An intelligent diagnosis method for storage operation and maintenance based on experience learning and hierarchical synchronization

Abstract

This invention relates to the field of information technology operation and maintenance technology, and provides a storage operation and maintenance intelligent diagnosis method based on experience learning and hierarchical synchronization. The method includes: intercepting diagnostic commands through an operation and maintenance terminal and associating them with system context information to generate a standardized experience data package; transmitting the standardized experience data package to a knowledge platform center based on the network connectivity status of the operation and maintenance terminal's environment; deduplicating, evaluating, and merging the received standardized experience data package through the knowledge platform center to update the global knowledge base; generating a full health status snapshot upon receiving a diagnostic request; performing feature vectorization matching between the current diagnostic information and historical experience, outputting the historical experience with the highest matching degree and its solution, and receiving verification feedback from operation and maintenance personnel to form a closed-loop learning process. This invention can realize the assetization and self-evolution mechanism of operation and maintenance knowledge, enhance the determinism and reliability of the diagnostic process, and significantly improve operation and maintenance efficiency and response speed.

Description

Technical Field

[0001] This invention relates to the field of information technology operation and maintenance technology, and in particular to a storage operation and maintenance intelligent diagnostic method based on experience learning and hierarchical synchronization. Background Technology

[0002] In the daily operation and maintenance of storage systems, fault diagnosis and problem handling are core aspects of ensuring business continuity and system stability. As storage systems continue to expand in scale and their architecture becomes increasingly complex, traditional operation and maintenance models relying on manual experience are facing severe challenges. Storage devices involve complex interactions at the hardware layer (disks, RAID, network), system layer (operating system, file system), and application layer (storage services, scheduling policies). The correlation between fault symptoms and root causes often exhibits non-linear, cross-layered relationships, making diagnosis extremely difficult. Therefore, building an intelligent diagnostic method that can accumulate expert experience, enable knowledge reuse, and possess self-evolving capabilities has become an essential requirement for improving the efficiency and quality of storage operation and maintenance.

[0003] However, existing technologies still have the following obvious shortcomings and defects in addressing the above challenges:

[0004] First, the problem of siloed experience is prominent. The diagnostic experience of operations and maintenance personnel exists only in their personal minds or scattered records, lacking a standardized expression and accumulation mechanism. Second, the phenomenon of repeated problem localization is serious. The same or similar faults occur repeatedly at different times and locations, but due to the lack of an effective experience reuse mechanism, each diagnosis requires starting from scratch, consuming a lot of manpower and time, resulting in low operational efficiency. Third, the level of automation and intelligence is low. Existing monitoring tools mainly remain at the alarm level, lacking the ability to learn from historical diagnoses and unable to continuously optimize their judgment accuracy as the number of diagnoses increases. Fourth, network isolation environments restrict knowledge flow. In actual production environments, due to strict network security policy requirements, production sites are usually physically isolated from external networks, and the valuable operational experience generated during their operation cannot be easily transmitted back to the knowledge center as in a laboratory environment.

[0005] Chinese patent CN121641369A discloses a deep learning-based operation and maintenance knowledge base analysis system and method. This solution lacks a structured encapsulation of operation and maintenance experience and a secure circulation mechanism across network isolation environments. Furthermore, its knowledge retrieval relies on difficult-to-interpret deep learning feature similarity matching and does not introduce a precise root cause diagnosis path based on deterministic health status snapshots.

[0006] Therefore, how to provide a storage operation and maintenance intelligent diagnostic method that can effectively accumulate expert experience, realize cross-environment knowledge flow, and has continuous learning capabilities has become an urgent technical problem to be solved. Summary of the Invention

[0007] In view of this, in order to overcome the shortcomings of the prior art, the present invention aims to provide a storage operation and maintenance intelligent diagnosis method based on experience learning and hierarchical synchronization.

[0008] This invention provides a storage operation and maintenance intelligent diagnostic method based on experience learning and hierarchical synchronization, the method comprising:

[0009] Step S1: Intercept diagnostic commands through the operation and maintenance terminal and associate them with system context information to generate a standardized experience data package;

[0010] Step S2: Based on the network connectivity of the environment where the maintenance terminal is located, select either online direct connection or offline export method to transfer the standardized experience data package to the knowledge platform center;

[0011] Step S3: The received standardized experience data packages are deduplicated, evaluated, and merged through the knowledge platform center to update the global knowledge base;

[0012] Step S4: Upon receiving a diagnostic request, select either experience-driven mode or root cause-driven mode for diagnosis. Root cause-driven mode generates a full health status snapshot by concurrently executing a predefined root cause check set.

