A private cloud and data center integrated management platform and method

Abstract

The application provides a private cloud and data center integrated management platform and method, relates to the technical field of data management, and comprises a subject boundary binding module; the subject boundary binding module is arranged on a gateway server and is used for realizing binding verification of an access identity and a data boundary; a sensitivity dynamic matching module is used for calculating a data sensitivity level; a compliance closed loop management module is used for receiving a sensitivity value output by the sensitivity dynamic matching module; a three-dimensional linkage scheduling module is connected with the compliance closed loop management module through a high-speed interconnection channel and receives a risk output by the compliance closed loop management module; through the construction of a collaborative system of four core modules of subject boundary binding, sensitivity dynamic matching, compliance closed loop management and three-dimensional linkage scheduling, the integrated improvement of private cloud and data center safety control and resource efficiency is realized; the subject boundary binding is realized through block chain storage and multi-dimensional verification.

Description

TECHNICAL FIELD

[0001] The present application relates to the technical field of data management, in particular to a private cloud and data center integrated management platform and method. BACKGROUND

[0002] In the existing private cloud and data center management platform, each functional module usually runs independently, such as identity verification, data sensitivity assessment, compliance management and resource scheduling, etc. The links are disconnected with each other, which leads to the disconnection of security policy and efficiency optimization. Specifically, the traditional platform has the following problems:

[0003] Functional isolation: identity verification is only based on simple IP or password verification, lacks dynamic binding with data boundaries, and cannot prevent IP spoofing or replay attacks; sensitivity assessment is often static, without real-time calculation combined with data flow characteristics, leading to inaccurate data protection level; compliance management relies on post-audit, lacks real-time monitoring in the kernel state and data encryption traceability, and is difficult to form a closed loop; resource scheduling is independent of security risks, only considers load status, ignores data sensitivity and compliance risks, and causes unreasonable resource allocation;

[0004] Security and efficiency disconnection: security measures (such as firewall interception) often affect data transmission and scheduling efficiency, and resource optimization may sacrifice security boundaries, such as unverified node access to core data, increasing the risk of leakage. In addition, the mapping relationship is stored in a centralized database, which is easy to tamper with, and the verification dimension is single, with low identification accuracy;

[0005] Technical fragmentation: existing solutions are mostly designed by piecing together, such as TPM verification only supporting single node authorization, which cannot meet the multi-node mutual authentication demand; data encryption and transmission channel are not specialized, leading to high delay and poor real-time performance. SUMMARY

[0006] In view of the deficiencies of the prior art, the present application provides a private cloud and data center integrated management platform and method.

[0007] A private cloud and data center integrated management platform, comprising a subject boundary binding module;

[0008] The subject boundary binding module is deployed on a gateway server and is used for realizing binding verification of access identity and data boundary. The input is an access request initiated by an external or internal party. By verifying whether the IP in the access request and the core data type match the pre-stored corresponding relationship, an interception instruction or a verification pass signal is output. For access requests that do not match, an interception instruction is sent to a hardware firewall. For access requests that match and have valid signatures, a verification pass signal is output.

[0009] The sensitivity dynamic matching module is configured to calculate a data sensitivity level, and an input of the sensitivity dynamic matching module is a verification pass signal output by the subject boundary binding module, a sensitivity value is calculated based on data features associated with the verification pass signal, and the sensitivity value is output;

[0010] The compliance closed-loop management module is configured to receive the sensitivity value output by the sensitivity dynamic matching module, and calculate and output a risk score in combination with a data flow track;

[0011] The three-dimensional linkage scheduling module is connected to the compliance closed-loop management module through a high-speed interconnection channel, receives the risk score output by the compliance closed-loop management module, calculates a resource adaptation degree in combination with resource state data, and outputs an optimal resource scheduling scheme based on the resource adaptation degree.

[0012] It should be further noted that, under the provision of the core architecture framework, subsequent clauses are implemented around this architecture; mainly including identity verification, sensitivity assessment, compliance judgment, and resource scheduling, solving the problem of functional isolation, security and efficiency disconnection of existing platforms; ensuring the integrity of the entire technical solution, and avoiding fragmentation;

[0013] Preferably, the pre-stored correspondence relationship of the subject boundary binding module is a unique mapping of core data types and authenticated physical storage node IPs, and the core data types include personal sensitive information, business core data, and key data.

