A distributed server encryption storage security protection and data processing method

Abstract

The application provides a distributed server encryption storage security protection and data processing method, including the following steps: deploying a monitoring agent; deploying a lightweight monitoring agent on each storage server node of a distributed cluster to collect multi-dimensional performance indexes of each storage server node in real time. The distributed server encryption storage security protection and data processing method provided by the application can comprehensively collect relevant data of each storage server node, thereby providing safe data support for subsequent node load analysis and storage scheduling decision, and meanwhile, by using an HKDF algorithm and combining multi-dimensional characteristic parameters, independent and exclusive key encryption is realized on each data fragment, and a safe processing mechanism of key ciphertext special line delivery and in-memory key plaintext instant erasure is used, so that the data storage encryption security level is improved.

Description

Technical Field

[0001] This application relates to the field of big data technology, and in particular to a distributed server encrypted storage security protection and data processing method. Background Technology

[0002] With the rapid development of big data and cloud computing technologies, the demand for massive data storage and security protection continues to rise. Against this backdrop, distributed server storage systems, with their advantages of elasticity, scalability, and high concurrency throughput, have become the core architecture for enterprise-level data storage, which also places higher demands on the server's data operation and management capabilities.

[0003] Existing distributed server data processing technologies primarily employ a scheme that generates a single, unified encryption key to encrypt and encapsulate all types of network data transmitted and stored within the cluster. This key is reused throughout the entire process, including data interaction, storage retrieval, decryption, and verification. However, relying solely on a fixed encryption algorithm for basic encryption and decryption operations lacks security and scheduling logic such as hierarchical key derivation, fragmented key matching, and dynamic scheduling of storage nodes. The encryption implementation is simplistic, and the cluster resource allocation mechanism is rigid, failing to dynamically schedule computing resources based on data task load and the real-time operating status of each node. This makes it difficult to fully utilize the surplus computing power of idle nodes, easily leading to computing power blockage and resulting in significantly low overall cluster resource utilization. Summary of the Invention

[0004] This application aims to at least partially solve one of the technical problems in the aforementioned technologies.

[0005] Therefore, one objective of this application is to propose a distributed server encrypted storage security protection and data processing method, which can comprehensively collect the operating data of each storage node, provide reliable data support for node load analysis and storage scheduling decisions, improve the resource utilization and operational stability of the distributed cluster, and at the same time, encrypt each data shard with an independent dedicated key, combined with the key encryption dedicated line distribution and the memory plaintext instant erasure mechanism, thereby improving the data storage encryption security level.

[0006] To achieve the above objectives, the first aspect of this application proposes a distributed server encrypted storage security protection and data processing method, comprising the following steps: Deploying a monitoring agent: Deploying a lightweight monitoring agent on each storage server node of the distributed cluster to collect multi-dimensional performance indicators of each storage server node in real time; Data collection and transmission: Establishing a normalized data reporting link to transmit the collected multi-dimensional performance indicators and the current node's own identifier to the cluster's management master node; Data preprocessing: Before weighted calculation, cleaning the data, including verification, correction of outliers, and supplementation of missing values; Deep data analysis: Calculating the comprehensive load score of each effective node using a weighted scoring formula, and identifying comprehensive load scores less than a preset global load threshold. The process involves: constructing a low-load storage candidate pool; data encryption preprocessing: before storing data, a key distribution request message is sent to an independent key management server. Using the HKDF hierarchical key derivation algorithm, a unique encryption key is generated for each data shard and transmitted to the data writing client or cluster gateway via an encrypted dedicated line; discrete storage scheduling: the node with the lowest overall load score in the low-load storage candidate pool is selected as the target writing node for the current shard. A constraint-based scheduling algorithm is used to store multiple shards of the same file on different nodes, and the encrypted ciphertext shards and corresponding metadata are simultaneously transmitted to the corresponding low-load storage nodes; security protection: based on access permissions, it is determined whether the current identity has access permissions for the specified shard requested.