[0013] Step S5: Perform feature vectorization matching between the current diagnostic information and historical experience, output the historical experience with the highest matching degree and its solution, and receive verification feedback from operation and maintenance personnel to form a closed-loop learning.

[0014] Optionally, in the storage operation and maintenance intelligent diagnosis method based on experience learning and hierarchical synchronization of the present invention, the specific method for generating standardized experience data packages in step S1 is as follows:

[0015] Intercept diagnostic commands initiated by the operation and maintenance terminal, and collect system context information when the command is executed. The system context information includes user identity, the target being diagnosed, system timestamp, and system status snapshot.

[0016] The diagnostic commands, the output results after command execution, and system context information are encapsulated and bound to form a preliminary experience record;

[0017] Receive and record the root causes and solutions summarized by operations and maintenance personnel during the diagnostic process, and integrate them into a complete standardized experience data package.

[0018] Optionally, in the storage operation and maintenance intelligent diagnosis method based on experience learning and hierarchical synchronization of the present invention, in step S1, the standardized experience data package is stored using a structured operation and maintenance experience template, which includes at least the following interrelated data groups:

[0019] Problem feature groups are used to record descriptions of fault symptoms and system health snapshots;

[0020] The diagnostic process group records the sequence of examination commands actually executed during the diagnostic process and the root cause that is ultimately determined.

[0021] Solution groups are used to record the specific operational steps for resolving faults and the associated automated script identifiers;

[0022] The validation feedback group is used to record the historical validity feedback results and reuse statistics of the experience data package.

[0023] Optionally, in the intelligent diagnostic method for storage operation and maintenance based on experience learning and hierarchical synchronization of the present invention, the specific implementation of transmitting the experience data package using the offline export method in step S2 is as follows:

[0024] The synchronization module deployed in the production environment periodically or when preset trigger conditions are met compares the local experience base with the last successfully exported baseline version to identify newly added, modified, or changed experience data packages.

[0025] The identified incremental experience data is packaged and an integrity check code is generated to form an offline transmission package that can be transmitted through physical media;

[0026] At the same time, a list file describing the details of this export is generated to enable offline reporting of operational experience in a completely isolated production environment.

[0027] Optionally, in the storage operation and maintenance intelligent diagnosis method based on experience learning and hierarchical synchronization of the present invention, step S3 involves deduplication, evaluation, and merging of the received standardized experience data packets in the following manner:

[0028] The knowledge platform center determines the logical relationship between newly received standard experience data packages and existing standard experience data packages based on their source identifiers, timestamp information, confidence scores, and feature vector similarity.

[0029] If the report is determined to be a duplicate of the same fault event, the number of verifications of the original experience data package is increased and its confidence level is improved.

[0030] If they are determined to be different manifestations of the same root cause or supplements to the solutions, merge the relevant information to enrich the content of the original experience data package.

[0031] If a diagnostic conclusion is determined to be substantially conflicting, the experience data package is marked as a conflict status pending manual review, and the evidence from both conflicting parties is stored together.

[0032] Optionally, in the storage operation and maintenance intelligent diagnosis method based on experience learning and hierarchical synchronization of the present invention, in step S3, after the knowledge platform center completes the global knowledge base update, it actively distributes the updated experience data package to each operation and maintenance terminal through the synchronization center, so as to realize the real-time sharing and collaborative evolution of operation and maintenance experience across the entire domain.

[0033] Optionally, in the storage operation and maintenance intelligent diagnosis method based on experience learning and hierarchical synchronization of the present invention, the specific implementation of selecting the root cause-driven mode for diagnosis in step S4 is as follows:

[0034] A standardized set of check items covering the key fault domains of the storage system is pre-built. These key fault domains include at least disk status, RAID array, network connectivity, memory usage, CPU load, and service process running status.

[0035] When a diagnosis is triggered, all check commands in the standardized check item set are executed concurrently or in parallel.

[0036] Collect the structured output results returned after each check command is executed, and summarize them to generate a unified, machine-readable full health status snapshot.

[0037] Optionally, in the intelligent diagnostic method for storage operation and maintenance based on experience learning and hierarchical synchronization of the present invention, the specific implementation of selecting the experience-driven mode for diagnosis in step S4 is as follows:

[0038] Receive fault description information manually entered by maintenance personnel or automatically reported by the monitoring system;

[0039] The fault phenomenon description information is matched with the problem feature layer of the experience templates stored in the knowledge base. The problem feature layer contains at least historical fault phenomenon descriptions and corresponding historical system health snapshots.

[0040] When the matching degree exceeds the preset similarity threshold, the historical diagnostic conclusions and solutions recorded in the successfully matched experience template are output.