[0014] The mapping relationship is written into the blockchain after being confirmed by double-node digital signature, and the signature adopts an SM2 asymmetric encryption algorithm; the binding verification includes four dimensions of IP prefix matching, identity public key verification, binding identifier consistency check, and timestamp validity check.

[0015] It should be noted that the core data types and verification dimensions are defined to provide specific basis for method clauses; the role is to provide a clear standard for binding verification, avoid ambiguity in defining core data, prevent the mapping relationship from being tampered with through blockchain storage, and improve the accuracy of identity recognition through multi-dimensional verification;

[0016] Preferably, the sensitivity dynamic matching module calculates the sensitivity value based on four types of data features, including data leakage influence range, transmission frequency, transmission range, and processing node historical behavior.

[0017] A multi-dimensional fusion model is adopted to realize feature operation and output quantitative scoring; the compliance closed-loop management module captures data flow tracks through kernel state monitoring, and the track data is stored after being encrypted, and the encryption adopts a national encryption algorithm; the high-speed interconnection channel of the three-dimensional linkage scheduling module is a dedicated data transmission channel, and supports low-latency data interaction.

[0018] It needs to be noted that the specific calculation of the feature, the encryption method and the transmission channel type are determined; the role is to fill in the details of the implementation of the core function, so that the sensitivity calculation has specific basis, the national secret encryption guarantees data security, and the special channel ensures that the dispatch data transmission is not blocked;

[0019] Preferably, it comprises the following steps:

[0020] S1: The subject boundary binding module initializes the correspondence between the Internet protocol address and the data type, generates a unique binding identifier, and generates the binding identifier by performing XOR operation on the Internet protocol address of the physical storage node, the data type identifier, the first signature and the second signature, and then performing hash operation; store the binding identifier in the block chain, the number of binding identifiers stored in each block is calculated based on the block capacity and the space occupied by each binding identifier and reserved redundancy;

[0021] S2: When initiating an access request, four verifications are performed: ① The Internet protocol address of the request is matched with the binding Internet protocol address; ② The identity of the request node is verified through the SM2 public key; ③ The binding identifier carried by the request is verified with the binding identifier stored in the block chain; ④ The difference between the request timestamp and the current time is within the valid range, which is determined by the replay attack test;

[0022] S3: After the four verifications are passed, a trigger signal is sent to the sensitivity dynamic matching module; if any verification fails, the sound and light alarm module is prompted, and the Internet protocol address of the request is frozen, and the freezing time is determined by the attack retry probability test.

[0023] It needs to be noted that the steps of binding and verification are specified, the block chain storage makes the mapping relationship unalterable, and the four verifications in order can prevent forgery and replay attacks, and the interception alarm strengthens the boundary security;

[0024] Preferably, the S1 comprises the following steps:

[0025] S11: Initialize the calculation model of the node trust degree factor, the calculation method is 1 minus the result of the number of controversial operations divided by the total number of operations, multiplied by the node qualification coefficient; the number of controversial operations is counted through the block chain log, the total number of operations is the total number of records in the same period, and the node qualification coefficient is based on the industry qualification standard;

[0026] S12: Collect data through network traffic monitoring protocol, extract risk attribute parameters, transmission frequency parameters and cross-domain attribute parameters, risk attribute parameters are quantified according to national information security standards, transmission frequency parameters are quantified according to the number of transmissions per unit time, and cross-domain attribute parameters are determined based on transmission range, and the sensitivity value is obtained by multiplying the three parameters with the node trust degree factor;

[0027] S13: Perform isolation strategy based on the mapping relationship between sensitivity value and hardware isolation capability: when the sensitivity value is in the low interval, use logical encryption; when the sensitivity value is in the medium interval, use physical subnet isolation; when the sensitivity value is in the high interval, use independent device isolation, and the interval threshold is determined by attack test.

[0028] Preferably, the S2 comprises:

[0029] S21: The TPM2.0 chip pre-stores the first node public key and the second node public key, and the chip and the independent device control module establish communication through a communication protocol;

[0030] S22: When receiving an operation instruction, extract the instruction signature and the instruction hash value, first perform a verification operation on the instruction signature and the instruction hash value through the first node public key, then calculate the hash value of the instruction signature, and perform a verification operation on the hash value and the second instruction hash value through the second node public key;

[0031] S23: When the results of the two verifications are both valid, output an allow code to allow the device to access; otherwise, output a prohibit code to block the operation; the verification time consumption is determined by taking the mean value of multiple groups of data measured to ensure that the real-time requirement is met.

