A regularized dispersion and convergence method for temporal network datasets

Abstract

The present invention discloses a method for regular dispersion and collection of a time series network data set, and the steps include: 1) encapsulating each data sequence in a data structure for storing a distributed time series data set according to a corresponding encapsulation algorithm rule; the data block of the encapsulated data sequence at least includes the type of the data block, the mark of the data sequence where the data block is located, the encapsulation time sequence mark of the data block, and the specific business data load in the data block; 2) when the time series data set continues to grow based on a specific business scenario and reaches a dispersion trigger condition, the data sequence By is split and separated into a data sequence Bys; 3) when the time series data set continues to grow based on a specific business scenario and reaches a collection trigger condition, the data sequence Bys is aggregated to the data sequence By and merged into the data sequence By. The present invention can realize the rapid and clear dispersion of the newly added data set, thereby realizing flexible processing of the data set.

Description

Technical Field

[0001] The present invention relates to the fields of computer technology and the Internet, and in particular to a method for regularized dispersion and collection of network data sets in an Internet environment. Background Art

[0002] With the continuous development of the Internet era, the growth of the volume of data sets over time under network-type data structures such as time series databases, blockchains, and edge computing, and the resources required for reading, transmitting, and operating data sets have become a bottleneck for node processing and storage resources in more and more distributed computing systems.

[0003] Currently, the growth in the size of data blocks of this type of distributed data set can be addressed by reducing packaging time, increasing packaging costs, and other processing methods, thereby reducing the size of the data set that needs to be packaged, thereby reducing the costs of subsequent data set transmission and processing. However, reducing packaging time may lead to the risk of insufficient data consistency, while increasing costs will lead to high project operating costs.

[0004] Therefore, the present invention proposes a method for regular dispersion and convergence of time series network data sets in a growing state for the Internet, which improves the flexibility and convenience of data storage of time series data in a distributed network state under the Internet, thereby reducing the single-point processing load of the distributed system. Summary of the invention

[0005] In view of the technical problems existing in the prior art, the purpose of the present invention is to provide a method for regular dispersion and collection of storage of time series network data sets for the Internet during their growth process. The method is suitable for applications where single-point loads are adjusted by splitting, collecting, storing and processing all data sets. The method is a method for associated dispersed storage of time series data sets used in public occasions.

[0006] In a given computer network, the time series network data set consists of several parallel data block sequences. The data sequence is composed of data blocks that are sequentially associated using feature values. The data blocks are composed of Figure 1 The data structure shown in the figure is composed of Figure 2 When the data block size of a data sequence exceeds the set capacity safety threshold or the business setting needs, the data sequence is split into even-numbered data sequences according to the parity characteristics of the eigenvalues ​​or other agreed rules, and the business data between the data sequences complement each other, such as Figure 5 and Figure 6 As shown; when the data block of the dispersed data sequence is lower than the set capacity threshold or the business setting needs, the dispersed data is merged and returned to the previous data sequence, such as Figure 4 .

[0007] The method for regular dispersion and collection of a time series data set of the present invention comprises the following steps:

