Data center node synchronous evaluation method and system based on block chain

Abstract

The invention relates to a block chain-based data center node synchronization evaluation method and system, and the method comprises the following steps: S1, randomly selecting a trusted node by each data center node, and calculating an expected synchronization progress in combination with synchronization information; s2, sorting is carried out according to interaction characteristics of untrusted nodes, nodes meeting conditions are selected to generate a synchronization request tuple containing a random synchronization sequence and an expected synchronization progress, and the synchronization request tuple is sent to the selected nodes; s3, receiving a synchronization response and a synchronization progress tuple of an untrusted node, detecting a data synchronization state through statistical analysis of a synchronization state and block confirmation time, generating a domain synchronization tuple, and performing whole network broadcasting; and S4, receiving the whole-network domain synchronization tuple, and judging a whole-network synchronization completion state according to the number of the synchronized domains and the whole-network synchronization progress index. According to the method, the system nodes are divided into the trusted domain and the non-trusted domain, the total data communication amount is reduced, and the overall efficiency is improved; and meanwhile, the nodes perform parallel and synchronous calculation, so that the overall efficiency of the system is improved.

Description

Technical Field

[0001] This invention relates to the field of blockchain technology, specifically to a blockchain-based data center node synchronization evaluation method and system, which improves the efficiency and reliability of data synchronization between data center nodes. Background Technology

[0002] With the widespread application of new information technologies such as 5G, the Internet of Things, and artificial intelligence, the demand for real-time storage and processing of massive amounts of data is increasing daily. Data centers, especially large-scale data centers, are increasingly becoming the fundamental support infrastructure for information technology transformation. Data centers are often composed of tens of thousands of nodes, and ensuring that these nodes work in a coordinated manner is one of the prerequisites for effectively utilizing the computing power of a data center. As data centers grow larger and larger, it is necessary to research an algorithm to evaluate the synchronization progress between nodes, so as to adopt appropriate computing algorithms based on the synchronization status, thereby maximizing the computing power efficiency of the data center.

[0003] Currently, the commonly used algorithm involves each node sending a synchronization status request to the management node when it has a data synchronization request. The management node then sends a synchronization status confirmation to all nodes in the network and returns the collected synchronization status to the requesting node. This method is simple and stable, but its efficiency decreases as the number of nodes increases, and it also carries the risk of single points of failure. Therefore, there is a need to research a synchronization progress evaluation method suitable for large-scale, especially distributed, data centers.

[0004] In large-scale distributed data centers, each node often interacts with only a small number of nodes, thus each node has its own set of trusted and untrusted nodes. Furthermore, by employing a blockchain data structure for data storage and update recording, each data update operation is recorded in the latest block. As the number of blocks continuously increases, when the data of all nodes across the network is synchronized to the latest state, each node's blocks and their number are identical. Summary of the Invention

[0005] The purpose of this invention is to provide a blockchain-based method for synchronizing and evaluating data center nodes, comprising the following steps:

[0006] Step S1: Each arbitrary data center node in the system randomly selects N trusted nodes in its trusted domain, reads the latest block sequence number of itself and the trusted nodes, and calculates the expected synchronization progress based on the synchronization information of itself and the trusted nodes.

[0007] Step S2: The data center node sorts the untrusted nodes according to the interaction characteristics of the M untrusted nodes in its untrusted domain, selects P untrusted nodes from them according to preset conditions, generates a synchronization request tuple including a random synchronization sequence and the calculated expected synchronization progress, and sends it to the selected untrusted nodes.

[0008] Step S3: Each selected untrusted node returns a synchronization response tuple and a synchronization progress tuple to the corresponding data center node. The corresponding data center node detects the data synchronization status by statistically analyzing the synchronization status and block confirmation time in the tuple, generates the domain synchronization tuple of the data center node, and broadcasts it to the entire network.

[0009] Step S4: Each data center node receives all domain synchronization tuples broadcast across the entire network. If the number of domains that have completed data synchronization exceeds the default value and the network-wide synchronization progress indicator exceeds the default value, then the network-wide synchronization is complete; otherwise, the network-wide synchronization is incomplete.

[0010] Furthermore, the domain of each arbitrary data center node A within the system includes all trusted nodes and all untrusted nodes of node A, wherein all trusted nodes constitute the trusted domain of node A, and all untrusted nodes constitute the untrusted domain of node A.

[0011] Furthermore, in step S1, each arbitrary data center node A within the system reads the sequence number idx of its latest local block at a preset time interval or when there is a data consistency verification requirement. A It randomly selects N trusted nodes from its trusted domain, and then sends requests to the N trusted nodes to obtain the latest local block sequence number idx of each node. u And wait for all nodes to return the sequence number of their latest block, where u∈[1,N]. The expected synchronization progress of node A's trusted domain is:

[0012]

[0013] Where α>β>0 is a default constant and α+β=1.