[0007] In addition, the distributed server encrypted storage security protection and data processing method proposed above in this application may also have the following additional technical features:

[0008] Specifically, in the step of deploying the monitoring agent, the real-time collection of multi-dimensional performance indicators of each storage server node includes CPU utilization, available memory capacity, disk I / O operations per second, network bandwidth inbound / outbound rate, and the remaining space percentage of the storage volume.

[0009] Specifically, in the data acquisition and transmission steps, establishing a normalized data reporting link includes building a Grip streaming communication interface based on the HTTP / 2 protocol, establishing a long connection channel between the monitoring agent and the management master node, and encrypting the long connection channel end-to-end using the TLS 1.3 security protocol to generate an encrypted transmission layer.

[0010] Specifically, in the data preprocessing step, data cleaning includes a sliding time window algorithm and a box plot outlier detection algorithm. The sliding time window algorithm constructs a continuous observation window for time series performance indicators and uses time series interpolation to complete and repair missing indicators. The box plot outlier detection algorithm defines the reasonable distribution range of values ​​by calculating the quartiles and interquartile ranges of the data, and identifies and removes extreme outlier data points that deviate from three standard deviations.

[0011] Specifically, in the deep data analysis step, the calculation of the comprehensive load score for each effective node using a weighted scoring formula includes the following steps: Feature substitution: Substituting the cleaned multidimensional performance indicators into the preset weighted scoring formula for linear weighted fusion calculation; Load score solution: Obtaining the corresponding real-time comprehensive load score through the above multiplication and addition operations, wherein the value of the comprehensive load score is positively correlated with the real-time load pressure of the node, and the lower the score, the higher the proportion of idle resources of the node and the greater the resource redundancy that can bear storage tasks; Low load candidate pool construction: Setting a load threshold, including nodes with a comprehensive load score less than the global load threshold into the initial selection set, and removing offline and abnormal faulty nodes based on the node heartbeat detection results to construct a low load storage candidate pool.

[0012] Specifically, in the data encryption preprocessing step, the generation of a unique encryption key for each data shard using the HKDF hierarchical key derivation algorithm includes the following steps: Parameter initialization: The key management server receives the encryption request message sent by the cluster management master node and extracts the three-dimensional information from the request message; Hierarchical iterative derivation: Based on the HKDF algorithm, multiple hash iteration diffusion operations are performed using salted hash calculation to eliminate key association and generate a unique encryption key for each data shard; Secure key distribution: The exclusive encryption key in ciphertext form is distributed to the data writing client or cluster gateway through an encrypted dedicated line, and the master key material and intermediate master key plaintext traces in memory are erased after distribution.

[0013] Specifically, the three dimensions of information in the extraction request message include a globally unique file identifier, the file splitting sequence number, and the target storage node for the encrypted fragment to be written.

[0014] Specifically, in the discrete storage scheduling process, the step of storing multiple fragments of the same file on different nodes using a constraint-based scheduling algorithm includes the following steps: Target node selection: Based on the low-load storage candidate pool, obtain the real-time comprehensive load score of each candidate node, sort them in ascending order of score, and prioritize the node with the lowest score and the highest resource redundancy as the preferred target node for the current fragment to be written; Constraint verification: Construct a set of occupied nodes for the current file, compare the identifier of the preferred target node with the set of occupied node identifiers, and select the physical node that does not exist in the set; Fragment discrete writing: Under the premise of satisfying the dual constraints of load optimization and node isolation, transmit the encrypted fragments and corresponding metadata and write them to the selected target storage node; Iterative update of scheduling status: Append the selected target storage node to the set of occupied node identifiers, and synchronously refresh the real-time comprehensive load score of the corresponding node in the low-load storage candidate pool, providing the latest node status data for the constraint scheduling of the next fragment, and realizing iterative optimization of the cluster-wide load.

[0015] Specifically, determining whether the current identity has access to the specified fragment of this request includes: when the identity token or digital certificate of the accessing subject fails the legitimacy and validity verification, it is determined as illegal access and the request link is blocked; when the identity token or digital certificate of the accessing subject passes the legitimacy and validity verification, a preset access control policy table is further queried to determine whether the identity identifier of the accessing subject has access rights.