[0032] The existing TPM verification is only single node, which cannot meet the multi-node mutual authentication requirement; in connection with the rest of the content, cross verification is performed through double public keys to ensure that the operation instruction is authorized by double nodes, avoid single node tampering, and improve the security of identity authentication;

[0033] Preferably, the S3 comprises:

[0034] S31: Collect data at a fixed period, which is determined by the CPU usage rate fluctuation frequency based on long-term CPU usage rate fluctuation statistics; the collected content includes the CPU usage rate of the node, the cross-border node round-trip delay, and the permission whitelist;

[0035] S32: When the CPU usage rate of the node reaches the load threshold, screen the idle nodes, and the screening condition is that the CPU usage rate is lower than the idle threshold and the cross-border node round-trip delay is lower than the response threshold; the load threshold is determined based on the CPU load and stability curve, the idle threshold is determined by resource utilization and response speed balance test, and the response threshold is determined based on user experience test; the number of concurrent users is the product of the maximum concurrency and the load redundancy coefficient, and the load redundancy coefficient is determined based on fault probability test;

[0036] S33: When calculating the adaptation degree, the weight coefficient corresponding to the delay characteristic in the cross-border scenario is adjusted according to the correlation between the delay and the scheduling effect, and other weight coefficients are adjusted in proportion to maintain the sum as 1; select the node with the highest adaptation degree as the target node, generate the candidate nodes, and the number of candidate nodes is determined based on the reliability requirement calculation.

[0037] S34: Monitor node status, switch to alternative node when CPU usage reaches early warning threshold or cross-border node round-trip delay reaches early warning threshold; load early warning threshold is determined based on performance decline test, delay early warning threshold is determined based on experience test.

[0038] By periodic collection and dynamic threshold adjustment, the scheduling adapts to resource state changes, and the alternative node mechanism improves reliability;

[0039] Preferably, it further comprises S4, which comprises the following steps:

[0040] S41: Use performance test tools to simulate user requests, requests contain fields in typical scenarios, and are designed based on actual application scenarios; set the starting number of users, step size, and target response time;

[0041] S42: Start with the starting number of users and increase the number of users in steps, each step runs for a certain period of time, and records the percentage response time;

[0042] S43: When the percentage response time first exceeds the target response time, take the previous step user number as the maximum concurrent candidate value under this response time;

[0043] S44: Repeat the above process several times, take the average of the multiple candidate values as the final maximum concurrency, ensure that the error meets the requirements, and the number of repetitions is determined by statistical significance test.

[0044] Preferably, it further comprises S5, which comprises the following steps:

[0045] S51: The execution monitoring submodule of the compliance closed-loop management module captures data flow tracks through kernel-level hook functions, and the tracks contain data identifiers, operation node Internet protocol addresses, operation types, and timestamp fields;

[0046] S52: Use the national secret algorithm to encrypt the track data, and the encryption key is generated by the double nodes through the key agreement protocol, and the key rotation period is determined through security testing to control the risk period;

[0047] S53: The encrypted track data is written into the blockchain according to the data identifier, a certain number of records are stored in each block, the block header stores the SM2 joint signature of the double nodes and the hash value of the previous block, forming a chain-like association structure;

[0048] S54: When conducting compliance audits, unlock the blockchain data through double-node signature authorization, sort the decrypted data by timestamp to generate a complete flow track report.

[0049] Preferably, it further comprises S6:

[0050] S61: Based on the local long-term compliance operation log and the violation record, the training data set is constructed, and the proportion of the two types of data in the actual scene is configured; a multi-layer fully connected sub-model is trained, and the model structure is designed based on the feature dimension and the prediction accuracy requirement;

[0051] S62: The sub-model parameters are processed by differential privacy, the privacy budget is determined by the combination of the basic budget and the compliance level correction coefficient, and the node compliance level is evaluated based on the long-term violation times; the processed parameters are uploaded to the fusion server through the homomorphic encryption protocol;

[0052] S63: The fusion server calculates the parameter mean to generate a global model, which is encrypted and distributed to the two nodes for synchronous updating;

[0053] S64: The global model receives data feature codes and operation instruction codes, outputs a risk score, and synchronously triggers a high-risk scheduling strategy to the three-dimensional linkage scheduling module when the score reaches a risk threshold; after accumulating a certain amount of new data, the model is updated repeatedly S61-S63.

