Computing power server operation and maintenance auditing method and system based on block chain and Internet of Things

Abstract

The invention provides a computing power server operation and maintenance auditing method and system based on a block chain and the Internet of Things. The method comprises the steps that physical and logic operation and maintenance behaviors are collected in a unified mode, a structured event chain is constructed, behavior weights are calculated in combination with time, identity and semantics, server state difference analysis is conducted with events as anchor points, and attribution credibility is quantified. And finally, a structured auditing record containing identity, time, state abstract and credibility score is generated, and the structured auditing record is sequentially chained and stored after digital signature, so that a responsibility evidence chain which cannot be tampered and can be verified is formed, and the problems that behaviors and results are disjointed and the evidence chain is incomplete in operation and maintenance auditing are effectively solved. According to the method, cross-layer association of physical and logic behaviors, accurate attribution of operation and maintenance results and long-term credible evidence storage of auditing evidences are realized, and the method is suitable for computing power infrastructure scenes with high safety and compliance requirements.

Description

Technical Field

[0001] This invention belongs to the field of data processing, and in particular relates to a method and system for auditing the operation and maintenance of computing servers based on blockchain and the Internet of Things. Background Technology

[0002] With the continuous expansion of computing infrastructure in cloud computing, artificial intelligence, and high-performance computing scenarios, large-scale computing servers in data centers and edge nodes require frequent hardware and software maintenance activities, including firmware upgrades, driver and kernel updates, network and security policy adjustments, hardware replacements, out-of-band management, and on-site troubleshooting. These activities often span both physical access and remote operation, and involve highly sensitive permissions and critical configurations. Any missing records, broken chains of responsibility, or tampered evidence will directly affect compliance audits and subsequent accountability. Current practices typically use IoT devices to record data center entry and exit, personnel identification, and local environmental events, supplemented by server-side login and command logs. Some summaries are then stored on the blockchain to enhance immutability. However, this approach has revealed three shortcomings in practical implementation: First, physical behaviors and logical operations lack a structured time-related mechanism, making it difficult to determine whether the person entering at the same time is the actual person responsible for performing remote or local operations, resulting in a lack of identity and behavior verification. Second, logs focus on process recording rather than result description, making it difficult to accurately express "which specific states of the system changed due to a certain operation" at the operational level, which makes it more prone to misjudgment under natural fluctuations and batch task interference. Third, the on-chain data is mostly scattered log fragments or simple hashes, lacking a unified structure and stable summary rules for auditing. There is no computable mapping relationship between on-chain immutability and off-chain verifiability, making it difficult to support automated verification, sequence replay, and accountability review. As regulators place higher demands on the verifiability, interpretability, and long-term traceability of critical infrastructure, simply overlaying multi-source logs and general evidence storage are no longer sufficient to meet the engineering implementation requirements of the complete chain of "operator - operation event - system impact - verifiable evidence". Summary of the Invention

[0003] The purpose of this invention is to design a computing server operation and maintenance auditing method and system based on blockchain and the Internet of Things, which can take into account cross-layer correlation, result attribution, summary stability and on-chain verification, and overcome the problems of physical and logical disconnection, unverifiable results and fragile on-chain structure in existing solutions.

[0004] To achieve the above objectives, a method for auditing the operation and maintenance of computing servers based on blockchain and the Internet of Things is provided in the first aspect of the present invention, the method comprising: Physical behavior data and logical operation data generated during the operation and maintenance of computing power servers are collected. The physical behavior data comes from IoT sensing devices, and the logical operation data comes from the server log interface. The physical behavior data and logical operation data are uniformly abstracted into structured operation and maintenance behavior records, and each record is coded in a regular way to generate a behavior feature vector. All behavior records are sorted in ascending order based on the behavior timestamps to construct a time-ordered chain of behavior events. Behavior consistency weights are calculated based on the time interval between adjacent events, identity consistency, and semantic relevance of the behavior. The semantic relevance of the behavior is determined based on the matching degree of behavior content keywords or operation type. At the same time, the behavior consistency weights are attenuated and adjusted according to the number of events within the time window to suppress the interference of short-term high-frequency operations on audit judgment. Using each event as a time anchor, a snapshot of the server's state before and after is obtained within a symmetrical time window. The snapshot is compared with a limited set of key state indicators that are strongly related to computing power operation and maintenance. A set of state differences is generated, and the credibility of the differential attribution is calculated by combining the behavior consistency weight, the importance of state items, and the magnitude of the difference. Based on the differential attribution credibility, state differential set, behavioral feature vector, operator identity, and behavioral time, a structured audit unit is constructed. After digitally signing the structured audit unit, it is written into the blockchain in chronological order to form an immutable and verifiable chain of evidence of responsibility.

