of converging current divider |
Publications (2)
| Publication Number | Publication Date |
|---|---|
| CN116827759A true CN116827759A (en) | 2023-09-29 |
| CN116827759B CN116827759B (en) | 2023-11-03 |
Family
ID=88124332
Family Applications (1)
| Application Number | Title | Priority Date | Filing Date |
|---|---|---|---|
| CN202311090831.8A Active CN116827759B (en) | 2023-08-29 | 2023-08-29 | Method and device for processing restarting instruction of converging current divider |
Country Status (1)
| Country | Link |
|---|---|
| CN (1) | CN116827759B (en) |
Citations (6)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN105228173A (en) * | 2014-07-02 | 2016-01-06 | 上海新联纬讯科技发展有限公司 | Possesses the wireless router method of operation of loading progress interactive function |
| US20200136890A1 (en) * | 2018-10-24 | 2020-04-30 | Affirmed Networks, Inc. | Anomaly detection and classification in networked systems |
| CN112380047A (en) * | 2020-09-25 | 2021-02-19 | 深圳中科德能科技有限公司 | Method and device for restarting switch, terminal equipment and storage medium |
| CN113067769A (en) * | 2021-03-31 | 2021-07-02 | 百度在线网络技术（北京）有限公司 | Router restarting method, device, equipment and storage medium |
| CN113645341A (en) * | 2021-07-16 | 2021-11-12 | 荣耀终端有限公司 | Power key false touch detection method and electronic equipment |
| WO2022056081A1 (en) * | 2020-09-10 | 2022-03-17 | Arris Enterprises Llc | A wireless device and a method for automatic recovery from failures |
-
2023
- 2023-08-29 CN CN202311090831.8A patent/CN116827759B/en active Active
Patent Citations (6)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN105228173A (en) * | 2014-07-02 | 2016-01-06 | 上海新联纬讯科技发展有限公司 | Possesses the wireless router method of operation of loading progress interactive function |
| US20200136890A1 (en) * | 2018-10-24 | 2020-04-30 | Affirmed Networks, Inc. | Anomaly detection and classification in networked systems |
| WO2022056081A1 (en) * | 2020-09-10 | 2022-03-17 | Arris Enterprises Llc | A wireless device and a method for automatic recovery from failures |
| CN112380047A (en) * | 2020-09-25 | 2021-02-19 | 深圳中科德能科技有限公司 | Method and device for restarting switch, terminal equipment and storage medium |
| CN113067769A (en) * | 2021-03-31 | 2021-07-02 | 百度在线网络技术（北京）有限公司 | Router restarting method, device, equipment and storage medium |
| CN113645341A (en) * | 2021-07-16 | 2021-11-12 | 荣耀终端有限公司 | Power key false touch detection method and electronic equipment |
Also Published As
| Publication number | Publication date |
|---|---|
| CN116827759B (en) | 2023-11-03 |
Similar Documents
| Publication | Publication Date | Title |
|---|---|---|
| US10579459B2 (en) | Log events for root cause error diagnosis | |
| US10552729B2 (en) | Baseboard management controller to deconfigure field replaceable units according to deep learning model | |
| US10795753B2 (en) | Log-based computer failure diagnosis | |
| US10581667B2 (en) | Method and network node for localizing a fault causing performance degradation of a service | |
| US11551085B2 (en) | Method, device, and computer program product for error evaluation | |
| CN106293798B (en) | Self-repairing method and system of electronic device and server | |
| US20180357143A1 (en) | Testing computing devices | |
| CN114579380A (en) | Artificial intelligence detection system and method for computer system fault | |
| US10938623B2 (en) | Computing element failure identification mechanism | |
| CN111859047A (en) | Fault solving method and device | |
| CN109684309A (en) | A kind of quality of data evaluating method and device, computer equipment and storage medium | |
| WO2024148857A1 (en) | Method and apparatus for filtering root cause of server fault, and non-volatile readable storage medium and electronic apparatus | |
| CN118041760A (en) | Intelligent self-repairing method for connector network | |
| Kalimulina | Analysis of system reliability with control, dependent failures, and arbitrary repair times | |
| US11563671B2 (en) | Routing engine switchover based on health determined by support vector machine | |
| CN112311574A (en) | Method, device and equipment for checking network topology connection | |
| CN116827759B (en) | Method and device for processing restarting instruction of converging current divider | |
| CN113839861A (en) | Routing engine switching based on health determined by support vector machine | |
| Simeonov et al. | Proactive software rejuvenation based on machine learning techniques | |
| US11665165B2 (en) | Whitelist generator, whitelist evaluator, whitelist generator/evaluator, whitelist generation method, whitelist evaluation method, and whitelist generation/evaluation method | |
| CN114385398A (en) | Request response state determination method, device, equipment and storage medium | |
| CN115544202A (en) | Alarm processing method, device and storage medium | |
| CN111949281A (en) | Database installation method based on AI configuration, user equipment and storage medium | |
| US11956117B1 (en) | Network monitoring and healing based on a behavior model | |
| CN117762533A (en) | Multi-loop throttle processing method and device for service data and electronic equipment |
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 |