[0016] Specifically, the further querying of the preset access control policy table to determine whether the identity of the accessing subject has query permissions includes: if the accessing subject has file-level access permissions but has not configured the independent permissions for the corresponding fragment, it is determined as illegal access and the request link is blocked; if the accessing subject's identity is legitimate and the fragment permissions are matched bidirectionally and pass the double verification condition, the data access request is allowed.

[0017] Compared with existing technologies, this application has the following advantages: The distributed server encrypted storage security protection and data processing method of this application, by deploying a lightweight deployment monitoring agent on each storage server node of the distributed cluster, can comprehensively collect relevant data from each storage server node, thereby providing secure data support for subsequent node load analysis and storage scheduling decisions, improving the overall resource utilization and operational stability of the distributed cluster. At the same time, by using the HKDF algorithm and combining multi-dimensional feature parameters, each data shard is encrypted with an independent dedicated key, and with the secure processing mechanism of key ciphertext offline transmission and instant memory key plaintext erasure, the data storage encryption security level is improved.

[0018] Additional aspects and advantages of this application will be set forth in part in the description which follows, and in part will be obvious from the description, or may be learned by practice of this application. Attached Figure Description

[0019] The above and / or additional aspects and advantages of this application will become apparent and readily understood from the following description of the embodiments taken in conjunction with the accompanying drawings, wherein:

[0020] Figure 1 This is a flowchart illustrating the steps of a distributed server encrypted storage security protection and data processing method according to an embodiment of this application;

[0021] Figure 2 The flowchart below shows the steps of the weighted scoring formula in this application.

[0022] Figure 3 This is a flowchart of the steps of the HKDF hierarchical key derivation algorithm in this application;

[0023] Figure 4 This is a flowchart of the steps of the constraint-based scheduling algorithm in this application. Detailed Implementation

[0024] The embodiments of this application 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 intended to explain this application, and should not be construed as limiting this application.

[0025] The following describes a distributed server encrypted storage security protection and data processing method according to an embodiment of this application, with reference to the accompanying drawings.

[0026] like Figures 1-4 As shown in the figure, a distributed server encrypted storage security protection and data processing method according to an embodiment of this application includes the following steps:

[0027] Deploy monitoring agents: Deploy lightweight monitoring agents on each storage server node of the distributed cluster to collect multi-dimensional performance metrics of each storage server node in real time.

[0028] In the deployment of the monitoring agent, the real-time collection of multi-dimensional performance indicators of each storage server node includes CPU utilization, available memory capacity, disk I / O operations per second, network bandwidth inbound / outbound rate, and the percentage of remaining space on the storage volume.

[0029] As illustrated by the example, using a lightweight monitoring agent to collect multi-dimensional performance metrics of node operation can provide feedback on the real-time operating load status of each storage node, restore the true operating conditions of the node, and provide real-time raw data support for subsequent data cleaning and analysis, node load scoring, and dynamic scheduling of storage shards, avoiding the problem of one-sided and distorted node load judgment caused by monitoring a single metric.

[0030] Data Acquisition and Transmission: Establish a regular data reporting link to transmit the collected multi-dimensional performance indicators and the current node's own identifier to the cluster's management master node.

[0031] In the data acquisition and transmission steps, establishing a normalized data reporting link includes building a Grip streaming communication interface based on the HTTP / 2 protocol, establishing a long connection channel between the monitoring agent and the management master node, and encrypting the long connection channel end-to-end through the TLS 1.3 security protocol to generate an encrypted transmission layer.

[0032] As illustrated by the example, the HTTP / 2 protocol features multiplexing, frame transmission, and header compression, enabling concurrent transmission of metrics from multiple nodes. This improves the throughput efficiency and transmission stability of cluster data reporting. Furthermore, through the TLS 1.3 end-to-end encryption mechanism, node performance data and node identity information can be encrypted and protected during transmission, preventing security risks such as eavesdropping, tampering, and forgery during data transmission. This provides secure data support for subsequent node load analysis and storage scheduling decisions.

