Method and system for recovering data in memory state of digital network data, and digital network system

Abstract

The present invention discloses a method, system and digital networking system for recovering data pragmatic memory state data, which relates to the field of digital networking, and includes: a master digital object warehouse node receives a call request and executes it, and broadcasts the call request to all slave digital object warehouse nodes at the same time; if any slave digital object warehouse node has an error in calling the result, the master digital object warehouse node sends the memory state data to the slave digital object warehouse node, and if the master digital object warehouse node has an error in calling the result, the memory state data is obtained from any correct slave digital object warehouse node. The present invention ensures that during the multi-point data pragmatic operation life cycle, after a computer node has problems such as a digital object crash, downtime, power failure, etc. and returns to normal, the data pragmatic memory state data can be automatically recovered, and the recovery method of the present invention has high practicality.

Description

Technical Field

[0001] The present invention relates to the field of digital networking, and in particular to a method and system for recovering pragmatic memory state data of digital networking data, and a digital networking system. Background Art

[0002] With the advent of the big data era, the value of data has gradually become prominent. Through the analysis of big data, many additional values ​​can be generated. For example, during the epidemic, through big data analysis, it is possible to accurately and quickly query which cities a person has visited, which trains he has taken, whether he has been to high-risk areas, and so on. To use data for analysis, data exchange and processing are required. In the data sharing and exchange scenario, there is a problem of supervision of large-scale shared data, that is, a serious trust problem. The demander needs to use the supplier's data and produce valuable results by operating on the data, but the supplier often does not trust the demander. Once the data is transmitted from the supplier to the demander, the demander can do anything, such as doing something different from the previously declared operation on the data, or even reselling the original data.

[0003] Based on the above big data application requirements, there are currently big data network systems that run data pragmatics based on memory (the so-called data pragmatics refers to a digital object. Data pragmatics is divided into general data pragmatics and clustered data pragmatics. General data pragmatics needs to be created and executed in a data pragmatics warehouse; clustered data pragmatics needs to be created and executed in multiple data pragmatics warehouses). Provide corresponding services for big data application scenarios. In the big data network system where the data pragmatics adopts the memory computing mode, the sharing and exchange of big data is realized in the form of data pragmatics, while in the big data network system where the data pragmatics adopts the memory computing mode, the data pragmatics adopts the memory-based computing mode and the random multi-point execution mode.

[0004] The memory-based computing model can well meet the requirements of data pragmatics execution time, improve the throughput of data pragmatics execution, have good overall performance, and no additional overhead. However, in the memory-based computing model, since data pragmatics is calculated, run, and stored in memory, when force majeure occurs, the device running the data pragmatics will experience problems such as digital object crash, downtime, power failure, etc., then the memory state data of the entire data pragmatics will be lost and cannot be recovered. Therefore, in the memory-based computing model, how to ensure the data pragmatics execution time requirements and the data pragmatics execution throughput, while ensuring that the data pragmatics memory state data can also be automatically restored after the relevant equipment encounters digital object crash, downtime, power failure, etc. and returns to normal, is an urgent problem to be solved. Summary of the invention

[0005] In view of the above problems, the present invention provides a method, device and digital networking system for recovering data pragmatics memory status data of a digital networking system. While ensuring the data pragmatics execution time requirements and data pragmatics execution throughput, it also ensures that the data pragmatics memory status data can be automatically restored after the relevant equipment encounters problems such as digital object crash, downtime, power outage, etc. and returns to normal.

[0006] An embodiment of the present invention provides a method for recovering data in a data pragmatic memory state, the method being applied to a plurality of digital object warehouse nodes in a data network, each of the plurality of digital object warehouse nodes running a data pragmatic, the data pragmatic being a multi-point data pragmatic, the plurality of digital object warehouse nodes forming a data pragmatic cluster, the method comprising:

[0007] The master digital object warehouse node in the data pragmatics cluster receives the call request and executes it, and at the same time, the master digital object warehouse node broadcasts the call request to all slave digital object warehouse nodes in the data pragmatics cluster;

[0008] The master digital object repository node receives all the call results fed back by the slave digital object repository nodes;

[0009] In the case that the calling result fed back by the first slave digital object repository node among all slave digital object repository nodes is wrong, the master digital object repository node sends the full state data of the multi-point data pragmatics running by itself to the first slave digital object repository node;