[0041] Optionally, in the storage operation and maintenance intelligent diagnosis method based on experience learning and hierarchical synchronization of the present invention, in step S5, the current diagnostic information and historical experience are matched by feature vectorization in the following manner: the diagnostic information of the new fault and the experience template in the historical experience library are respectively converted into multi-dimensional feature vectors, and the weighted similarity between the new fault vector and the historical experience vector is calculated to select the historical experience with the highest similarity as the recommendation basis.

[0042] Optionally, in the storage operation and maintenance intelligent diagnosis method based on experience learning and hierarchical synchronization of the present invention, the specific implementation of receiving verification feedback from operation and maintenance personnel to form a closed-loop learning in step S5 is as follows:

[0043] The operations and maintenance personnel confirm the effectiveness of the recommended solutions, or revise the recommendations when necessary, and record the specific content of the feedback.

[0044] Based on feedback from operations and maintenance personnel, the confidence score and historical reuse count of this experience data package are dynamically updated.

[0045] The entire diagnostic process is encapsulated into a new standardized experience data package, which includes a full health status snapshot generated during the diagnosis, the final diagnostic conclusions, and the solutions adopted. This new standardized experience data package is then transmitted in a layered and synchronized manner.

[0046] This invention, based on experience-based learning and hierarchical synchronization, provides an intelligent diagnostic method for storage operation and maintenance, which has the following beneficial technical effects:

[0047] I. Realize the assetization and self-evolution mechanism of operation and maintenance knowledge.

[0048] This enables the persistent accumulation, precise management, and intelligent application of operational knowledge. Through subsequent intelligent matching, verification feedback, and deduplication processes, the diagnostic knowledge base can continuously learn and autonomously evolve, overcoming the limitations of traditional static knowledge bases.

[0049] II. Enhance the certainty and reliability of the diagnostic process.

[0050] By proactively and concurrently executing a "root cause check set" covering the entire component layer, a snapshot of deterministic health status is directly obtained, avoiding subjective misjudgment from the source. Combined with high-precision historical template matching based on feature vectors, the diagnostic conclusions have a solid objective data foundation, and the accuracy is significantly improved.

[0051] Third, by adopting a "layered synchronization" strategy that adapts to complex network environments, barriers to knowledge flow can be broken down.

[0052] Under the premise of strictly adhering to the highest level of network security regulations, we can achieve secure, controllable, efficient aggregation and real-time distribution of operational experience across the entire domain, breaking down traditional data silos.

[0053] IV. Significantly improves operational efficiency and response speed.

[0054] It provides proven solutions directly for common faults, significantly reducing the average repair time. At the same time, predefined root cause check sets can be executed automatically and periodically, transforming operations and maintenance from a passive "fault response" to a proactive "health prevention," preventing problems before they occur.

[0055] Fifth, standardize and equalize team capabilities.

[0056] By leveraging the overall technical capabilities of the team, we reduce over-reliance on specific individuals. All diagnostic operations are logged on the platform, enabling full traceability of the maintenance process, auditability of operations, and reviewability of decisions, greatly improving the standardization and security of our work.

[0057] VI. It has excellent fault location capabilities.

[0058] By comparing full health snapshots, cases with the same root cause pattern can be quickly matched from historical experience, directly revealing their hidden root causes and providing effective solutions that have been tested in practice, greatly reducing the difficulty of troubleshooting difficult problems.

[0059] VII. It can lead to innovative product and service models.

[0060] For hardware and software manufacturers, integrating this system as a core function can create a strong technological barrier and product differentiation, allowing them to escape homogeneous competition. For end users, it can effectively reduce their total cost of ownership and business risk. Attached Figure Description

[0061] To more clearly illustrate the technical solutions of the embodiments of the present invention, the drawings used in the embodiments will be briefly introduced below. Obviously, the drawings described below are only some embodiments of the present invention. For those skilled in the art, other drawings can be obtained based on these drawings without creative effort.

[0062] Figure 1 This is a flowchart illustrating the intelligent diagnostic method for storage operation and maintenance based on experience learning and hierarchical synchronization according to an exemplary embodiment 1 of the present invention.

[0063] Figure 2 A flowchart illustrating the process of generating standardized experience data packages for the storage operation and maintenance intelligent diagnostic method based on experience learning and hierarchical synchronization according to an exemplary embodiment 1 of the present invention;

[0064] Figure 3 This is a schematic diagram of the structured experience template data organization of the intelligent diagnosis method for storage operation and maintenance based on experience learning and hierarchical synchronization according to an exemplary embodiment 1 of the present invention;

[0065] Figure 4 This is a schematic diagram of the offline export method for incremental synchronization of the storage operation and maintenance intelligent diagnosis method based on experience learning and hierarchical synchronization according to an exemplary embodiment 1 of the present invention.