[0033] Data preprocessing: Before the data is weighted, it is cleaned, including verification, correction of outliers and filling in missing values.

[0034] In the data preprocessing step, data cleaning includes a sliding time window algorithm and a box plot outlier detection algorithm. The sliding time window algorithm constructs a continuous observation window for time series performance indicators and uses time series interpolation to complete and repair missing indicators. The box plot outlier detection algorithm defines the reasonable distribution range of values ​​by calculating the quartiles and interquartile ranges of the data, and identifies and removes extreme outlier data points that deviate from three standard deviations.

[0035] As illustrated by the example, during the data acquisition and transmission process, the sliding time window algorithm adapts to the time-series acquisition characteristics of server performance indicators. It uses a continuous time observation window to lock the effective data of the preceding and following time series, and uses time-series interpolation to fill in the missing indicators caused by instantaneous network fluctuations and short-term node lag, ensuring the continuity and integrity of time-series data. At the same time, the box plot outlier detection algorithm can filter out extreme abnormal data caused by instantaneous hardware failures and abnormal acquisition modules, avoiding outliers from interfering with the overall data distribution, realizing standardized preprocessing of monitoring data, and providing a data foundation for subsequent in-depth data analysis.

[0036] Deep data analysis: The weighted scoring formula is used to calculate the comprehensive load score of each effective node, identify the comprehensive load scores that are less than the preset global load threshold, and construct a low-load storage candidate pool.

[0037] In the deep data analysis step, the calculation of the comprehensive load score for each effective node using a weighted scoring formula includes the following steps: Feature substitution: Substituting the cleaned multidimensional performance indicators into the preset weighted scoring formula for linear weighted fusion calculation; Load score solution: Obtaining the corresponding real-time comprehensive load score through the above multiplication and addition operations, wherein the value of the comprehensive load score is positively correlated with the real-time load pressure of the node, and the lower the score, the higher the proportion of idle resources of the node and the greater the resource redundancy that can bear storage tasks; Low load candidate pool construction: Setting a load threshold, including nodes with a comprehensive load score less than the global load threshold in the initial selection set, and removing offline and abnormal faulty nodes based on the node heartbeat detection results to construct a low load storage candidate pool.

[0038] For example, by linearly weighting and fusing pre-processed multi-dimensional indicators such as CPU utilization, available memory capacity, disk I / O operations per second, network bandwidth inbound / outbound rate, and remaining space percentage of storage volume, the weights can be reasonably allocated according to the impact of each performance indicator on node load, quantifying the overall load pressure of the nodes. At the same time, by setting a global load threshold, preliminary node screening is completed, eliminating offline, faulty, and abnormally responding invalid nodes, preventing failed nodes from being included in the scheduling candidate set, providing an effective node resource pool for the optimal scheduling of subsequent data sharding, and ensuring cluster load balancing and data writing stability.

[0039] Data encryption preprocessing: Before storing data, a key distribution request message is sent to an independent key management server. Through the HKDF hierarchical key derivation algorithm, a unique encryption key is generated for each data fragment and transmitted to the data writing client or cluster gateway through an encrypted dedicated line.

[0040] In the data encryption preprocessing step, the generation of a unique encryption key for each data shard using the HKDF hierarchical key derivation algorithm includes the following steps: Parameter initialization: The key management server receives the encryption request message sent by the cluster management master node and extracts the three-dimensional information from the request message; Hierarchical iterative derivation: Based on the HKDF algorithm, multiple hash iteration diffusion operations are performed using salted hash calculation to eliminate key association and generate a unique encryption key for each data shard; Secure key distribution: The exclusive encryption key in ciphertext form is distributed to the data writing client or cluster gateway via an encrypted dedicated line, and the master key material and intermediate master key plaintext traces in memory are erased after distribution.

[0041] The three dimensions of information extracted from the request message include the globally unique file identifier, the fragment sequence number after file splitting, and the target storage node for the ciphertext fragment to be written.