[0010] When the master digital operator digital object repository node executes the call request and determines that its own call result is wrong, the master digital operator digital object repository node sends a status data request to a second slave digital operator digital object repository node among all slave digital operator digital object repository nodes, where the second slave digital operator digital object repository node is any slave digital operator digital object repository node among all slave digital operator digital object repository nodes whose call results are correct;

[0011] The master digital object repository node receives the full state data of the multi-point data application running on the second slave digital object repository node sent by the second slave digital object repository node;

[0012] In the case where the first slave operator digital object warehouse node goes offline and comes back online, the first slave operator digital object warehouse node sends recovery information to the master operator digital object warehouse node, and the master operator digital object warehouse node sends the full state data of the multi-point data pragmatics running on itself to the first slave operator digital object warehouse node, or sends the incremental state data of the multi-point data pragmatics running on itself to the first slave operator digital object warehouse node, or does not send any data to the first slave operator digital object warehouse node according to the recovery information and the current operation mode of the data pragmatics cluster;

[0013] When the master digital object warehouse node goes offline and comes back online, and the master digital object warehouse node becomes a slave digital object warehouse node, and the second slave digital object warehouse node becomes a new master digital object warehouse node, the master digital object warehouse node sends the recovery information to the second slave digital object warehouse node, and the second slave digital object warehouse node sends the full state data of the multi-point data pragmatics running on itself to the master digital object warehouse node, or sends the incremental state data of the multi-point data pragmatics running on itself to the master digital object warehouse node, or does not send any data to the master digital object warehouse node according to the recovery information and the current operation mode of the data pragmatics cluster;

[0014] When the master operator digital object warehouse node goes offline and comes back online, and the master operator digital object warehouse node is still the master operator digital object warehouse node, the master operator digital object warehouse node performs the steps of: receiving a call request and executing it, and at the same time, the master operator digital object warehouse node broadcasts the call request to all slave operator digital object warehouse nodes in the data pragmatics cluster.

[0015] Optionally, before the main operator digital object repository node in the data pragmatics cluster receives the call request and executes it, the method further includes:

[0016] The master operator digital object repository node determines that the operation mode is a normal operation mode when the classification of the multi-point data pragmatics is a stateless data pragmatics;

[0017] When the classification of the multi-point data pragmatics by the master digital object warehouse node is stateful data pragmatics, and the total number of digital object warehouse nodes in the data pragmatics cluster is between the total number of digital object warehouse nodes currently operating normally, and the first preset condition is met, the master digital object warehouse node determines that the operation mode is the normal operation mode;

[0018] When the classification of the multi-point data pragmatics by the master digital object warehouse node is stateful data pragmatics, and the total number of digital object warehouse nodes in the data pragmatics cluster is between the total number of digital object warehouse nodes currently operating normally, and the second preset condition is met, the master digital object warehouse node determines that the operation mode is the stable operation mode;

[0019] In the case of the normal operation mode, each digital object warehouse node of the data operator in the data pragmatics cluster runs in a memory persistence mode;

[0020] In the case of the stable operation mode, each of the digital object warehouse nodes of the data operator records the full state data and the incremental state data of the multi-point data pragmatics to its own hard disk;

[0021] When the number of records of the incremental state data of the multi-point data pragmatics reaches a preset number, each of the digital object warehouse nodes of the operator updates and records the full state data of the multi-point data pragmatics, and deletes the incremental state data records of the multi-point data pragmatics;

[0022] The first preset condition is that the total number of the currently operating Shurui operator digital object warehouse nodes is greater than the total number of the Shurui operator digital object warehouse nodes in the data pragmatics cluster after a preset operation is performed, and the preset operation is: taking half of the total number of the Shurui operator digital object warehouse nodes in the data pragmatics cluster to obtain a first result, and rounding up the first result;

[0023] The second preset condition is that the total number of the Shurui operator digital object warehouse nodes that are currently operating normally is not greater than the total number after a preset calculation is performed on the total number of the Shurui operator digital object warehouse nodes in the data pragmatics cluster.