[0066] Figure 5 This is a schematic diagram of the process of resolving and merging knowledge base conflicts in the intelligent diagnosis method for storage operation and maintenance based on experience learning and hierarchical synchronization according to an exemplary embodiment 1 of the present invention.

[0067] Figure 6 This is a flowchart illustrating the knowledge base update and global distribution process of the storage operation and maintenance intelligent diagnosis method based on experience learning and hierarchical synchronization according to an exemplary embodiment 1 of the present invention.

[0068] Figure 7 This is a schematic diagram of the process of the intelligent diagnostic method for storage operation and maintenance based on experience learning and hierarchical synchronization according to an exemplary embodiment 1 of the present invention, which actively checks through a root cause-driven mode.

[0069] Figure 8 This is a flowchart illustrating the experience-driven pattern matching process of the intelligent diagnostic method for storage operation and maintenance based on experience learning and hierarchical synchronization according to an exemplary embodiment 1 of the present invention.

[0070] Figure 9 This is a flowchart illustrating the feature vectorization intelligent matching process of the storage operation and maintenance intelligent diagnosis method based on experience learning and hierarchical synchronization according to an exemplary embodiment 1 of the present invention.

[0071] Figure 10 This is a flowchart illustrating the verification feedback and closed-loop learning process of the storage operation and maintenance intelligent diagnostic method based on experience learning and hierarchical synchronization according to an exemplary embodiment 1 of the present invention. Detailed Implementation

[0072] The embodiments of the present invention will now be described in detail with reference to the accompanying drawings.

[0073] It should be noted that, in the absence of conflict, the following embodiments and features can be combined with each other; and, based on the embodiments of this disclosure, all other embodiments obtained by those skilled in the art without creative effort are within the scope of protection of this disclosure.

[0074] It should be noted that various aspects of embodiments within the scope of the appended claims are described below. It will be apparent that the aspects described herein can be embodied in a wide variety of forms, and any particular structure and / or function described herein is merely illustrative. Based on this disclosure, those skilled in the art will understand that one aspect described herein can be implemented independently of any other aspect, and two or more of these aspects can be combined in various ways. For example, any number of aspects set forth herein can be used to implement the device and / or practice the method. Additionally, this device and / or method can be implemented using structures and / or functionalities other than one or more of the aspects set forth herein.

[0075] Example 1

[0076] Exemplary embodiment 1 of the present invention provides a storage operation and maintenance intelligent diagnosis method based on experience learning and hierarchical synchronization. Figure 1 This is a flowchart illustrating the intelligent storage operation and maintenance diagnosis method based on experience learning and hierarchical synchronization according to an exemplary embodiment 1 of the present invention, as follows: Figure 1 As shown, in this embodiment, the method of the present invention is implemented in the following manner:

[0077] Step S1: Intercept diagnostic commands through the operation and maintenance terminal and associate them with system context information to generate a standardized experience data package.

[0078] Figure 2 A flowchart illustrating the process of generating standardized experience data packages for the storage operation and maintenance intelligent diagnosis method based on experience learning and hierarchical synchronization according to Exemplary Embodiment 1 of the present invention is shown below. Figure 2 As shown, in this embodiment, the specific method for generating standardized experience data packages is as follows:

[0079] Intercept diagnostic commands initiated by the operation and maintenance terminal, collect system context information when the command is executed, including user identity, the target being diagnosed, system timestamp, and system status snapshot; encapsulate and bind the diagnostic command, the output results after command execution, and the system context information to form a preliminary experience record; receive and record the root causes and solutions summarized by the operation and maintenance personnel for the diagnostic process, and integrate them into a complete standardized experience data package.

[0080] Figure 3 This is a schematic diagram illustrating the structured experience template data organization of the intelligent storage operation and maintenance diagnosis method based on experience learning and hierarchical synchronization according to an exemplary embodiment 1 of the present invention, as shown below. Figure 3 As shown, in this embodiment, the standardized experience data package is stored using a structured operation and maintenance experience template, which includes at least the following interrelated data groups:

[0081] Problem feature groups are used to record descriptions of fault symptoms and system health snapshots;

[0082] The diagnostic process group records the sequence of examination commands actually executed during the diagnostic process and the root cause that is ultimately determined.

[0083] Solution groups are used to record the specific operational steps for resolving faults and the associated automated script identifiers;

[0084] The validation feedback group is used to record the historical validity feedback results and reuse statistics of the experience data package.

[0085] Step S2: Based on the network connectivity of the environment where the maintenance terminal is located, select either online direct connection or offline export method to transfer the standardized experience data package to the knowledge platform center.