[0005] Furthermore, the physical behavior data includes entry and exit records provided by the access control device, facial recognition identity information output by the camera recognition module, and pass records provided by the fingerprint verification module; the logical operation data includes account login and operation command logs in the server's local logs, as well as key system management events output by the remote management controller.

[0006] Furthermore, the structured operation and maintenance behavior record is a five-tuple, including personnel identification, behavior timestamp, behavior type, source device identification, and behavior content.

[0007] Furthermore, the behavior type is limited to physical or logical, where physical corresponds to the behavior collected by IoT sensing devices, and logical corresponds to the operation behavior collected by the server-side log interface.

[0008] Furthermore, the rule-based encoding includes: mapping personnel identification to a number, normalizing behavior time to a relative time based on the daily start time, converting behavior type to 0 / 1 encoding, converting source device identification to a fixed integer encoding, and extracting keywords from behavior content and mapping them to a number.

[0009] Furthermore, the key status indicators include at least one of the following: GPU utilization and power consumption, critical service liveness status, main configuration file summary, controlled account set, sensitive directory modification summary, management port open / closed status, and out-of-band controller critical event flags.

[0010] Furthermore, when generating the state difference set, direct numerical comparison is used for numerical state items, and rule comparison or digest hash comparison is used for Boolean, enumerated, or structured state items.

[0011] Furthermore, the structured audit unit includes operator identity, behavior time, attribution credibility, state differential ordered digest, behavior semantic compressed digest, and digital signature, wherein the state differential ordered digest is generated by sorting the state item names, performing fixed-length normalization encoding, and hashing.

[0012] Furthermore, the digital signature adopts the elliptic curve signature algorithm SECP256K1 or the national cryptographic algorithm SM2, the signature object is the remaining fields in the structured audit unit except for the signature field, and the signature result length is fixed; the blockchain writing is executed through a smart contract, and the smart contract performs field integrity verification, time increment verification, and signature verification operations on the written data.

[0013] A second aspect of the present invention provides a computing server operation and maintenance audit system based on blockchain and the Internet of Things, the system comprising: The data acquisition module is used to collect physical behavior data and logical operation data generated during the operation and maintenance of the computing server. The physical behavior data comes from IoT sensing devices, and the logical operation data comes from the server log interface. The physical behavior data and logical operation data are uniformly abstracted into structured operation and maintenance behavior records, and each record is coded in a regular way to generate a behavior feature vector. The event chain construction module is used to sort all behavior records in ascending order based on the behavior timestamp, construct a time-ordered behavior event chain, and calculate the behavior consistency weight according to the time interval between adjacent events, identity consistency, and behavior semantic relevance. The behavior semantic relevance is determined based on the matching degree of behavior content keywords or operation type. At the same time, the behavior consistency weight is attenuated and adjusted according to the number of events within the time window to suppress the interference of short-term high-frequency operations on audit judgment. The differential attribution analysis module is used to take each event as a time anchor point, obtain a snapshot of the server's state before and after within a symmetrical time window, compare it with a limited number of key state indicators that are strongly related to computing power operation and maintenance, generate a state difference set, and calculate the differential attribution credibility by combining the behavior consistency weight, the importance of state items and the difference magnitude. The evidence chain generation module is used to construct a structured audit unit based on the differential attribution credibility, state differential set, behavioral feature vector, operator identity and behavior time. After digitally signing the structured audit unit, it is written into the blockchain in chronological order to form an immutable and verifiable chain of responsibility evidence.