[0024] Optionally, the recovery information includes: the operation mode of the data pragmatics cluster before the first slave Shurui operator digital object warehouse node goes offline, and the version number of the multi-point data pragmatics running after the first slave Shurui operator digital object warehouse node comes back online;

[0025] In the case where the first slave operator digital object warehouse node goes offline and comes back online, the first slave operator digital object warehouse node sends recovery information to the master operator digital object warehouse node, and the master operator digital object warehouse node sends the full state data of the multi-point data pragmatics running by itself to the first slave operator digital object warehouse node, or sends the incremental state data of the multi-point data pragmatics running by itself to the first slave operator digital object warehouse node, or does not send any data to the first slave operator digital object warehouse node, according to the recovery information and the current operation mode of the data pragmatics cluster, including:

[0026] In the case that the data pragmatics cluster is in the normal operation mode before the first slave operator digital object warehouse node goes offline, and the data pragmatics cluster is currently in any operation mode after the first slave operator digital object warehouse node comes back online, the master operator digital object warehouse node determines whether the version number of the multi-point data pragmatics running after the first slave operator digital object warehouse node comes back online is the same as the current state version number of the data pragmatics cluster. If they are the same, the master operator digital object warehouse node does not send any data to the first slave operator digital object warehouse node. If they are not the same, the master operator digital object warehouse node sends the full state data of the multi-point data pragmatics running by itself to the first slave operator digital object warehouse node, so that the first slave operator digital object warehouse node runs its own multi-point data pragmatics based on the received full state data;

[0027] When the data pragmatics cluster is in the stable operation mode before the first slave digital object warehouse node goes offline, and the data pragmatics cluster is in the normal operation mode after the first slave digital object warehouse node comes back online, the master digital object warehouse node sends the full state data of the multi-point data pragmatics running by itself to the first slave digital object warehouse node, so that the first slave digital object warehouse node runs its own multi-point data pragmatics based on the received full state data;

[0028] When the data pragmatics cluster is in the stable operation mode before the first slave digital object warehouse node goes offline, and the data pragmatics cluster is in the stable operation mode after the first slave digital object warehouse node comes back online, the first slave digital object warehouse node obtains the full state data and incremental state data of the multi-point data pragmatics from its own hard disk. At the same time, the master digital object warehouse node determines whether the version number of the multi-point data pragmatics running after the first slave digital object warehouse node comes back online is the same as the current state version number of the data pragmatics cluster. If they are the same, the master digital object warehouse node The operator digital object warehouse node does not send any data to the first slave operator digital object warehouse node. Otherwise, the master operator digital object warehouse node sends the newly added incremental status data of the multi-point data application running by itself to the first slave operator digital object warehouse node, so that the first slave operator digital object warehouse node runs its own multi-point data application based on the received newly added incremental status data and the full status data and incremental status data obtained from its own hard disk. The newly added incremental status data is the incremental status data that is not in the first slave operator digital object warehouse node's own hard disk due to the first slave operator digital object warehouse node being offline.

[0029] Optionally, the recovery information also includes: the operation mode of the data pragmatics cluster before the master data operator digital object warehouse node goes offline, and the version number of the multi-point data pragmatics running after the master data operator digital object warehouse node comes back online;

[0030] In the case where the master digital object warehouse node goes offline and comes back online, and the master digital object warehouse node becomes a slave digital object warehouse node, and the second slave digital object warehouse node becomes a new master digital object warehouse node, the master digital object warehouse node sends the recovery information to the second slave digital object warehouse node, and the second slave digital object warehouse node sends the full state data of the multi-point data pragmatics running on itself to the master digital object warehouse node, or sends the incremental state data of the multi-point data pragmatics running on itself to the master digital object warehouse node, or does not send any data to the master digital object warehouse node, according to the recovery information and the current operation mode of the data pragmatics cluster, including:

[0031] In the case that the data pragmatics cluster is in the normal operation mode before the master digital object warehouse node goes offline, and the data pragmatics cluster is currently in any operation mode after the master digital object warehouse node comes back online, the second slave digital object warehouse node determines whether the version number of the multi-point data pragmatics running after the master digital object warehouse node comes back online is the same as the current status version number of the data pragmatics cluster. If they are the same, the second slave digital object warehouse node does not send any data to the master digital object warehouse node. If they are not the same, the second slave digital object warehouse node sends the full status data of the multi-point data pragmatics running by itself to the master digital object warehouse node, so that the master digital object warehouse node runs its own multi-point data pragmatics based on the received full status data;

[0032] When the data pragmatics cluster is in the stable operation mode before the master digital object warehouse node goes offline, and the data pragmatics cluster is in the normal operation mode after the master digital object warehouse node comes back online, the second slave digital object warehouse node sends the full state data of the multi-point data pragmatics it runs to the master digital object warehouse node, so that the master digital object warehouse node runs its own multi-point data pragmatics based on the received full state data;