[0014] Further, step S2 includes the following steps:

[0015] Step S201: Each data center node A in the system contains M untrusted nodes within its untrusted domain, where the total amount of data interaction between the v-th untrusted node and node A is y. v The average data transmission rate between node A and node Z is z. v Its network activity in the untrusted domain of node A is defined as:

[0016] δ v =ζ*yv +η*z v

[0017] Where v∈[1,M], ζ,η>0 are default constants;

[0018] Step S202: Sort the M untrusted nodes in ascending order according to network activity, select the first, middle and last nodes in the sort, and then select a preset number of nodes from the remaining M-3 untrusted nodes, so that the number of selected nodes and their total network activity are both within a preset range.

[0019] Step S203: Let the number of selected untrusted nodes be P, and node A generates a random synchronization sequence. Each sequence element is a positive integer, and the synchronization request tuple will be included. Send to the corresponding k-th selected untrusted node.

[0020] Further, step S3 includes the following steps:

[0021] Step S301: When the k-th non-trusted node in [1,P] receives the corresponding synchronization request tuple, generate a synchronization response tuple. And send it to node A, where P is the number of untrusted nodes selected within the domain of each arbitrary data center node A in the system. These represent the k-th untrusted node's... The first block and the first Each block, For the k-th untrusted node Confirmation time of each block, For the k-th untrusted node The data status related to node A on each block, the synchronization response tuple also includes the synchronization progress tuple.

[0022] Step S302: Each arbitrary data center node A in the system performs statistical analysis on the tuple data returned from its P untrusted nodes. If the synchronization progress tuples are consistent with the local data of node A, and the confirmation time of the P untrusted nodes is... variance Δ A If the value is less than a preset constant, then the domain of node A is determined to have achieved data synchronization; otherwise, it is determined to have not achieved data synchronization.

[0023] Step S303: Each arbitrary data center node A within the system synchronizes its domain tuple. Broadcast across the entire network, where L is the synchronization result flag, L=1 indicates data synchronization is complete, and L=-1 indicates data synchronization is incomplete.

[0024] Further, in step S4, each arbitrary data center node A within the system receives all domain synchronization tuples broadcast across the entire network. If the number of domains that have completed data synchronization exceeds a preset value, and the network-wide synchronization progress indicator G exceeds a default value, then the network-wide synchronization is determined to be complete; otherwise, the network-wide synchronization is not complete. The network-wide synchronization progress indicator G is defined as follows:

[0025]

[0026] Where λ>0 is a preset constant.

[0027] Furthermore, the present invention adopts the following technical solution:

[0028] A blockchain-based data center node synchronization evaluation system includes the following modules:

[0029] Expected synchronization calculation module: used to randomly select multiple trusted nodes within each data center node, read the latest block sequence number of itself and the trusted nodes, and calculate the expected synchronization progress;

[0030] Untrusted Node Selection Module: This module sorts untrusted nodes within each data center node based on their interaction characteristics, selects multiple untrusted nodes that meet preset conditions, generates a random synchronization request tuple, and sends it to the selected untrusted nodes.

[0031] Synchronization Response Processing Module: Used to receive synchronization response tuples returned by untrusted nodes in each data center node, perform statistical analysis on the synchronization progress tuples, and determine the data synchronization status within the domain;

[0032] The network-wide synchronization judgment module is used to receive the domain synchronization tuples broadcast across the entire network in each data center node, and to determine whether the network-wide synchronization status has reached the preset value based on the number of domains that have completed data synchronization and the network-wide synchronization progress indicators.

[0033] The beneficial effects of this invention include:

[0034] 1. Divide system nodes into trusted and untrusted domains to reduce the total amount of system data communication and improve overall efficiency.

[0035] 2. By having each node calculate the synchronization progress of its domain in parallel, the utilization rate of system communication resources can be improved, thereby increasing overall efficiency. Attached Figure Description

[0036] Figure 1 This is a schematic diagram of the logical structure of the data center node in this invention.

[0037] Figure 2This is a schematic diagram of the data center node synchronization evaluation process of the present invention. Detailed Implementation

[0038] The present invention will be further described and illustrated below with reference to the accompanying drawings.