[0042] To illustrate, the three types of characteristic information—globally unique file identifier, fragment sequence number, and target storage node—are used together as input factors for key derivation. Combined with multiple salted hash iterative diffusion operations, the linear correlation between keys is eliminated, increasing the difficulty of cracking the keys. At the same time, the plaintext key in memory is cleared immediately after the key is distributed to prevent the risk of information leakage caused by the residue of plaintext key, thereby improving the encryption security and fault tolerance level of the data.

[0043] Discrete storage scheduling: The node with the lowest overall load score in the low-load storage candidate pool is selected as the target write node for the current fragment. Multiple fragments of the same file are stored on different nodes through a constraint scheduling algorithm. The encrypted ciphertext fragments and corresponding metadata are synchronously transmitted to the corresponding low-load storage nodes.

[0044] In the discrete storage scheduling process, the step of storing multiple fragments of the same file on different nodes using a constraint-based scheduling algorithm includes the following steps: Target node selection: Based on the low-load storage candidate pool, obtain the real-time comprehensive load score of each candidate node, sort them in ascending order of score, and prioritize the node with the lowest score and the highest resource redundancy as the preferred target node for the current fragment to be written; Constraint verification: Construct a set of occupied nodes for the current file, compare the identifier of the preferred target node with the set of occupied node identifiers, and select the physical node that does not exist in the set; Fragment discrete writing: Under the premise of satisfying the dual constraints of load optimization and node isolation, transmit the encrypted fragments and corresponding metadata and write them to the selected target storage node; Iterative update of scheduling status: Append the selected target storage node to the set of occupied node identifiers, and synchronously refresh the real-time comprehensive load score of the corresponding node in the low-load storage candidate pool, providing the latest node status data for the constraint scheduling of the next fragment, and realizing iterative optimization of the cluster-wide load.

[0045] To illustrate, by using a low-load storage candidate pool as the screening basis, the node with the highest resource redundancy is prioritized to undertake the write task, ensuring efficient utilization of cluster computing power and storage resources. Combined with the constraint verification of the already occupied node set, cross-node discrete storage of data shards of the same file is achieved, avoiding the centralized deployment of multiple shards on the same physical node and avoiding the problem of data loss or read failure of the entire file due to single node failure. At the same time, by iteratively updating the node occupancy status and load score in real time, the cluster status is dynamically updated after each shard write, so that subsequent shard scheduling is always based on the latest node load data, ensuring the real-time and accurate scheduling strategy and avoiding the problem of overall file unavailability caused by single node failure.

[0046] Security protection: Based on access permissions, determine whether the current identity has access permissions for the specified fragment requested in this transaction.

[0047] In the security protection steps, it is determined whether the current identity has the access rights to the specified fragment of this request, including: when the identity token or digital certificate of the accessing subject fails the legitimacy and validity verification, it is determined as illegal access and the request link is blocked; when the identity token or digital certificate of the accessing subject passes the legitimacy and validity verification, the preset access control policy table is further queried to determine whether the identity of the accessing subject has the access rights.

[0048] In the security protection steps, the preset access control policy table is further queried to determine whether the identity of the access subject has query permissions. This includes: if the access subject has file-level access permissions but has not configured the independent permissions for the corresponding fragment, it is determined as illegal access and the request link is blocked; if the access subject's identity is legitimate and the fragment permissions are matched bidirectionally and pass the double verification condition, the data access request is allowed.

[0049] For example, by prioritizing the verification of the authenticity and validity of the access subject's token and certificate, illegal access behaviors such as identity forgery and expired permissions are blocked. On the basis of successful identity verification, permissions are further verified in layers based on the preset access control policy table, which realizes differentiated control of data access permissions, reduces the risk of data leakage, and improves the security level of data access and the accuracy of permission control.