[0014] The beneficial technical effects of the present invention are at least as follows: To address the aforementioned issues, this invention provides a method and system for auditing the operation and maintenance of computing power servers based on blockchain and the Internet of Things. First, it uses standardized quintuples to abstract and integrate the minimum necessary fields of physical access and logical operations, constructing behavior records and codes that can be directly used for time-series organization and subsequent calculations. Second, it generates an event chain on the timeline, introducing a weighted mechanism that simultaneously considers time intervals, identity consistency, and behavioral semantic distance, and sets density regularization constraints for short-window dense behaviors to reduce the interference of collaboration and masquerading on audit judgments. Third, it uses events as time anchors to obtain snapshots of the server's state before and after within a symmetrical window, forming a differential set for a limited number of key indicators strongly related to computing power operation and maintenance. It improves attribution credibility by combining event weighting, state importance, and difference magnitude quantification methods, achieving verifiable result expression without introducing a black-box model. Finally, it uses stable ordered encoding and fixed-length concatenation to generate a differential digest, which, along with the operator, event time, credibility score, and behavior digest, generates a structured audit unit. After local signing, it is sequentially written into the blockchain, achieving a low-redundancy, strong-consistency, verifiable, and replayable chain of evidence of responsibility. The aforementioned technical elements are coupled together to form an engineering closed loop from data collection and unified representation, cross-layer correlation and risk suppression, event anchoring and result attribution to on-chain trusted evidence storage. This overcomes the problems of physical and logical disconnect, unverifiable results and fragile on-chain structure in existing solutions, and provides an implementable, verifiable and long-term maintainable operation and maintenance audit infrastructure for high security and strong compliance scenarios. Attached Figure Description

[0015] The present invention will be further described with reference to the accompanying drawings, but the embodiments in the drawings do not constitute any limitation on the present invention. For those skilled in the art, other drawings can be obtained based on the following drawings without creative effort.

[0016] Figure 1 This is a flowchart of the operation and maintenance auditing method for computing power servers based on blockchain and the Internet of Things according to the present invention.

[0017] Figure 2 This is a framework diagram of the computing server operation and maintenance audit system based on blockchain and IoT of the present invention. Detailed Implementation

[0018] Embodiments of the present invention are described in detail below. Examples of these embodiments are shown in the accompanying drawings, wherein the same or similar reference numerals denote the same or similar elements or elements having the same or similar functions throughout. The embodiments described below with reference to the accompanying drawings are exemplary and are only used to explain the present invention, and should not be construed as limiting the present invention.

[0019] In one or more embodiments, such as Figure 1 As shown, a method for auditing the operation and maintenance of computing servers based on blockchain and the Internet of Things is disclosed. The method includes the following: S1: Collect physical behavior data and logical operation data generated during the operation and maintenance of the computing server. The physical behavior data comes from IoT sensing devices, and the logical operation data comes from the server log interface. The physical behavior data and logical operation data are uniformly abstracted into structured operation and maintenance behavior records, and each record is coded in a regular way to generate a behavior feature vector. Specifically, the goal of this step is to collect and standardize the physical behavior and logical operation records generated by the computing power server during actual operation and maintenance, and to construct a standardized and structured behavioral information representation structure to support subsequent behavior chain construction and state attribution. Considering that physical behavior originates from IoT sensing devices (such as access control, cameras, and fingerprint readers), while logical operations originate from server log interfaces (such as login information, instruction execution logs, and API call records), this step achieves a standardized data design with high semantic expressiveness, modelability, and traceability by filtering these data sources, standardizing field abstraction, and encapsulating the structure.

[0020] This step directly collects the following five types of data: entry and exit records provided by the access control device, facial recognition identity information output by the camera recognition module, access records provided by the fingerprint verification module, account login and operation command logs in the server's local logs, and key system management events output by the remote management controller. All this data is acquired passively, without interfering with the device system.

[0021] First, all collected behavioral data are uniformly abstracted into a quintuple form, represented as the first... Structured record of each behavior: ; in, This indicates that the identification of the person involved in the action is provided by fingerprint, access control, or operating account; The local timestamp of the event is generated directly by the corresponding device; Indicates the behavior type, with its value limited to "physical" or "logical"; An identifier that indicates the device or interface from which the behavior originated, such as the access controller number or log interface path; This represents the content of the action, such as the specific command executed, access control action, or operation parameters. The above five-tuple record structure supports the complete expression of the subject, time, source, and semantic features of an action.