[0086] Figure 4This is a schematic diagram illustrating the offline export method for incremental synchronization of the storage operation and maintenance intelligent diagnostic method based on experience learning and hierarchical synchronization according to an exemplary embodiment 1 of the present invention. Figure 4 As shown, in this embodiment, the specific implementation method of transmitting the experience data package using the offline export method is as follows:

[0087] The synchronization module deployed in the production environment periodically or when preset trigger conditions are met compares the local experience base with the last successfully exported baseline version, identifying newly added, modified, or changed experience data packages; it packages the identified incremental experience data and generates an integrity check code to form an offline transmission package that can be transmitted through physical media; at the same time, it generates a list file describing the details of this export, which enables offline reporting of operation and maintenance experience in a completely isolated state in the production environment.

[0088] Step S3: The received standardized experience data packages are deduplicated, evaluated, and merged through the knowledge platform center to update the global knowledge base.

[0089] Figure 5 This is a schematic diagram illustrating the process of knowledge base conflict resolution and merging in the intelligent storage operation and maintenance diagnosis method based on experience learning and hierarchical synchronization according to exemplary embodiment 1 of the present invention. Figure 5 As shown, in this embodiment, the received standardized experience data packets are deduplicated, evaluated, and merged in the following manner:

[0090] The knowledge platform center determines the logical relationship between newly received standard experience data packages and existing standard experience data packages based on their source identifiers, timestamp information, confidence scores, and feature vector similarity. If the package is determined to be a duplicate report of the same fault event, the verification count of the original experience data package is increased and its confidence score is improved. If the package is determined to be a different manifestation of the same root cause or a supplement to the solution, relevant information is merged to enrich the content of the original experience data package. If the package is determined to be in a substantial conflict in the diagnostic conclusion, the experience data package is marked as a conflict pending manual review, and the evidence from both conflicting parties is stored together.

[0091] Figure 6 This is a flowchart illustrating the knowledge base update and global distribution process of the storage operation and maintenance intelligent diagnosis method based on experience learning and hierarchical synchronization according to Exemplary Embodiment 1 of the present invention. Figure 6 As shown in this embodiment, after completing the global knowledge base update, the knowledge platform center actively distributes the updated experience data package to each operation and maintenance terminal through the synchronization hub, so as to realize the real-time sharing and collaborative evolution of operation and maintenance experience across the entire domain.

[0092] Step S4: Upon receiving a diagnostic request, select either experience-driven mode or root cause-driven mode for diagnosis. Root cause-driven mode generates a full health status snapshot by concurrently executing a predefined root cause check set.

[0093] Figure 7 This is a schematic diagram illustrating the process of proactively checking storage operation and maintenance intelligent diagnostic method based on experience learning and hierarchical synchronization according to exemplary embodiment 1 of the present invention through a root cause-driven mode. Figure 7 As shown, in this embodiment, the specific implementation method of selecting the root cause-driven mode for diagnosis is as follows:

[0094] A standardized set of check items covering the critical fault domains of the storage system is pre-built. These critical fault domains include at least disk status, RAID array, network connectivity, memory usage, CPU load, and service process running status. When a diagnosis is triggered, all check commands in the standardized set of check items are executed concurrently or in parallel. The structured output results returned after the execution of each check command are collected and summarized to generate a unified, machine-readable full health status snapshot.

[0095] Figure 8 This is a flowchart illustrating the experience-driven pattern matching process of the intelligent storage operation and maintenance diagnosis method based on experience learning and hierarchical synchronization according to an exemplary embodiment 1 of the present invention. Figure 8 As shown, in this embodiment, the specific implementation method of selecting the experience-driven mode for diagnosis is as follows:

[0096] Receive fault description information manually entered by maintenance personnel or automatically reported by the monitoring system; match the fault description information with the problem feature layer of the experience template stored in the knowledge base. The problem feature layer contains at least historical fault descriptions and corresponding historical system health snapshots; when the matching degree exceeds the preset similarity threshold, output the historical diagnostic conclusions and solutions recorded in the successfully matched experience template.

[0097] Step S5: Perform feature vectorization matching between the current diagnostic information and historical experience, output the historical experience with the highest matching degree and its solution, and receive verification feedback from operation and maintenance personnel to form a closed-loop learning.

[0098] Figure 9 This is a flowchart illustrating the feature vectorization intelligent matching process of the storage operation and maintenance intelligent diagnosis method based on experience learning and hierarchical synchronization according to exemplary embodiment 1 of the present invention. Figure 9 As shown, in this embodiment, the current diagnostic information and historical experience are matched using feature vectorization in the following way: the diagnostic information of the new fault and the experience template in the historical experience library are respectively converted into multi-dimensional feature vectors. By calculating the weighted similarity between the new fault vector and the historical experience vector, the historical experience with the highest similarity is selected as the recommendation basis.

