10results about "Error identification" patented technology

The invention provides an embedded system crash prediction method and device based on multi-source dynamic perception, and the method comprises the steps: carrying out the stack boundary monitoring and / or interruption storm detection of a system, and obtaining a detection result; performing deadlock probability calculation and / or memory leak trend analysis on the system to obtain an analysis result; and performing risk assessment on the detection result and the analysis result to obtain a crash risk value, comparing the crash risk value with a preset threshold value, and when the crash risk value exceeds the preset threshold value, judging that the system has a crash risk. According to the method, stack boundary monitoring and / or interruption storm detection are / is carried out on the system, deadlock probability calculation and / or memory leakage trend analysis are / is carried out on the system, a more comprehensive detection result and an analysis result are obtained, risk assessment is carried out on the basis, and a crash risk value is obtained; and the crash risk value is compared with the preset threshold value to predict the system crash in advance, so that the problems that the misjudgment rate is high and the crash cannot be prevented in the prior art are solved.

A performance anomaly is detected in relation to a first data center equipment. An AI model is trained based on a plurality of anomaly patterns associated with a plurality of data center equipment and respective remediation processes associated with the anomaly patterns, to determine a remediation process that can resolve the detected performance anomaly associated with the first data center equipment. Upon execution of a machine-learning algorithm, the AI model compares real-time performance indicators recorded for the first data center equipment with performance indicators of the anomaly patterns and determines a matching anomaly pattern. The AI model determines a particular remediation process associated with the matching anomaly pattern. The remediation process is then implemented to resolve the performance anomaly associated with the first data center equipment.

An AI model is trained based on a plurality of anomaly patterns associated with a data center equipment to predict a performance anomaly associated with the data center equipment. Upon execution of a machine-learning algorithm, the AI model compares real-time performance indicators associated with the data center equipment to historical performance indicators of the anomaly patterns and determines a matching anomaly pattern. The AI model identifies a performance anomaly associated with the matching anomaly pattern and predicts that the performance anomaly is expected to occur in relation to the data center equipment. A remediation method is then implemented to avoid the performance anomaly from occurring.

A memory stores management information where identification information of each of a plurality of devices used by an information processing apparatus, first positional information, and second positional information are associated with one another. The first positional information indicates a position of a device storage storing the plurality of devices. The second positional information indicates a storage position of each of the plurality of devices in the device storage. A processor receives failed device information including identification information of a failed device among the plurality of devices from the information processing apparatus. The processor identifies the position of the device storage storing the failed device and the storage position of the failed device in the device storage from the identification information of the failed device included in the failed device information on the basis of management information.

Architectures and techniques are described that can leverage a container storage interface (CSI) driver to initiate volume health tests for a storage system in accordance with certain embodiments of this disclosure. The CSI driver can conform to a CSI specification or standard and can be used to expose the storage system to a container orchestration system. The volume health tests can be tailored to reduce or mitigate the impact of testing based on a status or state of the volumes. For instance, enforcement can be provided to ensure that only shorter, low-impact testing is performed on volumes having a published state, whereas longer, more intensive testing can be performed on volumes that are not in the published state.

The present disclosure provides an agent capability description method, a task execution method, an apparatus and an agent, and relates to the technical field of artificial intelligence, in particular to the technical fields of deep learning, large model, human-computer interaction, intelligent education, video generation and the like. The agent capability description method comprises: obtaining a task example for a to-be-tested agent; detecting, by a master agent, an example execution result to obtain an example detection result, the example execution result being determined based on the to-be-tested agent executing the task example, and the example detection result representing a matching degree between the example execution result and a task requirement condition for the task example; and describing, by the master agent, a task execution capability of the to-be-tested agent based on the example detection result and the task example to obtain capability description information for the to-be-tested agent.

The application discloses a deep learning-based management and control monitoring system and relates to the technical field of computers, which is used to solve the problem that traditional management and control monitoring systems usually rely on manual operation or simple automatic scripts and are difficult to cope with high concurrency and high reliability requirements in complex linkage scenarios; business actions are constructed into atomic action graphs, and graph rewriting and two-stage convergence batching are performed in combination with mutual exclusion matrices, so that parallelism is maximized and resource contention is avoided; an idempotent token is generated based on a plan fingerprint, and a send-only additional box is used to realize first-time successful reply reuse and at most once side effect semantics; through pre-rehearsal and state transition graphs, a set of main keys that can be aligned is provided, so that execution prerequisites can be quickly verified; three verification levels, i.e., an interface level, a state level and a side certificate level, are used as a decision mechanism, and the side certificate is a veto criterion, so that stop loss can be quickly performed when false success or late success occurs.

Methods and systems for managing operation of a deployment comprising data processing systems are disclosed. The operation may be managed by optimizing a diagnostic approach and / or a solution for an issue of an operation by a data processing system. The diagnostic approach may be optimized by using a trained inference model and / or at least one similarity map to identify at least one other data processing system. The at least one other data processing system may include contextual data for the issue. The data processing system may provide the issue and / or the contextual data to a large language model and / or a tree of thought model. The data processing system may use the tree of thought model to generate a remediative procedure for the issue.

Architectures and techniques are described that can leverage a container storage interface (CSI) driver to initiate volume health tests for a storage system in accordance with certain embodiments of this disclosure. The CSI driver can conform to a CSI specification or standard and can be used to expose the storage system to a container orchestration system. The volume health tests can be tailored to reduce or mitigate the impact of testing based on a status or state of the volumes. For instance, enforcement can be provided to ensure that only shorter, low-impact testing is performed on volumes having a published state, whereas longer, more intensive testing can be performed on volumes that are not in the published state.

The application discloses a consensus timeout response method and device in a consortium chain. One master copy node participating in consensus processing and a plurality of slave copy nodes with consensus states are determined, and the master copy node and the plurality of slave copy nodes are located in the same view. A challenge message of the master copy node responding to the consensus request timeout broadcast by the slave copy node in the view is collected, and the number of the slave copy nodes broadcasting the challenge message is counted. All the challenge messages are spliced into a challenge proof in the order of the serial identification codes of the slave copy nodes broadcasting the challenge messages. A confirmation message generated by responding to the challenge message broadcast by the slave copy node in the view is collected, and the number of the slave copy nodes broadcasting the confirmation message is counted. All the confirmation messages are spliced into a confirmation proof in the order of the serial identification codes of the slave copy nodes broadcasting the confirmation messages. Whether the master copy node is actually responding to the consensus request timeout is quickly judged according to the challenge proof and the confirmation proof.