[0033] When the data pragmatics cluster is in the stable operation mode before the master operator digital object warehouse node goes offline, and when the data pragmatics cluster is in the stable operation mode after the master operator digital object warehouse node comes back online, the master operator digital object warehouse node obtains the full state data and incremental state data of the multi-point data pragmatics from its own hard disk. At the same time, the second slave operator digital object warehouse node determines whether the version number of the multi-point data pragmatics running after the master operator digital object warehouse node comes back online is the same as the current state version number of the data pragmatics cluster. If they are the same, the second slave operator digital object warehouse node does not send any data to the master operator digital object warehouse node. If they are not the same, the second slave operator digital object warehouse node sends the newly added incremental state data of the multi-point data pragmatics running by itself to the master operator digital object warehouse node, so that the master operator digital object warehouse node runs its own multi-point data pragmatics based on the received newly added incremental state data and the full state data and incremental state data obtained from its own hard disk.

[0034] Optionally, the master operator digital object repository node determines whether the version number of the multipoint data pragmatics running after the first slave operator digital object repository node comes back online is the same as the current state version number of the data pragmatics cluster, including:

[0035] In the case that the digital object of the multi-point data application on the first slave digital operator digital object repository node is not corrupted, the recovery information sent by the first slave digital operator digital object repository node to the master digital operator digital object repository node carries the version number of the multi-point data application running after the first slave digital operator digital object repository node comes back online;

[0036] The master digital object repository node determines, based on the received version number of the multipoint data pragmatics running after the first slave digital object repository node comes back online, whether the version number of the multipoint data pragmatics running after the first slave digital object repository node comes back online is the same as the current state version number of the data pragmatics cluster;

[0037] In the case that the digital object of the multi-point data application on the first slave digital operator digital object repository node crashes, the recovery information sent by the first slave digital operator digital object repository node to the master digital operator digital object repository node does not carry the version number of the multi-point data application running after the first slave digital operator digital object repository node comes back online;

[0038] The master operator digital object warehouse node does not receive the version number of the multi-point data pragmatics running after the first slave operator digital object warehouse node comes back online, and determines that the version number of the multi-point data pragmatics running after the first slave operator digital object warehouse node comes back online is different from the current state version number of the data pragmatics cluster.

[0039] Optionally, the recovery method further includes:

[0040] Before the first slave digital object warehouse node goes offline, if the total number of digital object warehouse nodes in the data pragmatics cluster and the total number of digital object warehouse nodes currently operating normally meet the first preset condition, the master digital object warehouse node determines that the operation mode is the normal operation mode;

[0041] After the first slave digital object warehouse node goes offline and when the first slave digital object warehouse node has not come back online, if the total number of digital object warehouse nodes in the data pragmatics cluster and the total number of digital object warehouse nodes currently operating normally still meet the first preset condition, then the master digital object warehouse node maintains the operating mode as the normal operating mode; or

[0042] If the total number of the Shurui operator digital object warehouse nodes in the data pragmatics cluster and the total number of the Shurui operator digital object warehouse nodes currently operating normally meet the second preset condition, the master Shurui operator digital object warehouse node switches the operating mode to the stable operating mode;

[0043] When the first slave digital object warehouse node goes offline and has not yet come back online, and the operation mode is currently the normal operation mode, when the first slave digital object warehouse node comes back online, if the total number of digital object warehouse nodes in the data pragmatics cluster and the total number of digital object warehouse nodes currently operating normally still meet the first preset condition, then the master digital object warehouse node maintains the operation mode as the normal operation mode; or

[0044] If the total number of the Shurui operator digital object warehouse nodes in the data pragmatics cluster and the total number of the Shurui operator digital object warehouse nodes currently operating normally meet the second preset condition, the master Shurui operator digital object warehouse node switches the operating mode to the stable operating mode;

[0045] When the first slave digital object warehouse node goes offline and has not yet come back online, and the operating mode is currently the stable operating mode, when the first slave digital object warehouse node comes back online, if the total number of digital object warehouse nodes in the data pragmatics cluster and the total number of digital object warehouse nodes currently operating normally meet the first preset condition, the master digital object warehouse node switches the operating mode to the normal operating mode; or

[0046] If the total number of the Shurui operator digital object warehouse nodes in the data pragmatics cluster and the total number of the Shurui operator digital object warehouse nodes currently operating normally still meet the second preset condition, the main Shurui operator digital object warehouse node maintains the operating mode as the stable operating mode; after the main Shurui operator digital object warehouse node switches the operating mode from the stable operating mode to the normal operating mode, all Shurui operator digital object warehouse nodes in the data pragmatics cluster delete the full status data and incremental status data recorded to the hard disk during the stable operating mode.