[0008] 1) Select the corresponding encapsulation algorithm rules for each data sequence B1, ~, BN in the data structure used to store distributed time series data sets according to different business scenarios or other set criteria. It can be zip format compression, encryption algorithm or other custom algorithm rules. The data blocks in the same data sequence By are sequentially associated, y = 1 ~ N, where this association can be anchoring between data features of adjacent data blocks, or other custom association methods. The data sequences B1, ~, BN are finally encapsulated into their own data dependency structures according to their own business characteristics, such as Figure 4 The data blocks constituting the data sequence include at least four parts: Fy_Sh, F_DS, Ti_(y,x), and DATA, and may also include H_(y,(x-1)) and H_G(z-1) as the fifth and sixth parts, such as Figure 1 , where Fy_Sh is the type of data block, F_DS is the flag of the data sequence By where the data block is located (can be By, or a code or ordinal number), Ti_(y,x) is the encapsulation timing identifier of the xth data block in the yth column data sequence By, which can be the encapsulation time or other agreed identifier, DATA refers to the specific service data load in the data block, H_(y,(x-1)) is the digital characteristic value of the preceding data block (i.e., the x-1th data block) of the xth data block in the yth column data sequence By, where H_G(z-1) is the digital characteristic value of the super block (i.e., the z-1th) of the xth data block in the yth column data sequence By. Wherein, the super block H_G(z) is a data block in the super block sequence (a type of data sequence), and the data block of the super block sequence at least stores the data sequence flag F_DS of other data sequences, the data encapsulation timing Ti_(y,x), and the digital characteristic value H_G(z-1) of the preceding super block. The encapsulation and generation method of "super blocks" is similar to that of ordinary data blocks. At the same time, super blocks accept and save the digital characteristic values ​​of other data blocks. Super blocks can serve as other ordinary data blocks, or they can also be special data blocks.

[0009] The data block type Fy_Sh is used to characterize whether the data block is dispersed and the form of dispersion. The non-dispersed data sequence is treated as a conventional value. The dispersed type can be the serialization of the description parameter or the digital characteristic value of the description parameter. The description parameter organization form can be XML, JSON, YML or other custom types. The data load DATA constituting the data block includes T_1, T_2, ..., T_(M), where T_(m) (m = 1 to M) is a specific business data entry. The data entry can be a kV (i.e., key-Value) data structure or a data structure set by a specific business scenario. The data structure matches the data block type Fy_Sh.

[0010] 2) Based on the specific business scenario, the time series data set continues to grow and reaches the dispersion trigger condition (such as the data load is greater than the capacity threshold). The data sequence By then splits into the data sequence Bys. The data structure of the data sequence Bys is the same as that of the data sequence By, but the data load business data items in the data blocks of the data sequence By and the data sequence Bys are complementary sets that cooperate or depend on each other in the business scenario, and together constitute a business data set. Among them, the data sequence adopts rules to form a complementary set according to the business scenario requirements. Such rules include parity, modulus, mask and other agreed benchmark classifications of the characteristic values ​​of the data items, such as Figure 5 and Figure 6 , after taking the modulus or masking the characteristic value of the data entry, it is classified according to the specific business needs. Among them, the characteristic value of the data entry can be obtained by serialization or other custom methods, and the means of obtaining the characteristic value of the data entry is placed in the parameter structure of type Fy_Sh.

[0011] Among them, the decentralized trigger condition can be a data load capacity threshold, or other business scenario settings. Among them, DATA_ is the data load after the data sequence By is split, and DATA_s is the data load of the data sequence Bys. DATA_ is composed of T_1, T_3, ..., T_(M-1), and DATA_s is composed of T_2, T_4, ..., T_(M). T_(M) may have a specific business cooperation relationship with other data or no relationship. DATA_ and DATA_s constitute a complementary set. Among them, the key of the kV structure of the data entry T_(m) can be the digital characteristic value of Value, the encapsulation sequence or other business scenario settings.

[0012] 3) The time series data set is continuously growing based on specific business scenarios. When the global trigger condition (such as the time interval) is reached, the super block GB(z) obtains the digital feature value H_(y,x) of the data block of each data sequence. The super block data block includes Ti_G(z), DATA_G, H_(y')(y'=1~N,ys), and H_G(z-1). Among them, Ti_G(z) is the encapsulation time series identifier of the data block GB(z) of the super block sequence, DATA_G is the data load of the data block of the super block sequence, H_(y')(y'=1~N,ys) is the digital feature of the data block of the data sequence By(y'=1~N,ys), among which the data sequence Bys is the data sequence split from the data sequence By, and H_G(z-1) is the preceding data block (z-1th) of the zth data block of the super block sequence. The "data block" of the super block sequence is also called the "super block".