[0054] The application provides a private cloud and data center comprehensive management platform and method, which has the following beneficial effects:

[0055] By constructing the collaborative system of four core modules of subject boundary binding, dynamic sensitivity matching, compliance closed-loop management and three-dimensional linkage scheduling, the integration of private cloud and data center security control and resource efficiency is realized;

[0056] Specifically, through the subject boundary binding of block chain storage and multi-dimensional verification, the access identity and data boundary are ensured to be tamper-proof and accurately matched, and illegal access is prevented from the source; based on the dynamic sensitivity calculation of multi-feature fusion and the flow track monitoring of national encryption, the accurate risk assessment and transparent audit of the data life cycle are realized; finally, the real-time risk score and resource state are linked through a high-speed channel, and the three-dimensional scheduling module generates an optimal resource allocation scheme considering safety and performance. BRIEF DESCRIPTION OF DRAWINGS

[0057] Figure 1 The method flowchart of the application. DETAILED DESCRIPTION

[0058] The technical solutions in the embodiments of the application will be described clearly and completely below with reference to the drawings in the embodiments of the application. Obviously, the described embodiments are only part of the embodiments of the application, not all. Based on the embodiments in the application, all other embodiments obtained by those skilled in the art without creative labor are within the scope of protection of the application.

[0059] As Figure 1As an example, the application provides a private cloud and data center integrated management platform and method, which includes a subject boundary binding module;

[0060] The subject boundary binding module is deployed on a gateway server and is used to realize binding verification of access identity and data boundary. The input of the subject boundary binding module is an access request initiated by an external or internal source. The subject boundary binding module outputs an interception instruction or a verification pass signal by verifying whether the IP and the core data type in the access request match a pre-stored corresponding relationship. For an access request that does not match, the subject boundary binding module sends an interception instruction to a hardware firewall. For an access request that matches and has a valid signature, the subject boundary binding module outputs a verification pass signal.

[0061] The sensitivity dynamic matching module is used to calculate a data sensitivity level. The input of the sensitivity dynamic matching module is the verification pass signal output by the subject boundary binding module. The sensitivity dynamic matching module calculates a sensitivity value based on the data features associated with the verification pass signal and outputs the sensitivity value.

[0062] The compliance closed-loop management module is used to receive the sensitivity value output by the sensitivity dynamic matching module and calculate a risk score in combination with a data flow trajectory.

[0063] The three-dimensional linkage scheduling module is connected to the compliance closed-loop management module through a high-speed interconnection channel, receives the risk score output by the compliance closed-loop management module, calculates a resource adaptation degree in combination with resource state data, and outputs an optimal resource scheduling scheme based on the resource adaptation degree.

[0064] It should be further noted that, under the provision of the core architecture framework, subsequent clauses are implemented around this architecture. The main functions include identity verification, sensitivity evaluation, compliance judgment, and resource scheduling, which solve the problem of functional isolation, security, and efficiency disconnection of existing platforms. The entire technical solution is complete and not fragmented.

[0065] As an optional embodiment, the pre-stored corresponding relationship of the subject boundary binding module is a unique mapping of core data types and authenticated physical storage node IPs. The core data types include personal sensitive information, business core data, and key data.

[0066] The mapping relationship is written into a blockchain after being confirmed by double-node digital signature. The signature uses an SM2 asymmetric encryption algorithm. The binding verification includes four dimensions of IP prefix matching, identity public key verification, binding identifier consistency check, and timestamp validity check.

[0067] It should be noted that the core data types and verification dimensions are defined to provide specific basis for the method clauses. The role is to provide a clear standard for binding verification, avoid ambiguity in defining core data, prevent the mapping relationship from being tampered with by blockchain storage, and improve the accuracy of identity recognition through multi-dimensional verification.

[0068] As an optional embodiment: the sensitivity dynamic matching module calculates the sensitivity value based on four types of data characteristics: data leakage influence range, transmission frequency, transmission range, and processing node historical behavior;

[0069] A multi-dimensional fusion model is used to realize feature operation and output quantitative scores. The compliance closed-loop management module captures data flow tracks through kernel state monitoring, and the track data is stored after encryption. The encryption uses a national encryption algorithm. The high-speed interconnection channel of the three-dimensional linkage scheduling module is a special data transmission channel that supports low-latency data interaction.