[0099] Figure 10 This is a flowchart illustrating the verification feedback and closed-loop learning process of the storage operation and maintenance intelligent diagnostic method based on experience learning and hierarchical synchronization according to Exemplary Embodiment 1 of the present invention. Figure 10 As shown in this embodiment, the specific implementation method for receiving verification feedback from maintenance personnel to form a closed-loop learning loop is as follows:

[0100] The operations and maintenance personnel confirm the effectiveness of the recommended solution, or revise the recommendation results if necessary, and record the specific content of the feedback; based on the feedback results of the operations and maintenance personnel, dynamically update the confidence score and historical reuse count of the experience data package; encapsulate the complete process of this diagnosis into a new standardized experience data package, which includes the full health status snapshot generated by this diagnosis, the final diagnosis conclusion and the adopted solution, and transmit the new standardized experience data package in a layered synchronous manner.

[0101] Example 2

[0102] Exemplary embodiment 2 of the present invention provides a storage operation and maintenance intelligent diagnosis method based on experience learning and hierarchical synchronization. In this embodiment, each operation and maintenance diagnosis activity is standardized into an encapsulated, transmittable and learnable "experience data package", and a continuously evolving intelligent diagnosis ecosystem is constructed through a hierarchical synchronization strategy.

[0103] This embodiment provides a detailed description of the method of the present invention in a specific scenario.

[0104] In this scenario, the method of this embodiment is implemented in a system architecture that includes an operation and maintenance terminal, a synchronization hub, and a knowledge platform center.

[0105] The operation and maintenance terminal is a diagnostic tool deployed in various environments (laboratory / production). Its core is the "command encapsulation and capture module", which intercepts all diagnostic commands, automatically associates them with context, and generates structured experience drafts.

[0106] The synchronization hub is divided into a "direct-connect synchronization agent" (for laboratory use) and an "offline exporter" (for production environments). The latter can generate encrypted packages containing incremental / difference experiences.

[0107] The core of the knowledge platform center is the "experience processing engine," which is responsible for deduplication of experiences, quality assessment, intelligent merging of conflicts, and training of diagnostic models.

[0108] In this scenario, when a fault occurs, the monitoring system will display alarms indicating "high IO latency" and "scheduling lag" in the storage cluster.

[0109] Operations and maintenance personnel trigger diagnostics through the platform. The system automatically enters root cause-driven mode and concurrently executes the "root cause checkset," including: df -h (disk space), iostat -x (IO performance), vmstat (memory and CPU), systemctl status (service status), etc.

[0110] The test results showed that the root cause was an abnormality in the "System Disk Space" check ( / partition usage reached 98%), while all other checks, including data disk, network, and RAID, were normal. The system generated a healthy snapshot containing this abnormal state.

[0111] The system converts this snapshot into a feature vector and matches it against the knowledge base. A historical experience template was successfully matched, with a snapshot feature similarity of 92%. This template records a past failure caused by "the system disk log being full, leading to IO queue congestion and subsequently global scheduling lag."

[0112] The system recommends the following root cause and solution for this historical experience: "Clean up the / var / log directory, or migrate the logs to the data disk."

[0113] The operations and maintenance personnel quickly resolved the issue based on the solution and confirmed the resolution on the platform. The entire diagnostic process (from checking snapshots to the solution) was automatically packaged into a new experience data package.

[0114] This new experience is synchronized to the knowledge platform center through a hierarchical synchronization mechanism (direct connection from the lab in this example). The central engine determines that it is a different manifestation of the same root cause as the historical experience, and therefore merges the two experiences to increase the confidence of the root cause "system disk full" and the richness of the solutions.

[0115] The storage operation and maintenance intelligent diagnosis method based on experience learning and hierarchical synchronization in this invention has the following beneficial technical effects:

[0116] I. Realize the assetization and self-evolution mechanism of operation and maintenance knowledge.

[0117] By defining standardized "experience data templates," unstructured, experience-dependent operational data is successfully transformed into machine-readable, computable, and evaluable structured digital assets. This enables the persistent accumulation, precise management, and intelligent application of operational knowledge. Through subsequent intelligent matching, verification feedback, and deduplication processes, the diagnostic knowledge base can continuously learn and autonomously evolve, overcoming the limitations of traditional static knowledge bases.

[0118] II. Enhance the certainty and reliability of the diagnostic process.

[0119] Unlike traditional methods that infer from ambiguous phenomena, this invention directly obtains a snapshot of a deterministic health status by actively and concurrently executing a "root cause check set" covering the entire component layer. This avoids subjective misjudgments at the source and, combined with high-precision historical template matching based on feature vectors, provides a solid objective data foundation for diagnostic conclusions, significantly improving accuracy.