[0013] 4) Based on the specific business scenario, the time series data set continues to grow and reaches the trigger condition for collapsing (for example, the data load after splitting is less than the set threshold, and the data load is also the encapsulated volume of the data block of the data sequence). The data sequence Bys is merged into the data sequence By, and the data block type Fy_Sh is set to restore the undispersed type. The digital characteristic value H_s(y,(x-1)) of the preceding data block of the xth data block in the data sequence Bys is added to H_(y,(x-1)). The collapsing trigger condition can be a data load capacity threshold or other business scenario settings.

[0014] Compared with the prior art, a method for regular dispersion and convergence of network distributed time series data sets including time series databases, blockchains, etc. has the following beneficial effects:

[0015] By forming mutually related and complementary sets, the present invention can realize the rapid and clear dispersion of newly added data sets, thereby realizing flexible processing of data sets; it is suitable for the dispersion and aggregation of data sets in public network environments, including blockchain, cloud computing, edge computing and other environments, and is especially applicable to scenarios in various distributed data growth environments.

[0016] Of course, any product implementing the present invention does not necessarily need to achieve all the advantages described above at the same time. The parts not involved in the device are the same as the prior art or can be implemented by the prior art. The present invention has a simple structure and is easy to operate. BRIEF DESCRIPTION OF THE DRAWINGS

[0017] Figure 1 Schematic diagram of the data structure I of the data set proposed in the present invention.

[0018] Figure 2 Schematic diagram of data structure II of the data set proposed in the present invention.

[0019] Figure 3 Schematic diagram of the data structure III of the data set proposed in the present invention.

[0020] Figure 4 Schematic diagram of the data structure IV of the data set proposed in the present invention.

[0021] Figure 5 Schematic diagram of the data structure V of the data set proposed in the present invention.

[0022] Figure 6 Schematic diagram of the data structure VI of the data set proposed in the present invention. DETAILED DESCRIPTION

[0023] The technical solutions in the embodiments of the present invention will be described clearly and completely below in conjunction with the drawings in the embodiments of the present invention. Obviously, the described embodiments are only part of the embodiments of the present invention, rather than all the embodiments.

[0024] Specific implementation case 1, applied to parallel blockchain, is as follows:

[0025] In a blockchain computer network, the parallel blockchain consists of several parallel blockchain sub-chains. The blockchain is composed of data blocks anchored in sequence using characteristic values. The data blocks are composed of Figure 1 The data structure shown in the figure is composed of Figure 2 When the data block of the blockchain exceeds the set capacity security threshold or the business setting needs, the blockchain where the data block is located will be split into two blockchain sub-chains, and the business data of the two blockchains will complement each other, such as Figure 5 and Figure 6 As shown; when the data block of the split blockchain is lower than the set capacity threshold or the business setting requires, the dispersed data is merged and returned to the previous blockchain, such as Figure 4 .

[0026] The method for regularizing dispersion and collection of a time series data set in an implementation example of the present invention comprises the following steps:

[0027] 1) Encapsulate each data sequence B1, ~, BN in the data structure according to the specific business scenario, and use the hash value to anchor the data blocks in the same data sequence By in sequence, y = 1 ~ N. The data sequences B1, ~, BN are finally encapsulated into their own data dependency structures according to their own business characteristics, such as Figure 4 As shown. The data blocks that constitute the data sequence include 6 parts: Fy_Sh, F_DS, Ti_(y,x), DATA, H_(y,(x-1)), and H_G(z-1). Figure 1 , where Fy_Sh is the type of data block, F_DS is the flag of the data sequence By where the data block is located, Ti_(y,x) is the encapsulation time of the data block, DATA refers to the specific business data load carried by the data block, H_(y,(x-1)) is the digital characteristic value of the preceding data block (i.e., the x-1th) of the x-th data block in the y-th column, and H_G(z-1) is the digital characteristic value of the dependent super block (i.e., the z-1th) of the x-th data block in the y-th column.

[0028] Among them, the data block type Fy_Sh represents that the data block adopts a dispersed and dispersed form, and the description parameter organization form adopts JSON. The data load DATA constituting the data block includes T_1, T_2, ..., T_(M), where T_(m) (m=1~M) is a specific business data entry, and the data entry adopts a kV (key-Value) data structure, and the data structure matches the type Fy_Sh.