[0070] It should be noted that the specific calculation of the feature, encryption method and transmission channel type are specified. The purpose is to fill in the details of the implementation of the core function, provide a specific basis for sensitivity calculation, ensure data security through national encryption, and ensure smooth data transmission through a special channel.

[0071] As an optional embodiment, it includes:

[0072] S1: The subject boundary binding module initializes the correspondence between the Internet protocol address and the data type, generates a unique binding identifier, and generates it by performing XOR operation on the hash values of the Internet protocol address of the physical storage node, data type identifier, first signature and second signature, and then performing hash operation. Store the binding identifier in the blockchain. The number of binding identifiers stored in each block is calculated based on the block capacity and the space occupied by each binding identifier, and a redundancy is reserved;

[0073] S2: When initiating an access request, perform fourfold verification: ① The Internet protocol address of the request is matched with the bound Internet protocol address; ② Verify the identity of the request node through the SM2 public key; ③ Check whether the binding identifier carried by the request is consistent with the binding identifier stored in the blockchain; ④ Check whether the difference between the request timestamp and the current time is within the valid range, which is determined by the replay attack test;

[0074] S3: After passing the fourfold verification, send a trigger signal to the sensitivity dynamic matching module. If any verification fails, the sound and light alarm module will prompt and freeze the Internet protocol address of the request. The freezing duration is determined by the attack retry probability test.

[0075] It should be noted that the steps of binding and verification are specified. The blockchain storage makes the mapping relationship tamper-proof, and the ordered fourfold verification can prevent forgery and replay attacks, and strengthen the security of the boundary.

[0076] As an optional embodiment, S1 includes the following steps:

[0077] S11: initialize the calculation model of the node trust factor, the calculation method is 1 minus the number of controversial operations divided by the total number of operations, multiplied by the node qualification coefficient; the number of controversial operations is counted through the blockchain log, the total number of operations is the total number of records in the same period, and the node qualification coefficient is based on the industry qualification standard;

[0078] S12: collect data through network traffic monitoring protocol, extract risk attribute parameters, transmission frequency parameters and cross-domain attribute parameters, risk attribute parameters are quantified according to national information security standards, transmission frequency parameters are quantified according to the number of transmissions per unit time, and cross-domain attribute parameters are determined based on transmission range, and the three parameters are multiplied by the node trust factor to obtain the sensitivity value;

[0079] S13: execute the isolation strategy based on the mapping relationship between the sensitivity value and the hardware isolation capability: when the sensitivity value is in the low interval, logical encryption is used; when the sensitivity value is in the middle interval, physical subnet isolation is used; when the sensitivity value is in the high interval, independent device isolation is used, and the threshold values of each interval are determined through attack test.

[0080] As an optional embodiment, the S2 comprises:

[0081] S21: the TPM2.0 chip pre-stores the first node public key and the second node public key, and the chip and the independent device control module establish communication through a communication protocol;

[0082] S22: when receiving an operation instruction, extract the instruction signature and the instruction hash value, first execute the verification operation on the instruction signature and the instruction hash value through the first node public key, then calculate the hash value of the instruction signature, and execute the verification operation on the hash value and the second instruction hash value through the second node public key;

[0083] S23: when the two verification results are both valid, output the permission code to allow the device to access; otherwise, output the prohibition code to block the operation; the verification time is determined by taking the mean value of multiple groups of data through actual measurement to ensure that the real-time requirement is met.

[0084] The existing TPM verification is only for a single node, which cannot meet the multi-node mutual authentication requirement; in connection with the rest of the content, the double public key cross verification is used to ensure that the operation instruction is authorized by double nodes, avoid single node tampering, and improve the security of identity authentication;

[0085] As an optional embodiment, the S3 comprises:

[0086] S31: collect data at a fixed period, the period is determined by the CPU usage rate fluctuation frequency, based on long-term CPU usage rate fluctuation statistics; the collected content includes the CPU usage rate of the node, the cross-border node round-trip delay and the permission white list;

[0087] S32: When the CPU usage of the node reaches the load threshold, screen the idle nodes, and the screening condition is that the CPU usage is lower than the idle threshold, and the cross-border node round-trip delay is lower than the response threshold; the load threshold is determined based on the CPU load and stability curve, the idle threshold is determined by the balance test of resource utilization and response speed, and the response threshold is determined based on the user experience test; the number of concurrent users is the product of the maximum concurrency and the load redundancy coefficient, and the load redundancy coefficient is determined based on the fault probability test;