[0039] In a preferred embodiment of the present invention, such as Figure 1 As shown, the blockchain-based data center node synchronization evaluation method is applied in a system with 100 data center nodes. Each node can independently execute the following steps to evaluate the data synchronization status, such as... Figure 2 As shown, it includes the following steps:

[0040] Step S1, expected to be calculated synchronously:

[0041] In each arbitrary data center node A within the system, 10 trusted nodes (i.e., N=10) are first randomly selected from its trusted domain, and the latest block sequence numbers of these trusted nodes are read. Node A's own latest block sequence number is 105, while the latest block sequence numbers of the 10 selected trusted nodes are: 106, 104, 105, 107, 105, 106, 105, 104, 106, and 107, respectively.

[0042] Node A, using its own block number, calculates the expected synchronization progress according to the formula.

[0043]

[0044] Where α = 0.7 and β = 0.3, the calculation result is:

[0045]

[0046] According to the calculation results, the expected synchronization progress of node A is 105.6.

[0047] Step S2, Selection of untrusted nodes and generation of random synchronization requests:

[0048] Node A calculates its network activity δ from 50 untrusted nodes within its domain based on a combination of data interaction volume and average transmission rate. v Among the untrusted nodes of node A, the activity levels of the top 5 nodes are: 0.85, 0.80, 0.78, 0.76, and 0.75.

[0049] Node A selects the top 5 most active untrusted nodes and generates a random synchronization request tuple. The generated synchronization request tuple is: (2,105),(3,106),(1,104),(2,105),(3,107).

[0050] Step S3, Synchronous Response and Data Analysis:

[0051] Node A receives the synchronization response tuple returned by the untrusted node. The synchronization response tuple returned by the first untrusted node is (B 105,1 B 103,1 ,T 105,1 C 105,1 ), where B 105,1 This indicates the 105th block of the node, T. 105,1 For block confirmation time, C 105,1 This is the data status.

[0052] Node A performs statistical analysis on all returned response tuples and determines the synchronization status. If all synchronization progress tuples are consistent with Node A's local data, and the variance Δ of all acknowledgment times is within a certain range... A If the value is less than the preset value of 0.05, then node A is considered to have achieved data synchronization within the domain.

[0053] Step S4, Simultaneous evaluation across the entire network:

[0054] Node A receives all domain synchronization tuples broadcast across the entire network and calculates the network-wide synchronization progress index G. Let λ = 0.1, according to the formula:

[0055]

[0056] The calculation result is G = 10.5. If the number of synchronized domains exceeds 60% and G exceeds the default value of 10, then the entire network is considered synchronized.

[0057] This invention also includes a blockchain-based data center node synchronization evaluation system, comprising:

[0058] Expected synchronization calculation module: In each data center node, according to step S1, N trusted nodes are randomly selected, the latest block sequence number of itself and the trusted nodes is read, and the expected synchronization progress is calculated.

[0059] Untrusted Node Selection Module: In each data center node, according to step S2, the untrusted nodes are sorted based on their interaction characteristics, and P untrusted nodes that meet the preset conditions are selected. A random synchronization request tuple is generated and sent to the selected untrusted nodes.

[0060] Synchronization response processing module: In each data center node, according to step S3, it receives the synchronization response tuple returned by the untrusted node, performs statistical analysis on the synchronization progress tuple, and determines the data synchronization status within the domain.

[0061] The network-wide synchronization judgment module is used to receive the domain synchronization tuple broadcast across the network in each data center node according to step S4, and to determine whether the network-wide synchronization status has reached the preset value based on the number of domains that have completed data synchronization and the network-wide synchronization progress index.

[0062] The above description is only a preferred embodiment of the present invention and is not intended to limit the present invention. Any modifications, equivalent substitutions, improvements, etc., made within the spirit and principles of the present invention should be included within the protection scope of the present invention.

Claims

1. A blockchain-based data center node synchronization evaluation method, characterized in that, Includes the following steps: Step S1: Each arbitrary data center node in the system randomly selects N trusted nodes in its trusted domain, reads the latest block sequence number of itself and the trusted nodes, and calculates the expected synchronization progress based on the synchronization information of itself and the trusted nodes. Step S2: The data center node sorts the untrusted nodes according to the interaction characteristics of the M untrusted nodes in its untrusted domain, selects P untrusted nodes from them according to preset conditions, generates a synchronization request tuple including a random synchronization sequence and the calculated expected synchronization progress, and sends it to the selected untrusted nodes. Step S3: Each selected untrusted node returns a synchronization response tuple and a synchronization progress tuple to the corresponding data center node. The corresponding data center node detects the data synchronization status by statistically analyzing the synchronization status and block confirmation time in the tuple, generates the domain synchronization tuple of the data center node, and broadcasts it to the entire network. Step S4: Each data center node receives all domain synchronization tuples broadcast across the entire network. If the number of domains that have completed data synchronization exceeds the default value and the network-wide synchronization progress indicator exceeds the default value, then the network-wide synchronization is complete; otherwise, the network-wide synchronization is incomplete.