[0029] The json format of Fy_Sh in this case is described as follows:

[0030]

[0031] Among them, the TyofB key value indicates whether the current block is in a decentralized state, On means it is in a decentralized state, otherwise it is off; the kV key value indicates whether the key-value data structure is adopted, On means it is adopted.

[0032] 2) As the blockchain continues to grow, if the data load of the data block of the blockchain carrying the business exceeds the pre-set 2MB capacity threshold, the blockchain system will trigger the blockchain to be decentralized, and the data sequence By will then split into the data sequence Bys, and the data structure of the data sequence Bys is the same as that of By. The data items of the data load business data in the data blocks of the data sequence By and the data sequence Bys form a complementary set of interdependent relationships, and the data items of the two data sequences are classified based on the odd or even value of the rightmost bit of the characteristic value, such as Figure 5 and Figure 6 .

[0033] The characteristic value of the data entry may be obtained by serialization, and the means for obtaining the characteristic value of the data entry is placed in a parameter structure of type Fy_Sh.

[0034] The json format of the sequences By and Bys in this case is described as follows:

[0035]

[0036] Among them, the TyofB key value indicates whether the current block is in a decentralized state, On means it is in a decentralized state, otherwise it is off; the NofSh key value indicates the number of sequences after dispersion, which is currently 2; the TyofSh key value indicates the basis for dispersion, eo indicates the parity of the last right bit of the data entry index value; the kV key value indicates whether the key-value data structure is used, On means it is used.

[0037] Among them, DATA_ is the data payload after the data sequence By is split, and DATA_s is the data payload of the data sequence Bys. DATA_ is composed of T_1, T_2, ..., T_(M-1), and DATA_s is composed of T_2, T_4, ..., T_(M). DATA_ and DATA_s form a complementary set. Among them, the key of the kV structure of the data entry T_(m) adopts the digital characteristic value of Value.

[0038] 3) As the parallel blockchain continues to extend, a global response is triggered after a pre-set 10-minute time interval, and the super block GB(z) obtains the data digital feature value H_(y,x) of each data sequence. The super block data block includes Ti_G(z), DATA_G, H_(y) (y=1~N,ys), and H_G(z-1), where Ti_G(z) is the encapsulation time of the data block GB(z) of the super block sequence, DATA_G is the data load of the data block of the super block sequence, H_(y) (y=1~N,ys) is the data block digital feature of the data sequence By(y=1~N,ys), where the data sequence Bys is the data sequence split from the data sequence By, and H_G(z-1) is the preceding data block (z-1th) of the zth data block of the super block sequence.

[0039] 4) The decentralized blockchain carrying the business triggers convergence when the load of two consecutive data blocks is less than 50kB, and the data sequence Bys is aggregated to the data sequence By to form the data sequence By.

[0040] The sequence By and Bys here Fy_Sh's json format is re-described as follows:

[0041]

[0042] Specific implementation case 2, applied to serial blockchain, is as follows:

[0043] In a time-series network of computer networks in a serial blockchain, the blockchain is composed of data blocks that are sequentially linked using feature values. The data blocks are composed of Figure 1 The data structure shown in the figure is composed of Figure 2 When the data block of the blockchain exceeds the set capacity security threshold or the business setting needs, the cross-chain will be split into two blockchain sub-chains, and the business data of the two blockchains will complement each other, such as Figure 5 and Figure 6 As shown; when the data block of the split blockchain is lower than the set capacity threshold or the business setting needs, the decentralized blockchain is merged and returned to the previous data sequence, such as Figure 4 .