[0088] S33: When calculating the adaptation degree, the weight coefficient corresponding to the delay characteristic in the cross-border scenario is adjusted according to the correlation between the delay and the scheduling effect, and other weight coefficients are adjusted in proportion to maintain the sum of 1; the node with the highest adaptation degree is selected as the target node, and the candidate nodes are generated, and the number of candidate nodes is determined based on the reliability requirement;

[0089] S34: Monitor the node state, and switch to the candidate node when the CPU usage reaches the warning threshold or the cross-border node round-trip delay reaches the warning threshold; the load warning threshold is determined based on the performance decline test, and the delay warning threshold is determined based on the experience test.

[0090] Through periodic collection and dynamic threshold adjustment, the scheduling adapts to the resource state change, and the candidate node mechanism improves the reliability;

[0091] As an optional embodiment, it further includes S4, which includes the following steps:

[0092] S41: Use a performance test tool to simulate user requests, the request contains fields in typical scenarios, and is designed based on actual application scenarios; set the starting number of users, the step size, and the target response time;

[0093] S42: Start with the starting number of users and increase the number of users in steps, and record the percentage response time after stable operation for a certain period of time;

[0094] S43: When the percentage response time first exceeds the target response time, take the previous step user number as the maximum concurrency candidate value under the response time;

[0095] S44: Repeat the above process several times, take the average of the multiple candidate values as the final maximum concurrency, ensure that the error meets the requirements, and the number of repetitions is determined by statistical significance test.

[0096] As an optional embodiment, it further includes S5, which includes the following steps:

[0097] S51: The execution monitoring submodule of the compliance closed-loop management module captures the data flow trajectory through the kernel state hook function, and the trajectory includes data identification, operation node Internet protocol address, operation type, and timestamp field;

[0098] S52: The trajectory data is encrypted using a national secret algorithm, the encryption key is generated by the double nodes through a key agreement protocol, and the key rotation period is determined by selecting a controllable risk period through security testing;

[0099] S53: The encrypted trajectory data is written into the blockchain according to the data identifier, a certain number of records are stored in each block, the block header stores the SM2 joint signature of the double nodes and the hash value of the previous block, forming a chain-like association structure;

[0100] S54: During compliance audit, the double nodes are authorized to unlock the blockchain data through signature, and after decryption, the complete flow trajectory report is generated according to the timestamp.

[0101] As an optional embodiment, it also includes S6:

[0102] S61: The double nodes construct training data sets based on local long-term compliance operation logs and violation records respectively, and configure them according to the proportion of the two types of data in the actual scene; train multi-layer fully connected sub-models, and the model structure is designed based on feature dimension and prediction accuracy requirement;

[0103] S62: The sub-model parameters are processed by differential privacy, the privacy budget is determined by combining the basic budget and the compliance level correction coefficient, and the node compliance level is evaluated based on the long-term violation times; the processed parameters are uploaded to the fusion server through a homomorphic encryption protocol;

[0104] S63: The fusion server calculates the mean value of the parameters to generate a global model, and distributes it to the double nodes for synchronous update;

[0105] S64: The global model receives data feature codes and operation instruction codes, outputs risk scores, and when the score reaches the risk threshold, it is synchronized to the three-dimensional linkage scheduling module to trigger high-risk scheduling strategies; after accumulating a certain amount of new data, repeat S61-S63 to update the model.

[0106] Although embodiments of the present application have been shown and described, it will be understood by those of ordinary skill in the art that various changes, modifications, substitutions and alterations can be made without departing from the principles and spirit of the present application, and the scope of the present application is defined by the appended claims and their equivalents.

Claims

1. A private cloud and data center integrated management platform, including a subject boundary binding module; The subject boundary binding module, deployed on the gateway server, is used to implement the binding and verification of access identity and data boundary. Its input is the access request initiated from the outside or inside; by verifying whether the IP in the access request matches the core data type according to the pre-stored correspondence, it outputs an interception command or a verification pass signal; for access requests that do not match, it sends an interception command to the hardware firewall; for access requests that match and have a valid signature, it outputs a verification pass signal. The sensitivity dynamic matching module is used to calculate the data sensitivity level. Its input is the verification pass signal output by the subject boundary binding module. Based on the data features associated with the verification pass signal, the sensitivity value is calculated and output. The compliance closed-loop management module receives the sensitivity values ​​output by the sensitivity dynamic matching module and calculates and outputs a risk score based on the data flow trajectory. The three-dimensional linkage scheduling module is connected to the compliance closed-loop management module through a high-speed interconnection channel. It receives the risk score output by the compliance closed-loop management module, calculates the resource suitability by combining it with resource status data, and outputs the optimal resource scheduling plan based on the resource suitability.