[0120] Third, by adopting a "layered synchronization" strategy that adapts to complex network environments, barriers to knowledge flow can be broken down.

[0121] Under the premise of strictly adhering to the highest level of network security regulations, we can achieve secure, controllable, efficient aggregation and real-time distribution of operational experience across the entire domain, breaking down traditional data silos.

[0122] IV. Significantly improves operational efficiency and response speed.

[0123] By accurately matching historical experience templates, proven solutions can be directly provided for common faults, significantly reducing the average repair time. At the same time, predefined root cause check sets can be executed automatically and periodically, transforming operations and maintenance from a passive "fault response" to a proactive "health prevention," preventing problems before they occur.

[0124] Fifth, standardize and equalize team capabilities.

[0125] By solidifying expert diagnostic experience into reusable knowledge templates, junior engineers can quickly resolve complex problems with system guidance, thereby leveling up the team's overall technical capabilities and reducing over-reliance on specific individuals. Furthermore, all diagnostic operations are logged on the platform, ensuring full traceability of the maintenance process, auditability of operations, and reviewability of decisions, significantly improving the standardization and security of the work.

[0126] VI. It has excellent fault location capabilities.

[0127] By comparing full health snapshots, cases with the same root cause pattern can be quickly matched from historical experience, directly revealing their hidden root causes and providing effective solutions that have been tested in practice, greatly reducing the difficulty of troubleshooting difficult problems.

[0128] VII. It can lead to innovative product and service models.

[0129] For example, a "smart operations and maintenance knowledge platform" or a "diagnosis-as-a-service" SaaS product can be built, providing unprecedented value-added services to the market. For hardware or software vendors, integrating this system as a core function can create a strong technological barrier and product differentiation, allowing them to escape homogeneous competition. For end users, it can effectively reduce their total cost of ownership and business risks. By improving the level and efficiency of operations and maintenance automation, customers can reduce their reliance on expensive, highly skilled experts, directly reducing labor costs. More importantly, through preventative maintenance and rapid fault repair, it can effectively ensure the continuity and stability of customers' core businesses, thereby improving customer satisfaction and loyalty, and bringing long-term sustainable business returns to service providers.

[0130] Through the above description of the embodiments, those skilled in the art can clearly understand that each embodiment can be implemented by means of software plus necessary general-purpose hardware platforms, and of course, it can also be implemented by hardware. Based on this understanding, the above technical solutions, in essence or the part that contributes to the prior art, can be embodied in the form of a software product. This computer software product can be stored in a computer-readable storage medium, such as ROM / RAM, magnetic disk, optical disk, etc., including several instructions to cause a computer device (which may be a personal computer, server, or network device, etc.) to execute the methods of various embodiments or some parts of embodiments.

[0131] The above description is merely a specific embodiment of the present invention, but the scope of protection of the present invention is not limited thereto. Any variations or substitutions that can be easily conceived by those skilled in the art within the technical scope disclosed in the present invention should be included within the scope of protection of the present invention. Therefore, the scope of protection of the present invention should be determined by the scope of the claims.

Claims

1. A storage operation and maintenance intelligent diagnostic method based on experience learning and hierarchical synchronization, characterized in that, The method includes: Step S1: Intercept diagnostic commands through the operation and maintenance terminal and associate them with system context information to generate a standardized experience data package; Step S2: Based on the network connectivity of the environment where the maintenance terminal is located, select either online direct connection or offline export method to transfer the standardized experience data package to the knowledge platform center; Step S3: The received standardized experience data packages are deduplicated, evaluated, and merged through the knowledge platform center to update the global knowledge base; Step S4: Upon receiving a diagnostic request, select either experience-driven mode or root cause-driven mode for diagnosis. Root cause-driven mode generates a full health status snapshot by concurrently executing a predefined root cause check set. Step S5: Perform feature vectorization matching between the current diagnostic information and historical experience, output the historical experience with the highest matching degree and its solution, and receive verification feedback from operation and maintenance personnel to form a closed-loop learning.

2. The intelligent storage operation and maintenance diagnosis method based on experience learning and hierarchical synchronization according to claim 1, characterized in that, In step S1, the specific method for generating the standardized experience data package is as follows: Intercept diagnostic commands initiated by the operation and maintenance terminal, and collect system context information when the command is executed. The system context information includes user identity, the target being diagnosed, system timestamp, and system status snapshot. The diagnostic commands, the output results after command execution, and system context information are encapsulated and bound to form a preliminary experience record; Receive and record the root causes and solutions summarized by operations and maintenance personnel during the diagnostic process, and integrate them into a complete standardized experience data package.