[0044] The method for regular dispersion and collection of a time series data set in an implementation example of the present invention specifically includes the following steps:

[0045] 1) Encapsulate each data sequence B1, ~, BN in the data structure according to the specific business scenario, and associate the data blocks in the same data sequence By in chronological order, y = 1 ~ N, such as Figure 4 Among them, the data blocks that constitute the data sequence include 5 parts: Fy_Sh, F_DS, Ti_(y,x), DATA, and H_(y,(x-1)). Figure 1 , where Fy_Sh is the type of data block, F_DS is the flag of the data sequence By where the data block is located, Ti_(y,x) is the encapsulation time number of the data block, DATA is the business data load specifically carried by the data block, and H_(y,(x-1)) is the digital characteristic value of the preceding data block (i.e., the x-1th) of the xth data block in the yth column.

[0046] The data block type Fy_Sh indicates that the data block adopts a dispersed and dispersed form, and the description parameter organization form adopts the YML format. The data load DATA constituting the data block includes T_1, T_2, ..., T_(M), where T_(m) (m=1~M) is a specific business data entry, and the data entry adopts a kV (key-Value) data structure, and the data structure matches the type Fy_Sh.

[0047] The yml format of Fy_Sh in this case is described as follows:

[0048]

[0049] Among them, the TyofB key value indicates whether the current block is in a decentralized state, On means it is in a decentralized state, otherwise it is off; the kV key value indicates whether the key-value data structure is adopted, On means it is adopted.

[0050] 2) When the data load of a blockchain data block exceeds the preset 2MB capacity threshold, the blockchain system will trigger the blockchain to be dispersed. The data sequence By will then split into the data sequence Bys. The data structure of the data sequence Bys is the same as that of By, but the data load business data items in the data blocks of the data sequences By and Bys are complementary sets. The data items of the two data sequences are classified based on the odd or even eigenvalues, such as Figure 5 and Figure 6 The characteristic value of the data entry can be obtained by serialization or other custom methods, and the method for obtaining the characteristic value of the data entry is placed in the parameter structure of type Fy_Sh.

[0051] The YML format of the sequences By and Bys in this case is described as follows:

[0052]

[0053] Among them, the TyofB key value indicates whether the current block is in a decentralized state, On means it is in a decentralized state, otherwise it is off; the NofSh key value indicates the number of sequences after dispersion, which is currently 2; the TyofSh key value indicates the basis for dispersion, eo indicates the parity of the last right bit of the data entry index value; the kV key value indicates whether the key-value data structure is used, On means it is used.

[0054] The dispersion trigger condition can be a data load capacity threshold or other business scenario settings. DATA_ is the data load after the data sequence By is split, and DATA_s is the data load of the data sequence Bys. DATA_ is composed of T_1, T_2, ..., T_(M-1), DATA_s is composed of T_2, T_4, ..., T_(M), and DATA_ and DATA_s form a complementary set. The key of the kV structure of the data entry T_(m) adopts the digital characteristic value of Value.

[0055] 3) The decentralized blockchain triggers convergence when the load of two consecutive data blocks is less than 50kB, and the data sequence Bys is aggregated to the data sequence By to form the data sequence By. The convergence trigger condition can be a data load capacity threshold or other business scenario settings.

[0056] The sequence By and Bys here Fy_Sh yml format is re-described as follows:

[0057]

[0058] Specific implementation case 3, applied to time series data processing of edge computing, is as follows:

[0059] A time series network dataset of an edge computing computer network. The time series dataset consists of several parallel data block sequences. The data sequence is composed of data blocks encapsulated using the zip rule and anchored in time sequence. The data blocks are composed of Figure 1 The data structure shown in the figure is composed of Figure 2 When the data block of a data sequence exceeds the set capacity safety threshold or the business setting needs, the data sequence is split into two data sequences, and the business data sets of the two data sequences complement each other, such as Figure 5 and Figure 6 As shown; when the data block of the dispersed data sequence is lower than the set capacity threshold or the business setting needs, the dispersed data is merged and returned to the previous data sequence, such as Figure 4 .