[0047] Optionally, the master digital object warehouse node goes offline and comes back online, and the master digital object warehouse node becomes a slave digital object warehouse node, and the second slave digital object warehouse node becomes the master digital object warehouse node, including:

[0048] When the main Shurui operator digital object warehouse node is offline and has not come back online after a preset period of time, all other Shurui operator digital object warehouse nodes in the data pragmatics cluster disconnect from the main Shurui operator digital object warehouse node and send a re-election request to the Shurui digital network identity resolution system in the digital network;

[0049] After receiving the re-election request sent by all the digital object repository nodes except the master digital object repository node, the digital network identification resolution system selects the second slave digital object repository node from all the digital object repository nodes except the master digital object repository node as the new master digital object repository node;

[0050] After the second slave digital object repository node becomes the new master digital object repository node, a connection request is sent to all digital object repository nodes except the master digital object repository node, so that all digital object repository nodes except the master digital object repository node are connected to the second slave digital object repository node;

[0051] When the master digital object repository node is back online, the second slave digital object repository node sends the connection request to the master digital object repository node, so that the master digital object repository node connects to the second slave digital object repository node as a slave digital object repository node.

[0052] Optionally, the master operator digital object warehouse node goes offline and comes back online, and the master operator digital object warehouse node is still the master operator digital object warehouse node, including:

[0053] When the master Shurui operator digital object warehouse node goes offline and comes back online within the preset time, other Shurui operator digital object warehouse nodes in the data pragmatics cluster will not sense that the master Shurui operator digital object warehouse node is offline, and still maintain the connection with the master Shurui operator digital object warehouse node;

[0054] After the master digital object warehouse node comes back online, the following steps are performed: receiving a call request and executing it, and at the same time, the master digital object warehouse node broadcasts the call request to all slave digital object warehouse nodes in the data pragmatics cluster.

[0055] The embodiment of the present invention also provides a data pragmatic memory state data recovery system, the recovery system comprising: a plurality of digital object warehouse nodes in a digital network, the plurality of digital object warehouse nodes comprising: a master digital object warehouse node and a slave digital object warehouse node; each of the plurality of digital object warehouse nodes runs data pragmatics, the data pragmatics is a multi-point data pragmatics, the plurality of digital object warehouse nodes form a data pragmatics cluster, the master digital object warehouse node comprising:

[0056] A receiving and executing broadcast module is used to receive and execute a call request, and broadcast the call request to all slave operator digital object warehouse nodes in the data pragmatics cluster;

[0057] The result receiving module is used to receive all the call results fed back from the Shurui operator digital object warehouse node;

[0058] A first sending module is used to send the full state data of the multi-point data pragmatics running by itself to the first slave digital object warehouse node when the call result fed back by the first slave digital object warehouse node among all slave digital object warehouse nodes is wrong;

[0059] a data request module, configured to, when executing the call request and determining that its own call result is wrong, send a status data request to a second slave operator digital object repository node among all slave operator digital object repository nodes, wherein the second slave operator digital object repository node is any slave operator digital object repository node among all slave operator digital object repository nodes whose call results are correct;

[0060] A data receiving module, configured to receive full state data of a multi-point data application running on the second slave digital object repository node, sent by the second slave digital object repository node;

[0061] an information receiving and data sending module, used for receiving recovery information sent by the first slave digital object warehouse node when the first slave digital object warehouse node goes offline and comes back online, and sending the full state data of the multi-point data pragmatics running by itself to the first slave digital object warehouse node, or sending the incremental state data of the multi-point data pragmatics running by itself to the first slave digital object warehouse node, or not sending any data to the first slave digital object warehouse node according to the recovery information and the current operation mode of the data pragmatics cluster;

[0062] The information sending and data receiving module is used to send the recovery information to the second slave digital object warehouse node when the module goes offline and comes back online, and the module becomes a slave digital object warehouse node, and the second slave digital object warehouse node becomes a new master digital object warehouse node, and receive the full state data or incremental state data of the multi-point data pragmatics of the second slave digital object warehouse node running, which is sent by the second slave digital object warehouse node according to the recovery information and the running mode;

[0063] The execution module is used to execute the steps of receiving a call request and executing it when it goes offline and comes back online, and when it is still the master digital object warehouse node: broadcasting the call request to all slave digital object warehouse nodes in the data pragmatics cluster.