3. The intelligent storage operation and maintenance diagnosis method based on experience learning and hierarchical synchronization according to claim 1, characterized in that, In step S1, the standardized experience data package is stored using a structured operation and maintenance experience template, which includes at least the following interrelated data groups: Problem feature groups are used to record descriptions of fault symptoms and system health snapshots; The diagnostic process group records the sequence of examination commands actually executed during the diagnostic process and the root cause that is ultimately determined. Solution groups are used to record the specific operational steps for resolving faults and the associated automated script identifiers; The validation feedback group is used to record the historical validity feedback results and reuse statistics of the experience data package.

4. The intelligent storage operation and maintenance diagnosis method based on experience learning and hierarchical synchronization according to claim 1, characterized in that, In step S2, the specific implementation method for transmitting the experience data package using the offline export method is as follows: The synchronization module deployed in the production environment periodically or when preset trigger conditions are met compares the local experience base with the last successfully exported baseline version to identify newly added, modified, or changed experience data packages. The identified incremental experience data is packaged and an integrity check code is generated to form an offline transmission package that can be transmitted through physical media; At the same time, a list file describing the details of this export is generated to enable offline reporting of operational experience in a completely isolated production environment.

5. The intelligent storage operation and maintenance diagnosis method based on experience learning and hierarchical synchronization according to claim 1, characterized in that, In step S3, the received standardized experience data packets are deduplicated, evaluated, and merged in the following manner: The knowledge platform center determines the logical relationship between newly received standard experience data packages and existing standard experience data packages based on their source identifiers, timestamp information, confidence scores, and feature vector similarity. If the report is determined to be a duplicate of the same fault event, the number of verifications of the original experience data package is increased and its confidence level is improved. If they are determined to be different manifestations of the same root cause or supplements to the solutions, merge the relevant information to enrich the content of the original experience data package. If a diagnostic conclusion is determined to be substantially conflicting, the experience data package is marked as a conflict status pending manual review, and the evidence from both conflicting parties is stored together.

6. The intelligent storage operation and maintenance diagnosis method based on experience learning and hierarchical synchronization according to claim 1, characterized in that, In step S3, after completing the global knowledge base update, the knowledge platform center actively distributes the updated experience data package to each operation and maintenance terminal through the synchronization hub to achieve real-time sharing and collaborative evolution of operation and maintenance experience across the entire domain.

7. The intelligent storage operation and maintenance diagnosis method based on experience learning and hierarchical synchronization according to claim 1, characterized in that, In step S4, the specific implementation method for selecting the root cause-driven mode for diagnosis is as follows: A standardized set of check items covering the key fault domains of the storage system is pre-built. These key fault domains include at least disk status, RAID array, network connectivity, memory usage, CPU load, and service process running status. When a diagnosis is triggered, all check commands in the standardized check item set are executed concurrently or in parallel. Collect the structured output results returned after each check command is executed, and summarize them to generate a unified, machine-readable full health status snapshot.

8. The intelligent storage operation and maintenance diagnosis method based on experience learning and hierarchical synchronization according to claim 1, characterized in that, In step S4, the specific implementation method for selecting the experience-driven mode for diagnosis is as follows: Receive fault description information manually entered by maintenance personnel or automatically reported by the monitoring system; The fault phenomenon description information is matched with the problem feature layer of the experience templates stored in the knowledge base. The problem feature layer contains at least historical fault phenomenon descriptions and corresponding historical system health snapshots. When the matching degree exceeds the preset similarity threshold, the historical diagnostic conclusions and solutions recorded in the successfully matched experience template are output.

9. The intelligent storage operation and maintenance diagnosis method based on experience learning and hierarchical synchronization according to claim 1, characterized in that, In step S5, the current diagnostic information and historical experience are matched using feature vectorization in the following way: the diagnostic information of the new fault and the experience template in the historical experience library are converted into multi-dimensional feature vectors respectively. By calculating the weighted similarity between the new fault vector and the historical experience vector, the historical experience with the highest similarity is selected as the recommendation basis.

10. The intelligent storage operation and maintenance diagnosis method based on experience learning and hierarchical synchronization according to claim 1, characterized in that, In step S5, the specific implementation method for receiving verification feedback from operations and maintenance personnel to form a closed-loop learning process is as follows: The operations and maintenance personnel confirm the effectiveness of the recommended solutions, or revise the recommendations when necessary, and record the specific content of the feedback. Based on feedback from operations and maintenance personnel, the confidence score and historical reuse count of this experience data package are dynamically updated. The entire diagnostic process is encapsulated into a new standardized experience data package, which includes a full health status snapshot generated during the diagnosis, the final diagnostic conclusions, and the solutions adopted. This new standardized experience data package is then transmitted in a layered and synchronized manner.

Images