[0060] The method for regularizing dispersion and collection of a time series data set in an implementation example of the present invention comprises the following steps:

[0061] 1) Encapsulate each data sequence B1, ~, BN in the data structure using the zip rule according to the specific business scenario. The data blocks in the same data sequence By are anchored in sequence using time sequence, y = 1 ~ N, such as Figure 4 Among them, the data blocks that constitute the data sequence include Fy_Sh, F_DS, Ti_(y, x), and DATA, a total of 4 parts, such as Figure 1 , where Fy_Sh is the type of data block, F_DS is the flag of the data sequence By where the data block is located, Ti_(y, x) is the encapsulation time of the data block, and DATA refers to the specific business data load.

[0062] The data payload DATA constituting the data block includes T_1, T_2, ..., T_(M), wherein T_(m) (m=1 to M) is a specific business data entry, and the data entry adopts a kV (key-Value) data structure, and the data structure matches the type Fy_Sh.

[0063] Here the encoding format of Fy_Sh is described as follows:

[0064] Fy_Sh=[00 00 01 00 00 01]

[0065] Among them, the mask Mask = [00 00 01 01 01 01] is set according to the specific business scenario. It indicates that the current data block is not in a scattered state and uses a key-value data structure.

[0066] 2) As the time series data set grows, the business data load at the edge computing node exceeds the preset 2MB capacity threshold, triggering the data sequence to be dispersed. The data sequence By is then split into the data sequence Bys. The data structure of the data sequence Bys is the same as that of By, but the data load business data items in the data blocks of the data sequence By and the data sequence Bys are complementary sets. The data items of the two data sequences are classified based on the odd and even eigenvalues, such as Figure 5 and Figure 6 The characteristic value of the data entry may be obtained by serialization or other custom methods, and the method for obtaining the characteristic value of the data entry is placed in a parameter structure of type Fy_Sh.

[0067] The encoding format of Fy_Sh in the sequence By and Bys is described as follows:

[0068] Fy_Sh=[00 00 00 01 01 01]

[0069] Among them, the mask Mask = [00 00 01 01 01 01] is set according to the specific business scenario. It indicates that the data block is in a scattered state. The number of sequences after dispersion is currently 2. The parity of the last right bit of the data entry index value adopts a key-value data structure.

[0070] The dispersion trigger condition can be a data load capacity threshold or other business scenario settings. DATA_ is the data load after the data sequence By is split, and DATA_s is the data load of the data sequence Bys. DATA_ is composed of T_1, T_2, ..., T_(M-1), DATA_s is composed of T_2, T_4, ..., T_(M), and DATA_ and DATA_s form a complementary set. The key of the kV structure of the data entry T_(m) adopts the digital characteristic value of Value.

[0071] 3) When the load of two consecutive data blocks of the dispersed data sequence is less than 50 kB, the data sequence Bys is converged to form the data sequence By. The convergence trigger condition may be a data load capacity threshold or other business scenario settings.

[0072] The encoding format of the sequence By and Bys Fy_Sh is described as follows:

[0073] Fy_Sh=[00 00 00 00 00 01]

[0074] It indicates that the current data block is not in a scattered state and uses a key-value data structure.

[0075] Specific implementation case 4, using a time series network database, is as follows:

[0076] In the time series network database of computer network, there are several parallel data block sequences, and the data sequence is composed of data blocks anchored in sequence using feature values. Figure 1 The data structure shown in the figure is composed of Figure 2 When the data block of a data sequence exceeds the set capacity safety threshold or the business setting needs, the data sequence is split into two data sequences, and the business data of the two data sequences complement each other, such as Figure 5 and Figure 6As shown; when the data block of the dispersed data sequence is lower than the set capacity threshold or the business setting needs, the dispersed data is merged and returned to the previous data sequence, such as Figure 4 .