2. The private cloud and data center integrated management platform according to claim 1, characterized in that: The pre-stored correspondence of the subject boundary binding module is a unique mapping between core data types and certified physical storage node IPs. Core data types include personal sensitive information, core business data, and key data. The mapping relationship is written to the blockchain after being confirmed by dual-node digital signatures. The signature uses the SM2 asymmetric encryption algorithm. The binding verification includes four dimensions: IP prefix matching, identity public key verification, binding identifier consistency verification, and timestamp validity verification.

3. The private cloud and data center integrated management platform according to claim 1, characterized in that: The sensitivity dynamic matching module calculates sensitivity values ​​based on four types of data characteristics: the scope of impact of data leakage, transmission frequency, transmission range, and historical behavior of processing nodes. A multi-dimensional fusion model is used to perform feature calculations and output a quantitative score. The compliance closed-loop management module captures data flow trajectories through kernel-mode monitoring. The trajectory data is stored after encryption, and the encryption adopts the national cryptographic algorithm. The high-speed interconnection channel of the three-dimensional linkage scheduling module is a dedicated data transmission channel, which supports low-latency data interaction.

4. A method for integrated management of private cloud and data center, applicable to the integrated management platform for private cloud and data center as described in any one of claims 1 to 3, characterized in that: include: S1: The main boundary binding module initializes the correspondence between Internet Protocol addresses and data types, and generates a unique binding identifier. The generation method is to perform an XOR operation on the Internet Protocol address of the physical storage node, the data type identifier, and the hash values ​​of the first signature and the second signature, and then perform a hash operation. The binding identifiers are stored on the blockchain. The number of binding identifiers stored in each block is calculated based on the block capacity and the space occupied by each binding identifier, with redundancy reserved. S2: When initiating an access request, perform four-fold verification: ① Match the prefix of the requested Internet Protocol address with the bound Internet Protocol address; ② Verify the identity of the requesting node using the SM2 public key; ③ Verify whether the binding identifier carried in the verification request is consistent with the binding identifier stored in the blockchain; ④ Verify whether the difference between the request timestamp and the current time is within a valid range. This range is determined through replay attack testing. S3: After all four verifications pass, a trigger signal is sent to the sensitivity dynamic matching module; if any verification fails, an audible and visual alarm module will be activated and the requested Internet Protocol address will be frozen. The duration of the freeze will be determined by an attack retry probability test.

5. The method for integrated management of private cloud and data center according to claim 4, characterized in that: S1 includes the following steps: S11: Initialize the calculation model for the node trust factor. The calculation method is to subtract the number of disputed operations from the total number of operations from 1, and then multiply it by the node qualification coefficient. The number of disputed operations is counted through blockchain logs, the total number of operations is the total number of operation records in the same period, and the node qualification coefficient is assessed based on industry qualification standards. S12: Collect data through network traffic monitoring protocol, extract risk attribute parameters, transmission frequency parameters, and cross-domain attribute parameters. Risk attribute parameters are quantified according to national information security standards, transmission frequency parameters are quantified according to the number of transmissions per unit time, and cross-domain attribute parameters are determined based on the transmission range. Multiply the three parameters by the node trust factor to obtain the sensitivity value. S13: Execute isolation strategies based on the mapping relationship between sensitivity values ​​and hardware isolation capabilities: when the sensitivity value is in the low range, logical encryption is used; when the sensitivity value is in the middle range, physical subnet isolation is used; when the sensitivity value is in the high range, independent device isolation is used. The threshold for each range is determined through attack testing.

6. The method for integrated management of private cloud and data center according to claim 4, characterized in that: S2 includes: S21: The TPM2.0 chip pre-stores the public keys of the first and second nodes, and the chip establishes communication with the independent device management module through a communication protocol; S22: When an operation instruction is received, extract the instruction signature and instruction hash value. First, perform a verification operation on the instruction signature and instruction hash value using the first node's public key. Then, calculate the hash value of the instruction signature and perform a verification operation on the hash value and the first instruction hash value using the second node's public key. S23: When both verification results are valid, output an allow code to allow device access; otherwise, output a prohibit code to block the operation; the verification time is determined by taking the average of multiple sets of measured data to ensure that real-time requirements are met.