[0022] Based on a unified structure, to facilitate subsequent behavioral chain modeling and sequence construction, each behavioral record is encoded to generate a feature vector. : ; in, It is a function that maps identity identifiers to numbers, and it comes from the permission database; To normalize the results of the behavior time, the start time of each day is used as the reference benchmark; Convert the behavior type to 0 / 1 encoding (physical 0, logical 1); It is a fixed integer code representing the source device number; It is the result of keyword extraction and number mapping of behavioral content. This encoding process is entirely based on rule construction, without relying on model reasoning or deep feature extraction, ensuring interpretability, reproducibility, and computational friendliness.

[0023] S2: Sort all behavior records in ascending order based on the behavior timestamp to construct a time-ordered chain of behavior events, and calculate the behavior consistency weight based on the time interval between adjacent events, identity consistency, and semantic relevance of the behavior. The semantic relevance of the behavior is determined based on the matching degree of behavior content keywords or operation type. At the same time, the behavior consistency weight is attenuated and adjusted according to the number of events within the time window to suppress the interference of short-term high-frequency operations on audit judgment. Specifically, the goal of this step is to establish a clearly structured, time-ordered, and identity-traceable chain of behavioral events based on the standardized operational behavior records built in the first step. This chain maps the inherent logical relationships between behaviors through three dimensions: time, identity, and behavioral semantics, providing an accurate and traceable event index for subsequent system state differential analysis and responsibility attribution. This step not only addresses the temporal integration of physical and logical behaviors but also introduces a behavior consistency weight structure specifically designed for auditing computing servers, solving the problem in traditional auditing systems where "though behavior records are complete, it's impossible to determine whether they were performed by the same person."

[0024] First, all behavior records are analyzed based on their timestamps. Sort the data in ascending order to generate a time-consistent sequence of actions. Each sorted record is assigned a unique event number. This establishes a set of event chains: ; in, For the sorted number The identity of the person who operated the event. The trigger time for this behavior. It encodes a vector for its structure. This chain not only ensures traceability in time, but also preserves the identity information and semantic content of the original behavior, providing a key reference for system state modeling and authentication.

[0025] To enhance the semantic coupling between events in the behavior chain, a behavior consistency weight function based on multi-dimensional factors is introduced to express the probability that adjacent behavior events belong to the same operation and maintenance process. The weight function is designed as follows: ; in, Indicates the time interval between two adjacent events; This is an identity matching penalty item. If two events are executed by different people, this item will have a value of [value missing]. Otherwise ; This is the behavioral semantic distance, calculated as the cosine distance between two vectors. Weight parameters. , , Adjustments can be made based on engineering experience, generally meeting the requirements. This ensures that mismatched identities have the greatest impact on weighting.

[0026] In auditing computing servers, "inconsistent identity" is a high-risk factor, "large time intervals" indicate that they may belong to different tasks, and "significant differences in behavioral semantics" indicate a significant change in operation type. This multi-factor weighting mechanism differs from common time-series modeling methods and is more in line with the dual requirements of "behavioral continuity" and "identity consistency" in audit scenarios.

[0027] To further enhance the auditing capabilities of behavioral chains, this step also defines an event weight adjustment mechanism with constraints. A special regularization term is introduced into the weight calculation to limit the event density of operations performed by the same identity within a short period, preventing a single identity from masking collaborative operations through intensive activity. ; in, Indicates within the time window Inside, by identity The number of events generated This is a coefficient used to control the strength of the density penalty term. The regularization term acts as... The multiplier, the final action event weight is updated to This is used as a reference weight in the subsequent state attribution stage. Its function is to weaken the dominance of short-term, high-frequency operations in the behavior chain, prevent the audit system from mistakenly judging short-term concentrated operations as legitimate continuous behavior, and thus improve the accuracy of abnormal behavior identification.

[0028] S3: Using each event as a time anchor point, obtain a snapshot of the server's state before and after within a symmetrical time window, compare it with a limited set of key state indicators that are strongly related to computing power operation and maintenance, generate a set of state differences, and calculate the credibility of the differential attribution by combining the behavior consistency weight, the importance of state items and the magnitude of difference. Specifically, this step, based on the established event chain, establishes a one-to-one mapping of "event-state change" for each event around the key operational items of the computing server, and provides a differential trust quantification for auditing. The input is the two outputs of the previous step: one is the set of event chains. Each event , Indicates the operator's identity. Indicates the time of the event. The first is the behavior encoding vector retained from the previous step; the second is the set of weights between events. Each item is calculated in the previous step according to time interval, identity consistency and semantic similarity and then corrected by event density regularization.