[0077] The method for regular dispersion and collection of a time series data set in an implementation example of the present invention comprises the following steps:

[0078] 1) Encapsulate each data sequence B1, ~, BN in the data structure of the time series network database according to the specific business scenario, and use the hash value to anchor the data blocks in the same data sequence By in sequence, y = 1 ~ N. The data sequences B1, ~, BN are finally encapsulated into their own data dependency structures according to their own business characteristics, such as Figure 4 Among them, the data blocks that constitute the data sequence include 6 parts: Fy_Sh, F_DS, Ti_(y,x), DATA, H_(y,(x-1)), and H_G(z-1). Figure 1 , where Fy_Sh is the type of data block, F_DS is the flag of the data sequence By where the data block is located, Ti_(y,x) is the encapsulation time of the data block, DATA is the specific business data load carried by the database sequence, H_(y,(x-1)) is the digital characteristic value of the preceding data block (i.e., the x-1th) of the x-th data block in the y-th column, and H_G(z-1) is the digital characteristic value of the dependent super block (i.e., the z-1th) of the x-th data block in the y-th column.

[0079] The description parameter organization form of the data block type Fy_Sh adopts JSON. The data load DATA of the data block includes T_1, T_2, ..., T_(M), where T_(m) (m=1~M) is a specific business data entry, and the data entry adopts a TT (time series-transaction) data structure, which matches the type Fy_Sh.

[0080] The json format of Fy_Sh in this case is described as follows:

[0081]

[0082] Among them, the TyofB key value indicates whether the current block is decentralized, On means it is decentralized, otherwise it is off; the kV key value indicates whether the key-value data structure is adopted, On means it is adopted; the TT key value indicates whether the "timing-transaction" data structure is adopted, On means it is adopted.

[0083] 2) As the time series database continues to grow, the data load of the sequence data block carrying the business exceeds the preset 2MB capacity threshold, and the database system will trigger the sequence to be dispersed. The data sequence By will then split into the data sequence Bys. The data structure of the data sequence Bys is the same as that of By, but the data load business data items in the data blocks of the data sequences By and Bys are complementary sets. The data items of the two data sequences are classified based on the odd and even eigenvalues, such as Figure 5 and Figure 6 The characteristic value of the data entry may be obtained in a serialized manner, and the means for obtaining the characteristic value of the data entry is placed in a parameter structure of type Fy_Sh.

[0084] The json format of the sequences By and Bys in this case is described as follows:

[0085]

[0086] Among them, the TyofB key value indicates whether the current block is in a decentralized state, On means it is in a decentralized state, otherwise it is off; the NofSh key value indicates the number of sequences after dispersion, which is currently 2; the TyofSh key value indicates the basis for dispersion, eo indicates the parity of the last right bit of the data entry index value; the kV key value indicates whether the key-value data structure is adopted, Off means it is adopted; the TT key value indicates whether the "time series-transaction transaction" data structure is adopted, On means it is adopted.

[0087] Among them, DATA_ is the data payload after the data sequence By is split, and DATA_s is the data payload of the data sequence Bys. DATA_ is composed of T_1, T_2, ..., T_(M-1), and DATA_s is composed of T_2, T_4, ..., T_(M). DATA_ and DATA_s form a complementary set. Among them, the key of the kV structure of the data entry T_(m) adopts the digital characteristic value of Value.

[0088] 3) As the time series database continues to expand, the global response of the database is triggered after a preset 10-minute time interval, and the super block GB(z) obtains the data digital feature value H_(y,x) of each data sequence. The super block data block includes Ti_G(z), DATA_G, H_(y) (y=1~N,ys), and H_G(z-1), where Ti_G(z) is the encapsulation time of the data block GB(z) of the super block sequence, DATA_G is the data load of the data block of the super block sequence, H_(y) (y=1~N,ys) is the data block digital feature of the data sequence By(y=1~N,ys), where the data sequence Bys is the data sequence split from the data sequence By, and H_G(z-1) is the preceding data block (z-1th) of the zth data block of the super block sequence.

[0089] 4) The scattered data sequences are triggered to be gathered when the load of two consecutive data blocks is less than 50 kB, and the data sequence Bys is gathered to the data sequence By to form the data sequence By. The gathering trigger condition may be a data load capacity threshold or other business scenario settings.

