35 results about "Software aging" patented technology

An approach is provided that rejuvenates a software application to reduce the effects of software aging. An active replica corresponding to a software application is identified. If rejuvenation of the software application is appropriate, a new replica is created and state information is transferred from the active replica to the new replica. In addition, client requests are redirected to the new replica. After the state data has been transferred and requests have been redirected, the active replica is terminated. Once the active replica has been terminated, the new replica becomes the active replica. When rejuvenation is again proper, another new replica is created and the state data is transferred from the new active replica to the new replica and requests are redirected to the new replica. This process repeats whenever rejuvenation of the application is needed.

The invention provides a software aging streaming monitoring system, which belongs to the research field of distributed systems. For the host nodes in the existing distributed system, the data collection module is designed to collect the memory information of each host node and upload the information to the message queue module. Then, according to the memory data in the message queue module, the real-time exception discovery module uses the bucket algorithm to judge whether the memory in each host node is overloaded, and the software aging prediction module predicts the failure and abnormal time of each host node. Finally, the operation and maintenance personnel obtain the data processing results from the display module. The invention solves the problem of high false alarm rate and missing alarm rate of abnormal early warning based on the threshold of traditional distributed system monitoring tools, and solves the problem that the traditional software aging analysis method is difficult to cope with the daily TB-level streaming data of the distributed system.

The invention discloses a software aging prediction method and device based on multi-model comparison, belonging to the field of software aging, collecting aging indicators from target software systems, processing them into time series data as pre-input of models; aiming at aging data scale, designing Including the aging prediction model of machine learning and neural network, calculate the prediction error of each model, select the model with the smallest error as the candidate model; calculate whether there is a significant difference between the model and other models, if the difference is obvious, select the model for the final aging prediction model. The invention solves the problem that the prediction result of a single model may affect the decision-making, and the user can automatically select a suitable model according to the characteristics of aging data and prediction errors, avoiding the early or late implementation of proactive maintenance measures and reducing the reliability of the software. sexual influence. More models can be expanded, and the optimal prediction model can be selected for different aging data scales to help system operation and maintenance.

The invention provides a method for predicting concurrent business of different types of virtual machines, and relates to the technical field of cloud computing. A business concurrency prediction method for different types of virtual machines. Firstly, the historical business concurrency of virtual machines is collected and preprocessed, and then the virtual machine business concurrency is judged based on the improved 1nearest neighbor-dynamic time adjustment method 1NN-DTW type; finally, the classification regression tree is used to fit the business concurrency without periodic changes; the Fourier series FS and classification regression tree CART are used to fit the business concurrency with periodic changes; the invention provides different types of virtual machines The business concurrency prediction method based on this method predicts the concurrency of each business of the virtual machine, which can provide a basis for the increase or decrease of the virtual machine in the next step, and at the same time help to accurately estimate the software aging status of the virtual machine, so as to improve the performance of the virtual machine. performance and reliability purposes.

Load test is executed with an appropriate frequency which does not lead to a decrease in software development efficiency and a decrease in the precision of software aging detection. Load test of a version of software under test is executed in accordance with an execution criterion, presence or absence of a software aging problem is detected by comparing a test result of the load test with a test result of load test of a previous version of the software to be compared, and frequency of execution of subsequent load test is adjusted by changing the execution criterion based on a result of the detection.

The invention discloses a cross-project software aging prediction method. First, the data in the source project and the target project are preprocessed, and then the joint distribution domain is used to adapt to reduce the difference between the marginal distribution and the conditional distribution, and then the undersampling method and the improved The subclass discriminant analysis method alleviates the class imbalance problem, and finally uses a machine learning classifier (logistic regression, etc.) to make predictions. The invention considers the conditional distribution difference between the source item and the target item of the software aging defect data set, and further adopts an improved subclass discriminant analysis method to alleviate the extremely serious class imbalance problem. It solves the problem of low accuracy and robustness of traditional cross-project software aging defect prediction methods, and helps developers find and remove software aging-related defects during the development and testing phase, avoiding losses caused by software aging problems. The feasibility of the invention has been verified on real software, and can be extended to other software to predict defects related to software aging.

The invention discloses an Active Learning-based software aging prediction method in a project. By collecting static metrics of codes in the software, Active Learning is used to select samples and label them as a training set to predict the remaining samples without class labels. Active Learning is used for sample selection and manual labeling to form a training set. A combination of oversampling and undersampling is used to alleviate the class imbalance problem, and a machine learning classifier is used for prediction. The present invention considers that there are few samples in the software aging defect data set, and the collection is time-consuming and labor-intensive. The method of combining under-sampling and over-sampling is used to alleviate the problem of extreme class imbalance, which helps developers to find software aging-related defects during the development and testing stage. Removed to avoid losses caused by software aging issues. The feasibility of the invention has been verified on real software, and can be extended to other software to predict defects related to software aging.

The invention relates to a system re-graduating method of group environment applied server which tests weather the current server node is aging by aging formation, if it is aging, it sets the node as testing point, ascertains the current server node condition information of the EJB by the condition replicating operation and replicates the node condition to the backup node of anther server; the server calls the local operating system Windows API-ExitWindowsEx to restart the server node which comes into the group after restarting and in step with the other nodes and replicates the new condition of the current backup node to the local node; the applied server accepts the EJB application.

The invention discloses a software aging detection method, and the software aging detection method comprises the following steps: acquiring the total usage amount of piles corresponding to test software and target software at different moments, wherein the test software simulates the operation of the target software, and the test software is injected into a memory leak hole; determining a divergence value corresponding to each moment according to a first total usage amount corresponding to the test software and a second total usage amount corresponding to the target software, wherein the divergence value is a difference amount between the first total usage amount and the second total usage amount at the same moment; when each divergence value is less than a preset threshold, judging that the target software is not aged. The invention further discloses a software aging detection device and a computer readable storage medium. According to the method, the aging of the software is accurately determined.