[0029] by As a time anchor point, in Internally, status snapshots are captured through the server-side data collection interface. And select a set of status items from the fixed monitoring list. Compare them. It includes only a limited set of metrics strongly relevant to computing power operation and maintenance, such as: GPU utilization and power consumption, critical service liveness status, main configuration file summary, controlled account set, sensitive directory modification summary, management port open / closed status, out-of-band controller critical event markers, etc. All metrics are directly provided by local server probes or out-of-band management interfaces, and are compared according to field rules rather than through secondary modeling. Events are defined based on snapshot pairs. The state difference set is as follows: ; In the formula, For the status item name; and They are respectively exist and The values ​​in the table are determined by the following criteria: Numerical items are compared directly using numerical comparison, while Boolean / enumeration / structured items are compared using rule-based comparison or digest comparison (e.g., whether the hash has changed or the enumeration value has switched). To avoid calculation biases caused by different units of measurement, the values ​​of the participating metrics are normalized according to the rules within the item before the difference comparison, and the same mapping scheme is maintained for subsequent calculations.

[0030] Considering the natural fluctuations and batch operations in computing power scenarios, this step incorporates the importance and stability of the event weights and state items provided in the previous step into the trusted metrics, defining events. The confidence level of the differential attribution is: ; in, The weights, adjusted for the timing and density of adjacent events, are given by the calculation in the previous step and are directly referenced here. For state items The overall weighting coefficient is determined in two parts during the deployment phase: first, structural importance (such as "administrator account change" and "main configuration file modification" have higher weights), and second, volatility confidence factor (based on the inverse quantification of the frequency of the same type of volatility within a recent event-free window). For measuring discrepancies, numerical types are normalized to the absolute difference, Boolean / enumerated types are mapped to fixed difference values ​​based on inconsistency criteria, and structured summary types are mapped to normalized differences based on edit / set differences. The above three-factor design corresponds to the three dimensions of "reliable event context, significant state change, and significant change magnitude," which can suppress the interference of natural noise and batch low-risk operations on attribution results; when the same identity completes configuration and permission linkage modifications within a tight timeframe, and The combined amplification can significantly improve Discrimination level.

[0031] Through the above process, this step outputs two types of results. The first is a set of state differences organized by event number. Each Directly by The comparison revealed that all values ​​within a given item could be traced back to the specific data collection interface and comparison rules. The second aspect is the differential attribution credibility set. Each The event weights from the previous step are calculated using the formula in the previous section. With respect to the item rights in this step Difference measurement The two outputs above will not be further transformed and will be directly used for the next step of audit summary construction and on-chain writing; if window parameters are required... When adjusting the settings, follow the principle of "ensuring data collection coverage and avoiding cross-event leakage" by adjusting the settings within the recommended range, and maintain consistent configuration for the same environment to ensure comparability of results.

[0032] S4: Based on the differential attribution credibility, state differential set, behavioral feature vector, operator identity and behavior time, construct a structured audit unit, digitally sign the structured audit unit, and write it into the blockchain in chronological order to form an immutable and verifiable chain of responsibility evidence.

[0033] Specifically, this step, based on the state difference set and attribution credibility set constructed in the previous stage, further organizes them into structured audit records and achieves immutable on-chain evidence storage. The goal is to construct a chain of responsibility containing fields such as operator identity, behavior time, state impact summary, and credibility score, and to write this chain to the blockchain to create a long-term verifiable archive of operational actions. Unlike traditional log compression schemes, the audit chain designed in this step is centered on a "behavior-consequence" structure, directly mapped to an on-chain data structure, and embeds minimum digest, digital signature, and chained timestamp mechanisms to ensure its integrity and traceability.

[0034] To generate structured audit units that can be used for evidence storage in blockchain systems, each event is defined. The audit log structure is as follows: ; in and events The operator's identity and the timing of the action both come from the event chain structure defined in step two. The attribution credibility score for step three; Representing the difference set An ordered structure summary is generated using a hash construct defined below; For behavior vectors The compressed summary retains the ability to recognize behavioral semantics without revealing the specific content; Yes The result of irreversible digital signature of the first five fields is generated by the system's private key and is used to ensure the anti-counterfeiting capability of on-chain records.