2. The blockchain-based data center node synchronization evaluation method as described in claim 1, characterized in that, Each arbitrary data center node A within the system contains all trusted nodes and all untrusted nodes of node A within its domain, wherein all trusted nodes constitute the trusted domain of node A, and all untrusted nodes constitute the untrusted domain of node A.

3. The blockchain-based data center node synchronization evaluation method as described in claim 1, characterized in that, In step S1, each arbitrary data center node A within the system reads the sequence number idx of its latest local block at a preset time interval or when there is a data consistency verification requirement. A It randomly selects N trusted nodes from its trusted domain, and then sends requests to the N trusted nodes to obtain the latest local block sequence number idx of each node. u And wait for all nodes to return the sequence number of their latest block, where u∈[1,N], the expected synchronization progress of node A's trusted domain is: Where α>β>0 is a default constant and α+β=1.

4. The blockchain-based data center node synchronization evaluation method as described in claim 1, characterized in that, Step S2 includes the following steps: Step S201: Each data center node A in the system contains M untrusted nodes within its untrusted domain, where the total amount of data interaction between the v-th untrusted node and node A is y. v The average data transmission rate between node A and node Z is z. v Its network activity in the untrusted domain of node A is defined as: d v =ζ*y v +η*z v Where v∈[1,M], ζ, η>0 are default constants; Step S202: Sort the M untrusted nodes in ascending order according to network activity, select the first, middle and last nodes in the sort, and then select a preset number of nodes from the remaining M-3 untrusted nodes, so that the number of selected nodes and their total network activity are both within a preset range. Step S203: Let the number of selected untrusted nodes be P, and node A generates a random synchronization sequence. Each sequence element is a positive integer, and the synchronization request tuple will be included. Send to the corresponding k-th selected untrusted node.

5. The blockchain-based data center node synchronization evaluation method as described in claim 1, characterized in that, Step S3 includes the following steps: Step S301: When the k-th non-trusted node in [1,P] receives the corresponding synchronization request tuple, generate a synchronization response tuple. And send it to node A, where P is the number of untrusted nodes selected within the domain of each arbitrary data center node A in the system. These represent the k-th untrusted node's... The first block and the first Each block, For the k-th untrusted node Confirmation time of each block, For the k-th untrusted node The data status related to node A on each block, the synchronization response tuple also includes the synchronization progress tuple. Step S302: Each arbitrary data center node A in the system performs statistical analysis on the tuple data returned from its P untrusted nodes. If the synchronization progress tuples are consistent with the local data of node A, and the confirmation time of the P untrusted nodes is... variance Δ A If the value is less than a preset constant, then the domain of node A is determined to have achieved data synchronization; otherwise, it is determined to have not achieved data synchronization. Step S303: Each arbitrary data center node A within the system synchronizes its domain tuple. Broadcast across the entire network, where L is the synchronization result flag, L=1 indicates data synchronization is complete, and L=-1 indicates data synchronization is incomplete.

6. The blockchain-based data center node synchronization evaluation method as described in claim 1, characterized in that, In step S4, each arbitrary data center node A within the system receives all domain synchronization tuples broadcast across the entire network. If the number of domains that have completed data synchronization exceeds a preset value, and the network-wide synchronization progress indicator G exceeds the default value, then the network-wide synchronization is considered complete; otherwise, the network-wide synchronization is incomplete. The network-wide synchronization progress indicator G is defined as follows: Where λ>0 is a preset constant.

7. A blockchain-based data center node synchronization evaluation system, characterized in that, Includes the following modules: Expected synchronization calculation module: used to randomly select multiple trusted nodes within each data center node, read the latest block sequence number of itself and the trusted nodes, and calculate the expected synchronization progress; Untrusted Node Selection Module: This module sorts untrusted nodes within each data center node based on their interaction characteristics, selects multiple untrusted nodes that meet preset conditions, generates a random synchronization request tuple, and sends it to the selected untrusted nodes. Synchronization Response Processing Module: Used to receive synchronization response tuples returned by untrusted nodes in each data center node, perform statistical analysis on the synchronization progress tuples, and determine the data synchronization status within the domain; The network-wide synchronization judgment module is used to receive the domain synchronization tuples broadcast across the entire network in each data center node, and to determine whether the network-wide synchronization status has reached the preset value based on the number of domains that have completed data synchronization and the network-wide synchronization progress indicators.

Images