[0064] Optionally, the main digital operator digital object warehouse node further includes: an operation mode determination module, the operation mode determination module is used to:

[0065] In the case where the multi-point data pragmatics is classified as a stateless data pragmatics, determining the operation mode to be a normal operation mode;

[0066] When the multi-point data pragmatics is classified as stateful data pragmatics, and the total number of the digital object repository nodes of the data pragmatics cluster is between the total number of the digital object repository nodes of the currently normally operating data pragmatics, and a first preset condition is satisfied, determining that the operation mode is the normal operation mode;

[0067] When the multi-point data pragmatics is classified as stateful data pragmatics, and the total number of digital object warehouse nodes of the data pragmatics cluster is between the total number of digital object warehouse nodes of the currently normally operating data pragmatics, and a second preset condition is satisfied, determining that the operation mode is the stable operation mode;

[0068] The master digital object repository node and the slave digital object repository node both include: an operation record module, and the operation record module is used to:

[0069] In the case of the normal operation mode, the system runs in a memory persistence mode;

[0070] In the case of the stable operation mode, the full state data and the incremental state data of the multi-point data language are recorded in their own hard disks respectively;

[0071] When the number of records of the incremental status data of the multi-point data language reaches a preset number, the full status data of the multi-point data language is updated and recorded respectively, and the incremental status data records of the multi-point data language are deleted respectively;

[0072] The first preset condition is that the total number of the currently operating Shurui operator digital object warehouse nodes is greater than the total number of the Shurui operator digital object warehouse nodes in the data pragmatics cluster after a preset operation is performed, and the preset operation is: taking half of the total number of the Shurui operator digital object warehouse nodes in the data pragmatics cluster to obtain a first result, and rounding up the first result;

[0073] The second preset condition is that the total number of the Shurui operator digital object warehouse nodes that are currently operating normally is not greater than the total number after a preset calculation is performed on the total number of the Shurui operator digital object warehouse nodes in the data pragmatics cluster.

[0074] Optionally, the recovery information includes: the operation mode of the data pragmatics cluster before the first slave Shurui operator digital object warehouse node goes offline, and the version number of the multi-point data pragmatics running after the first slave Shurui operator digital object warehouse node comes back online;

[0075] The module for receiving information and sending data is specifically used for:

[0076] In the case that the data pragmatics cluster is in the normal operation mode before the first slave operator digital object warehouse node goes offline, and the data pragmatics cluster is currently in any operation mode after the first slave operator digital object warehouse node comes back online, determine whether the version number of the multi-point data pragmatics running after the first slave operator digital object warehouse node comes back online is the same as the current state version number of the data pragmatics cluster; if they are the same, no data is sent to the first slave operator digital object warehouse node; if they are not the same, the full state data of the multi-point data pragmatics running by itself is sent to the first slave operator digital object warehouse node, so that the first slave operator digital object warehouse node runs its own multi-point data pragmatics based on the received full state data;

[0077] Before the first slave operator digital object warehouse node goes offline, the data pragmatics cluster is in the stable operation mode, and after the first slave operator digital object warehouse node comes back online, the data pragmatics cluster is in the normal operation mode, sending the full state data of the multi-point data pragmatics running by itself to the first slave operator digital object warehouse node, so that the first slave operator digital object warehouse node runs its own multi-point data pragmatics based on the received full state data;

[0078] When the data pragmatics cluster is in the stable operation mode before the first slave operator digital object warehouse node goes offline, and the data pragmatics cluster is in the stable operation mode after the first slave operator digital object warehouse node comes back online, the first slave operator digital object warehouse node obtains the full state data and incremental state data of the multi-point data pragmatics from its own hard disk, and at the same time, determines whether the version number of the multi-point data pragmatics running after the first slave operator digital object warehouse node comes back online is the same as the current state version number of the data pragmatics cluster. If they are the same, no data is sent to the first slave operator digital object warehouse node. If they are not the same, the newly added incremental state data of the multi-point data pragmatics running by itself is sent to the first slave operator digital object warehouse node, so that the first slave operator digital object warehouse node runs its own multi-point data pragmatics based on the received newly added incremental state data and the full state data and incremental state data obtained from its own hard disk. The newly added incremental state data is the incremental state data that is not in the first slave operator digital object warehouse node's own hard disk due to the first slave operator digital object warehouse node being offline.