[0090] The sequence By and Bys here Fy_Sh's json format is re-described as follows:

[0091]

[0092] The above is only an embodiment of the present application and is not intended to limit the present application. For those skilled in the art, the present application may have various changes and variations. Any modification, equivalent replacement, improvement, etc. made within the spirit and principle of the present application should be included in the scope of the claims of the present application.

Claims

1. A method for regular dispersion and collection of a temporal network data set, comprising the steps of: 1) Encapsulate each data sequence B1, ..., BN in the data structure for storing distributed time series data sets according to the corresponding encapsulation algorithm rules; the data blocks in the encapsulated data sequence By at least include Fy_Sh, F_DS, Ti_(y,x), DATA, H_(y,(x-1)) and H_G(z-1); where Fy_Sh is the type of the data block, F_DS is the mark of the data sequence By where the data block is located, Ti_(y,x) is the encapsulation timing mark or other agreed mark of the xth data block in the yth column data sequence By, and D ATA refers to the specific service data load in the data block, H_(y,(x-1)) is the digital characteristic value of the preceding data block of the x-th data block in the y-th column data sequence By, y=1~N; H_G(z-1) is the digital characteristic value of the super-block corresponding to the x-th data block in the y-th column data sequence By; the super-block of the x-th data block in the data sequence By is the (z-1)th super-block in the super-block sequence, which at least stores the data sequence flag F_DS of the data sequence By, the data encapsulation timing Ti_(y,x) and the digital characteristic value H_G(z-1) of the preceding super-block; 2) When the time series data set continues to grow based on specific business scenarios and reaches the dispersion trigger condition, the data sequence By is split into the data sequence Bys; the data sequence Bys has the same data structure as the data sequence By, and the data load business data items in the data blocks of the data sequence By and the data sequence Bys are mutually cooperative complementary sets; DATA_ is the data payload after the data sequence By is split, and DATA_s is the data payload of the data sequence Bys; 3) When the time series data set continues to grow based on specific business scenarios and reaches the global trigger condition, the super block GB(z) in the super block sequence obtains the digital feature value H_(y,x) of the data block of each data sequence. The super block GB(z) includes Ti_G(z), DATA_G, H_(y') and H_G(z-1); where Ti_G(z) is the encapsulation time series identifier of the super block GB(z), DATA_G is the data payload; for H_(y') is the digital feature of the data block of the data sequence By', y'=1~N and ys; 4) When the time series data set continues to grow based on specific business scenarios and reaches the convergence trigger condition, the data sequence Bys is aggregated into the data sequence By, and the data block type Fy_Sh is set to the restored undispersed type, and the digital characteristic value H_s(y,(x-1)) of the preceding data block of the xth data block in the data sequence Bys is added to H_(y,(x-1)).

2. The method according to claim 1, characterized in that DATA_ is composed of T_1, T_3, …, T_(M-1), DATA_s is composed of T_2, T_4, …, T_(M), and DATA_ and DATA_s form complementary sets; where M is the total number of data entries before the data sequence By is split.

3. The method according to claim 1, characterized in that The data entry T_(m) is stored in a key-value pair structure, where the key is the digital characteristic value, encapsulation sequence or other business scenario setting of the key value, and m=1~M.

4. The method according to claim 1, characterized in that The dispersion trigger condition is a data load capacity threshold; the global trigger condition is a set time interval; and the convergence trigger condition is that the data load after the split is less than the set threshold.

5. The method according to claim 1, characterized in that According to different business scenarios or other set criteria, select the encapsulation algorithm rules corresponding to each data sequence for encapsulation.

6. The method according to claim 1, characterized in that The distributed time series data set is a time series network data set of a parallel blockchain, and the data block is a data block of a blockchain.

7. The method according to claim 1, characterized in that The distributed time-series data set is a time-series network data set of a serial blockchain, and the data block is a data block of a blockchain.

8. The method according to claim 1, characterized in that The distributed time series data set is a time series network data set of an edge computing network or the data structure is a data structure of a time series network database.

Images