7. The method for integrated management of private cloud and data center according to claim 4, characterized in that: S3 includes: S31: Data is collected at a fixed period, which is determined by the frequency of CPU utilization fluctuations and is derived from long-term CPU utilization fluctuation statistics; the collected data includes the CPU utilization of nodes, cross-border node round-trip latency, and permission whitelist. S32: When the CPU utilization of a node reaches the load threshold, idle nodes are selected. The selection criteria are that the CPU utilization is lower than the idle threshold and the round-trip latency of cross-border nodes is lower than the response threshold. The load threshold is determined based on the CPU load and stability curve, the idle threshold is determined through resource utilization and response speed balance test, and the response threshold is determined based on user experience test. The number of concurrent users is the product of the maximum concurrency and the load redundancy coefficient. The load redundancy coefficient is determined based on the failure probability test. S33: When calculating the fit, the weight coefficient corresponding to the latency characteristic in cross-border scenarios is adjusted according to the correlation between latency and scheduling effect, and other weight coefficients are adjusted proportionally to keep the sum of 1; the node with the highest fit is selected as the target node, and candidate nodes are generated. The number of candidate nodes is calculated and determined based on reliability requirements. S34: Monitor node status. When CPU utilization reaches the warning threshold or cross-border node round-trip delay reaches the warning threshold, switch to the backup node. The load warning threshold is determined based on performance degradation test, and the delay warning threshold is determined based on experience test.

8. The method for integrated management of private cloud and data center according to claim 4, characterized in that: It also includes S4, which includes the following steps: S41: Use performance testing tools to simulate user requests. The requests include fields from typical scenarios and are designed based on real-world application scenarios. Set the initial number of users, step size, and target response time. S42: Starting from the initial number of users, increase the number of users in a step-by-step manner, with each level running stably for a certain duration, and record the percentage response time. S43: When the percentage response time exceeds the target response time for the first time, take the number of users in the previous tier as the candidate value for the maximum concurrency under that response time; S44: Repeat the above process multiple times, and take the average of the multiple candidate values ​​as the final maximum concurrency to ensure that the error meets the requirements. The number of repetitions is determined by statistical significance testing.

9. The method for integrated management of private cloud and data center according to claim 4, characterized in that: It also includes S5, which includes the following steps: S51: The execution monitoring submodule of the compliance closed-loop management module captures the data flow trajectory through kernel-mode hook functions. The trajectory includes data identifier, operation node Internet protocol address, operation type, and timestamp fields. S52: The trajectory data is encrypted using the national cryptographic algorithm. The encryption key is generated by the two nodes through a key negotiation protocol. The key rotation cycle is determined by selecting a risk-controllable cycle through security testing. S53: The encrypted trajectory data is partitioned and written to the blockchain according to the data identifier. Each block stores a certain number of records. The block header stores the joint signature of the two nodes SM2 and the hash value of the previous block, forming a chain-related structure. S54: During compliance audits, blockchain data is unlocked through dual-node signature authorization, and after decryption, a complete flow trajectory report is generated by sorting by timestamp.

10. The method for integrated management of private cloud and data center according to claim 4, characterized in that: Also includes S6: S61: The two nodes respectively build training datasets based on local long-term compliance operation logs and violation records, and are configured according to the proportion of the two types of data in the actual scenario; train multi-layer fully connected sub-models, and the model structure is designed based on feature dimension and prediction accuracy requirements; S62: Sub-model parameters undergo differential privacy processing. The privacy budget is determined by a combination of the base budget and the compliance level correction coefficient. The node compliance level is assessed based on the number of long-term violations. The processed parameters are uploaded to the fusion server via a homomorphic encryption protocol. S63: The average parameter calculation of the fusion server generates a global model, which is then encrypted and distributed to the two nodes for synchronous updates. S64: The global model receives data feature codes and operation instruction codes, outputs risk scores, and when the scores reach the risk threshold, it is synchronized to the three-dimensional linkage scheduling module to trigger a high-risk scheduling strategy; after accumulating a certain amount of new data, the model is updated by repeating S61-S63.

Images