[0050] In summary, the distributed server encrypted storage security protection and data processing method of this application, by deploying a lightweight monitoring agent on each storage server node of the distributed cluster, can comprehensively collect relevant data from each storage server node, thereby providing secure data support for subsequent node load analysis and storage scheduling decisions, improving the overall resource utilization and operational stability of the distributed cluster. At the same time, by using the HKDF algorithm and combining multi-dimensional feature parameters, each data shard is encrypted with an independent dedicated key, coupled with a secure processing mechanism for dedicated offline key ciphertext distribution and immediate erasure of plaintext keys in memory, thereby improving the data storage encryption security level.

[0051] In the description of this specification, the terms "first" and "second" are used for descriptive purposes only and should not be construed as indicating or implying relative importance or implicitly specifying the number of indicated technical features. Thus, a feature defined as "first" or "second" may explicitly or implicitly include at least one of that feature. In the description of this application, "multiple" means at least two, such as two, three, etc., unless otherwise explicitly specified.

[0052] In the description of this specification, the references to terms such as "one embodiment," "some embodiments," "example," "specific example," or "some examples," etc., refer to specific features, structures, materials, or characteristics described in connection with that embodiment or example, which are included in at least one embodiment or example of this application. In this specification, the illustrative expressions of the above terms do not necessarily refer to the same embodiment or example. Furthermore, the specific features, structures, materials, or characteristics described may be combined in any suitable manner in one or more embodiments or examples. Moreover, without contradiction, those skilled in the art can combine and integrate the different embodiments or examples described in this specification, as well as the features of different embodiments or examples.

[0053] Although embodiments of this application have been shown and described above, it is understood that the above embodiments are exemplary and should not be construed as limiting this application. Those skilled in the art can make changes, modifications, substitutions and variations to the above embodiments within the scope of this application.

Claims

1. A method for encrypted storage security protection and data processing of a distributed server, characterized in that, Includes the following steps: Deploy monitoring agents: Deploy lightweight monitoring agents on each storage server node of the distributed cluster to collect multi-dimensional performance metrics of each storage server node in real time. Data collection and transmission: Establish a regular data reporting link to transmit the collected multi-dimensional performance indicators and the current node's own identifier to the cluster's management master node; Data preprocessing: Before the data is weighted, it is cleaned, including verification, correction of outliers and filling in missing values; Deep data analysis: The weighted scoring formula is used to calculate the comprehensive load score of each effective node, identify the comprehensive load scores that are less than the preset global load threshold, and construct a low-load storage candidate pool. Data encryption preprocessing: Before storing data, a key distribution request message is sent to an independent key management server. Through the HKDF hierarchical key derivation algorithm, a unique encryption key is generated for each data fragment and transmitted to the data writing client or cluster gateway through an encrypted dedicated line. Discrete storage scheduling: The node with the lowest overall load score in the low-load storage candidate pool is selected as the target write node for the current fragment. Multiple fragments of the same file are stored on different nodes through a constraint scheduling algorithm. The encrypted ciphertext fragments and corresponding metadata are synchronously transmitted to the corresponding low-load storage nodes. Security protection: Based on access permissions, determine whether the current identity has access permissions for the specified fragment requested in this transaction.

2. The distributed server encrypted storage security protection and data processing method according to claim 1, characterized in that, In the deployment of the monitoring agent, the real-time collection of multi-dimensional performance indicators of each storage server node includes CPU utilization, available memory capacity, disk I / O operations per second, network bandwidth inbound / outbound rate, and the percentage of remaining space on the storage volume.

3. The distributed server encrypted storage security protection and data processing method according to claim 2, characterized in that, In the data acquisition and transmission steps, establishing a normalized data reporting link includes building a Grip streaming communication interface based on the HTTP / 2 protocol, establishing a long connection channel between the monitoring agent and the management master node, and encrypting the long connection channel end-to-end through the TLS 1.3 security protocol to generate an encrypted transmission layer.

4. The distributed server encrypted storage security protection and data processing method according to claim 3, characterized in that, In the data preprocessing step, data cleaning includes a sliding time window algorithm and a box plot outlier detection algorithm. Sliding time window algorithm: For time series performance indicators, a continuous observation window is constructed, and time series interpolation is used to complete and repair missing indicators; Box plot outlier detection algorithm: By calculating the quartiles and interquartile ranges of the data, it defines the reasonable distribution range of the values, identifies and removes extreme outlier data points that deviate from three standard deviations.