[0035] in, The generation method is as follows, ensuring strict consistency with the variables mentioned above to avoid homonyms: From Extract all state items Sort by status item name and their corresponding preceding and following values. After performing fixed-length normalized encoding and concatenation, the resulting data is input into a hash function to calculate the digest value. ; in The normalized encoding function, whose rules have been defined and used previously. The generation method should remain completely consistent. Here... For splicing symbols, The standard hash function used for on-chain deployment, such as SHA-256. This digest structure is compressible. The full state differential content is written into the chain in a low-data-volume form while maintaining verifiability.

[0036] Audit Unit Once constructed, they are organized into an audit chain set in chronological order: ; Write data sequentially through the blockchain smart contract interface. The contract interface must support field integrity verification, time increment verification, signature verification, and on-chain state updates. This step does not alter the original meaning of the event data; it only creates a structural summary of the state differences to ensure minimal on-chain data storage while maintaining high verifiability.

[0037] Furthermore, to ensure the reproducibility of signature operations, the SECP256K1 curve from the elliptic curve signature algorithm is used as the default signature scheme, and the system uses the private key of the chain-end registered node to generate the signature by default. The signature result only applies to the digest field, and the signature length is a fixed 64 bytes, meeting the on-chain storage compression requirements. If the deployment environment has additional requirements for the signature scheme, it can be replaced with a similar national cryptographic algorithm such as SM2, while keeping the field interfaces unchanged.

[0038] In one or more embodiments, such as Figure 2 As shown, a computing server operation and maintenance audit system based on blockchain and the Internet of Things is disclosed. The system includes: The data acquisition module is used to collect physical behavior data and logical operation data generated during the operation and maintenance of the computing server. The physical behavior data comes from IoT sensing devices, and the logical operation data comes from the server log interface. The physical behavior data and logical operation data are uniformly abstracted into structured operation and maintenance behavior records, and each record is coded in a regular way to generate a behavior feature vector. The event chain construction module is used to sort all behavior records in ascending order based on the behavior timestamp, construct a time-ordered behavior event chain, and calculate the behavior consistency weight according to the time interval between adjacent events, identity consistency, and behavior semantic relevance. The behavior semantic relevance is determined based on the matching degree of behavior content keywords or operation type. At the same time, the behavior consistency weight is attenuated and adjusted according to the number of events within the time window to suppress the interference of short-term high-frequency operations on audit judgment. The differential attribution analysis module is used to take each event as a time anchor point, obtain a snapshot of the server's state before and after within a symmetrical time window, compare it with a limited number of key state indicators that are strongly related to computing power operation and maintenance, generate a state difference set, and calculate the differential attribution credibility by combining the behavior consistency weight, the importance of state items and the difference magnitude. The evidence chain generation module is used to construct a structured audit unit based on the differential attribution credibility, state differential set, behavioral feature vector, operator identity and behavior time. After digitally signing the structured audit unit, it is written into the blockchain in chronological order to form an immutable and verifiable chain of responsibility evidence.

[0039] It is worth noting that the specific workflow of the blockchain and IoT-based computing server operation and maintenance audit system provided in this embodiment of the invention is the same as that of the blockchain and IoT-based computing server operation and maintenance audit method described in the above embodiments, and will not be repeated here.

[0040] This invention also provides a computing server operation and maintenance auditing device based on blockchain and the Internet of Things, including a processor, a memory, and a computer program stored in the memory and configured to be executed by the processor. When the processor executes the computer program, it implements the steps described in the above embodiments of the computing server operation and maintenance auditing method based on blockchain and the Internet of Things. Figure 1 The steps S1 to S4 described above; or, when the processor executes the computer program, it implements the functions of each module in the above system embodiments.

[0041] For example, the computer program may be divided into one or more modules, which are stored in the memory and executed by the processor to complete the present invention. The one or more modules may be a series of computer program instruction segments capable of performing specific functions, which describe the execution process of the computer program in the blockchain and IoT-based computing server operation and maintenance auditing device.