[0079] Optionally, the recovery information also includes: the operation mode of the data pragmatics cluster before the master data operator digital object warehouse node goes offline, and the version number of the multi-point data pragmatics running after the master data operator digital object warehouse node comes back online;

[0080] The sending information receiving data module is specifically used for:

[0081] Sending the recovery information to the second slave operator digital object repository node;

[0082] receiving the full state data of the multi-point data pragmatics sent by the second slave operator digital object repository node, and running its own multi-point data pragmatics based on the received full state data;

[0083] Alternatively, the full status data and incremental status data of the multi-point data pragmatics are obtained from its own hard disk, and at the same time, the newly added incremental status data of the multi-point data pragmatics sent by the second slave operator digital object warehouse node is received, and based on the received newly added incremental status data and the full status data and incremental status data obtained from its own hard disk, its own multi-point data pragmatics is run.

[0084] The embodiment of the present invention further provides a digital networking system, the digital networking system comprising: a digital networking identification resolution system, a plurality of digital storage object warehouse nodes, and a plurality of digital operator digital object warehouse nodes;

[0085] The Shurui digital network identification resolution system is used to manage whether the Shurui operator digital object warehouse node and the Shurui storage digital object warehouse node join the digital network system;

[0086] The plurality of digital object repository nodes of the data operator are used to execute any of the above methods for recovering data in the data pragmatic memory state;

[0087] The plurality of data storage digital object repository nodes are used to store at least one of the following information:

[0088] The operating result of the multi-point data pragmatics;

[0089] Status data of the multi-point data pragmatics;

[0090] The multiple data points include execution branch information.

[0091] The present invention provides a method for recovering data pragmatics memory status data. A master digital object warehouse node receives a call request and executes it, and also broadcasts the call request to all slave digital object warehouse nodes; then receives the call results fed back by all slave digital object warehouse nodes; when the call result fed back by the first slave digital object warehouse node is wrong, that is, the data pragmatics of the multi-point data pragmatics running on the first slave digital object warehouse node crashes, or the sequence number of the cross-data pragmatics call is inconsistent with that of other digital object warehouse nodes, or the current status version number of the multi-point data pragmatics is inconsistent with that of other digital object warehouse nodes, the master digital object warehouse node sends the full status data of the multi-point data pragmatics running on itself to the first slave digital object warehouse node, thereby recovering the multi-point data pragmatics memory status data running on the first slave digital object warehouse node.

[0092] When the master operator digital object warehouse node executes a call request and determines that its own call result is wrong, that is, the data pragmatics of the multi-point data pragmatics running on the master operator digital object warehouse node crashes, or the sequence number of the cross-data pragmatics call is inconsistent with that of other operator digital object warehouse nodes, or the current state version number of the multi-point data pragmatics is inconsistent with that of other operator digital object warehouse nodes, the master operator digital object warehouse node sends a status data request to any slave operator digital object warehouse node among all slave operator digital object warehouse nodes with correct call results, for example, the second slave operator digital object warehouse node, and the master operator digital object warehouse node receives the full state data of the multi-point data pragmatics running on the second slave operator digital object warehouse node sent by the second slave operator digital object warehouse node, thereby restoring the memory state data of the multi-point data pragmatics running on the master operator digital object warehouse node.

[0093] For the case where the Shu Rui Operator Digital Object Warehouse node goes offline and comes back online due to various factors: when the first slave Shu Rui Operator Digital Object Warehouse node goes offline and comes back online, the first slave Shu Rui Operator Digital Object Warehouse node sends recovery information to the master Shu Rui Operator Digital Object Warehouse node, and the master Shu Rui Operator Digital Object Warehouse node sends the full status data of the multi-point data pragmatics running on itself to the first slave Shu Rui Operator Digital Object Warehouse node, or sends the incremental status data of the multi-point data pragmatics running on itself to the first slave Shu Rui Operator Digital Object Warehouse node, or does not send any data to the first slave Shu Rui Operator Digital Object Warehouse node; thereby restoring the multi-point data pragmatics memory status data running on the first slave Shu Rui Operator Digital Object Warehouse node.

[0094] When the master Shurui operator digital object warehouse node goes offline and comes back online, and the master Shurui operator digital object warehouse node becomes a slave Shurui operator digital object warehouse node, and the second slave Shurui operator digital object warehouse node becomes the new master Shurui operator digital object warehouse node, the master Shurui operator digital object warehouse node sends recovery information to the second slave Shurui operator digital object warehouse node, and the second slave Shurui operator digital object warehouse node sends the full status data of the multi-point data pragmatics running on itself to the master Shurui operator digital object warehouse node, or sends the incremental status data of the multi-point data pragmatics running on itself to the master Shurui operator digital object warehouse node, or does not send any data to the master Shurui operator digital object warehouse node, thereby restoring the multi-point data pragmatics memory status data running on the master Shurui operator digital object warehouse node.