5. A distributed server encrypted storage security protection and data processing method according to claim 4, characterized in that, In the deep data analysis step, the calculation of the comprehensive load score for each effective node using a weighted scoring formula includes the following steps: Feature substitution: Substitute the cleaned multidimensional performance indicators into the preset weighted scoring formula for linear weighted fusion calculation; Load score calculation: The corresponding real-time comprehensive load score is obtained through the above multiplication and addition operations. The value of the comprehensive load score is positively correlated with the real-time load pressure of the node. The lower the score, the higher the proportion of idle resources of the node and the greater the resource redundancy that can support storage tasks. Low-load candidate pool construction: Set a load threshold, include nodes with a comprehensive load score less than the global load threshold into the initial selection set, and remove offline and abnormal faulty nodes based on the node heartbeat detection results to construct a low-load storage candidate pool.

6. A distributed server encrypted storage security protection and data processing method according to claim 5, characterized in that, In the data encryption preprocessing step, the generation of a unique encryption key for each data fragment using the HKDF hierarchical key derivation algorithm includes the following steps: Parameter initialization: The key management server receives the encryption request message sent by the cluster management master node and extracts three dimensions of information from the request message; Layered Iterative Derivation: Based on the HKDF algorithm, multiple hash iterations and diffusion operations are performed using salted hash calculations to eliminate key association and generate a unique encryption key for each data fragment. Secure key distribution: The exclusive encrypted key in ciphertext form is distributed to the data writing client or cluster gateway via an encrypted dedicated line, and the master key material and intermediate master key plaintext traces in memory are erased after distribution.

7. A distributed server encrypted storage security protection and data processing method according to claim 6, characterized in that, The three dimensions of information in the extraction request message include a globally unique file identifier, the file splitting sequence number, and the target storage node for the encrypted fragment to be written.

8. A distributed server encrypted storage security protection and data processing method according to claim 7, characterized in that, In the discrete storage scheduling process, storing multiple fragments of the same file on different nodes using a constrained scheduling algorithm includes the following steps: Target node selection: Based on the low-load storage candidate pool, obtain the real-time comprehensive load score of each candidate node, sort them in ascending order of score, and prioritize the node with the lowest score and the highest resource redundancy as the first choice target node for the current shard to be written. Constraint verification: Construct a set of occupied nodes for the current file, compare the identifier of the preferred target node with the set of occupied node identifiers, and select the physical node that does not exist in the set; Fragmented Discrete Write: Under the dual constraints of load optimization and node isolation, the ciphertext is fragmented and the corresponding metadata is transmitted and written to the selected target storage node. Iterative update of scheduling status: The selected target storage node is added to the set of occupied node identifiers, and the real-time comprehensive load score of the corresponding node in the low-load storage candidate pool is refreshed simultaneously to provide the latest node status data for the constraint scheduling of the next shard, thereby realizing iterative optimization of the cluster-wide load.

9. A distributed server encrypted storage security protection and data processing method according to claim 8, characterized in that, The step of determining whether the current identity has access rights to the specified fragment requested in this transaction includes: When the identity token or digital certificate of the accessing entity fails the legitimacy and validity verification, it is determined as an illegal access and the request link is blocked; When the identity token or digital certificate of the accessing subject passes the legitimacy and validity verification, the system further queries the preset access control policy table to determine whether the accessing subject's identity identifier has the necessary access permissions.

10. A distributed server encrypted storage security protection and data processing method according to claim 9, characterized in that, The further querying of the preset access control policy table to determine whether the access subject's identity has query permissions includes: If the accessing entity has file-level access permissions but has not configured independent permissions for the corresponding fragment, it is determined as illegal access and the request link is blocked; The data access request is allowed when the identity of the accessing entity is legitimate and the two-way matching of fragment permissions passes the dual verification condition.

Images