[0042] The blockchain and IoT-based computing server operation and maintenance auditing device can be a desktop computer, laptop, handheld computer, or cloud server, among other computing devices. This device may include, but is not limited to, processors and memory. Those skilled in the art will understand that the blockchain and IoT-based computing server operation and maintenance auditing device may also include input / output devices, network access devices, buses, etc.

[0043] The processor can be a Central Processing Unit (CPU), or other general-purpose processors, digital signal processors (DSPs), application-specific integrated circuits (ASACs), field-programmable gate arrays (FPGAs), or other programmable logic devices, discrete gate or transistor logic devices, discrete hardware components, etc. The general-purpose processor can be a microprocessor or any conventional processor. This processor is the control center of the blockchain and IoT-based computing server operation and maintenance auditing equipment, connecting all parts of the equipment via various interfaces and lines.

[0044] The memory can be used to store the computer programs and / or modules. The processor, by running or executing the computer programs and / or modules stored in the memory and calling the data stored in the memory, realizes various functions of the blockchain and IoT-based computing server operation and maintenance auditing device. The memory may mainly include a program storage area and a data storage area. The program storage area may store the operating system, at least one application required for a function, etc.; the data storage area may store data created based on the operation of the air conditioner controller, etc. In addition, the memory may include high-speed random access memory, and may also include non-volatile memory, such as hard disk, memory, plug-in hard disk, smart media card (SMC), secure digital card (SD), flash card, at least one disk storage device, flash memory device, or other volatile solid-state storage device.

[0045] The modules integrated into the blockchain and IoT-based computing server operation and maintenance auditing equipment, if implemented as software functional units and sold or used as independent products, can be stored in a computer-readable storage medium. Based on this understanding, all or part of the processes in the above embodiments of the present invention can also be implemented by a computer program instructing related hardware. The computer program can be stored in a computer-readable storage medium, and when executed by a processor, it can implement the steps of the various method embodiments described above. The computer program includes computer program code, which can be in the form of source code, object code, executable files, or certain intermediate forms. The computer-readable medium can include: any entity or device capable of carrying the computer program code, recording media, USB flash drives, portable hard drives, magnetic disks, optical disks, computer memory, read-only memory (ROM), random access memory (RAM), electrical carrier signals, telecommunication signals, and software distribution media, etc.

[0046] Those skilled in the art will understand that all or part of the processes in the above embodiments can be implemented by a computer program instructing related hardware. The program can be stored in a computer-readable storage medium, and when executed, it can include the processes of the embodiments of the above methods. The storage medium can be a magnetic disk, optical disk, read-only memory (ROM), or random access memory (RAM), etc.

[0047] The above description represents the preferred embodiments of the present invention. It should be noted that those skilled in the art can make various improvements and modifications without departing from the principles of the present invention, and these improvements and modifications are also considered to be within the scope of protection of the present invention.

Claims

1. A method for auditing the operation and maintenance of computing servers based on blockchain and the Internet of Things, characterized in that, The method includes: Physical behavior data and logical operation data generated during the operation and maintenance of computing power servers are collected. The physical behavior data comes from IoT sensing devices, and the logical operation data comes from the server log interface. The physical behavior data and logical operation data are uniformly abstracted into structured operation and maintenance behavior records, and each record is coded in a regular way to generate a behavior feature vector. All behavior records are sorted in ascending order based on the behavior timestamps to construct a time-ordered chain of behavior events. Behavior consistency weights are calculated based on the time interval between adjacent events, identity consistency, and semantic relevance of the behavior. The semantic relevance of the behavior is determined based on the matching degree of behavior content keywords or operation type. At the same time, the behavior consistency weights are attenuated and adjusted according to the number of events within the time window to suppress the interference of short-term high-frequency operations on audit judgment. Using each event as a time anchor, a snapshot of the server's state before and after is obtained within a symmetrical time window. The snapshot is compared with a limited set of key state indicators that are strongly related to computing power operation and maintenance. A set of state differences is generated, and the credibility of the differential attribution is calculated by combining the behavior consistency weight, the importance of state items, and the magnitude of the difference. Based on the differential attribution credibility, state differential set, behavioral feature vector, operator identity, and behavioral time, a structured audit unit is constructed. After digitally signing the structured audit unit, it is written into the blockchain in chronological order to form an immutable and verifiable chain of evidence of responsibility.

2. The method for operation and maintenance auditing of computing power servers based on blockchain and the Internet of Things as described in claim 1, characterized in that, The physical behavior data includes entry and exit records provided by the access control device, facial recognition identity information output by the camera recognition module, and pass records provided by the fingerprint verification module; the logical operation data includes account login and operation command logs in the server's local logs, as well as key system management events output by the remote management controller.

3. The method for operation and maintenance auditing of computing power servers based on blockchain and the Internet of Things as described in claim 1, characterized in that, The structured operation and maintenance behavior record is a five-tuple, including personnel identification, behavior timestamp, behavior type, source device identification, and behavior content.

4. The method for operation and maintenance auditing of computing power servers based on blockchain and the Internet of Things as described in claim 3, characterized in that, The behavior type is limited to physical or logical, where physical corresponds to the behavior collected by IoT sensing devices and logical corresponds to the operation behavior collected by the server-side log interface.

5. The method for operation and maintenance auditing of computing power servers based on blockchain and the Internet of Things according to claim 3, characterized in that, The rule-based coding includes: mapping personnel identification to a number, normalizing behavior time to a relative time based on the daily start time, converting behavior type to 0 / 1 code, converting source device identification to a fixed integer code, and extracting keywords from behavior content and mapping them to a number.

6. The method for operation and maintenance auditing of computing power servers based on blockchain and the Internet of Things according to claim 1, characterized in that, The key status indicators include at least one of the following: GPU utilization and power consumption, critical service liveness status, master configuration file summary, controlled account set, sensitive directory modification summary, management port open / closed status, and out-of-band controller critical event flags.

7. The method for operation and maintenance auditing of computing power servers based on blockchain and the Internet of Things according to claim 1, characterized in that, When generating the state difference set, direct numerical comparison is used for numerical state items, and rule comparison or digest hash comparison is used for Boolean, enumerated, or structured state items.

8. The method for operation and maintenance auditing of computing power servers based on blockchain and the Internet of Things according to claim 1, characterized in that, The structured audit unit includes operator identity, behavior time, attribution credibility, state differential ordered digest, behavior semantic compressed digest, and digital signature. The state differential ordered digest is generated by sorting the state item names, performing fixed-length normalization encoding, and hashing.

9. The method for operation and maintenance auditing of computing power servers based on blockchain and the Internet of Things according to claim 1, characterized in that, The digital signature uses the elliptic curve signature algorithm SECP256K1 or the national cryptographic algorithm SM2. The signature object is the remaining fields in the structured audit unit except for the signature field, and the signature result has a fixed length. The blockchain writing is executed through a smart contract. The smart contract performs field integrity verification, time increment verification, and signature verification operations on the written data.

10. A computing server operation and maintenance audit system based on blockchain and the Internet of Things, characterized in that, The system includes: The data acquisition module is used to collect physical behavior data and logical operation data generated during the operation and maintenance of the computing server. The physical behavior data comes from IoT sensing devices, and the logical operation data comes from the server log interface. The physical behavior data and logical operation data are uniformly abstracted into structured operation and maintenance behavior records, and each record is coded in a regular way to generate a behavior feature vector. The event chain construction module is used to sort all behavior records in ascending order based on the behavior timestamp, construct a time-ordered behavior event chain, and calculate the behavior consistency weight according to the time interval between adjacent events, identity consistency, and behavior semantic relevance. The behavior semantic relevance is determined based on the matching degree of behavior content keywords or operation type. At the same time, the behavior consistency weight is attenuated and adjusted according to the number of events within the time window to suppress the interference of short-term high-frequency operations on audit judgment. The differential attribution analysis module is used to take each event as a time anchor point, obtain a snapshot of the server's state before and after within a symmetrical time window, compare it with a limited number of key state indicators that are strongly related to computing power operation and maintenance, generate a state difference set, and calculate the differential attribution credibility by combining the behavior consistency weight, the importance of state items and the difference magnitude. The evidence chain generation module is used to construct a structured audit unit based on the differential attribution credibility, state differential set, behavioral feature vector, operator identity and behavior time. After digitally signing the structured audit unit, it is written into the blockchain in chronological order to form an immutable and verifiable chain of responsibility evidence.

Images