[0095] When the master operator digital object warehouse node goes offline and comes back online, and the master operator digital object warehouse node is still the master operator digital object warehouse node, the master operator digital object warehouse node only needs to execute the steps: receive the call request and execute it.

[0096] Through the above-mentioned multiple recovery methods, it is ensured that during the multi-point data pragmatic operation life cycle, if the relevant equipment encounters problems such as digital object crash, downtime, power failure, etc. and returns to normal, the data pragmatic memory status data can be automatically restored. The recovery method of the present invention has high practicality. BRIEF DESCRIPTION OF THE DRAWINGS

[0097] Various other advantages and benefits will become apparent to those of ordinary skill in the art by reading the detailed description of the preferred embodiments below. The accompanying drawings are only for the purpose of illustrating the preferred embodiments and are not to be considered as limiting the present invention. Moreover, the same reference symbols are used throughout the accompanying drawings to represent the same components. In the accompanying drawings:

[0098] Figure 1 It is a flow chart of a method for recovering data in a data pragmatic memory state according to an embodiment of the present invention;

[0099] Figure 2 It is a block diagram of a data pragmatic memory state data recovery system according to an embodiment of the present invention;

[0100] Figure 3 It is a schematic diagram of the results obtained by performing a fault recovery test on three recovery methods using data pragmatics with different state quantities under the same network delay in an embodiment of the present invention;

[0101] Figure 4 It is a schematic diagram of the results obtained by performing a fault recovery test on data pragmatics with the same state quantity size under different TCP packet delays in an embodiment of the present invention. DETAILED DESCRIPTION

[0102] In order to make the above-mentioned purposes, features and advantages of the present invention more obvious and easy to understand, the present invention is further described in detail below in conjunction with the accompanying drawings and specific embodiments. It should be understood that the specific embodiments described herein are only used to explain the present invention, are only part of the embodiments of the present invention, not all of the embodiments, and are not used to limit the present invention.

[0103] The data pragmatic memory state data recovery method proposed in the embodiment of the present invention proposes a recovery method for the current memory-based computing mode, referring to Figure 1 , shows a flow chart of a method for recovering data in a data pragmatic memory state according to an embodiment of the present invention. The recovery method is applied to multiple Shurui operator digital object warehouse nodes in a data network. Each of the multiple Shurui operator digital object warehouse nodes runs a data pragmatic, and the data pragmatic is a multi-point data pragmatic. The multiple Shurui operator digital object warehouse nodes form a data pragmatic cluster. The recovery method includes:

[0104] Step 101: The master digital object warehouse node in the data pragmatics cluster receives and executes the call request, and at the same time, the master digital object warehouse node broadcasts the call request to all slave digital object warehouse nodes in the data pragmatics cluster.

[0105] In the embodiment of the present invention, due to its own characteristics, multi-point data pragmatics are generally deployed and run on multiple digital object warehouse nodes of the data network. Since multi-point data pragmatics in the big data network system based on memory operation of data pragmatics can be implemented by sharding with data pragmatics as the granularity to achieve scalability and low cost, when the multi-point data pragmatics is started and run, the user can define how many digital object warehouse nodes of the digital operator are needed to run the multi-point data pragmatics.

[0106] As an example: a multi-point data pragmatics is stored on 20 Shurui operator digital object warehouse nodes. When the user needs to use the multi-point data pragmatics, he can choose 9 Shurui operator digital object warehouse nodes to run the multi-point data pragmatics, or he can choose 20 Shurui operator digital object warehouse nodes to run the multi-point data pragmatics.

[0107] Before starting the multi-point data pragmatics, the user needs to choose to initiate the multi-point data pragmatics request on a certain Shurui operator digital object warehouse node. Generally, the Shurui operator digital object warehouse node is used as the master Shurui operator digital object warehouse node in the data pragmatics cluster. At the same time, the user also needs to choose how many Shurui operator digital object warehouse nodes to run the multi-point data pragmatics on. The Shurui operator digital object warehouse nodes other than the master Shurui operator digital object warehouse node are slave Shurui operator digital object warehouse nodes in the data pragmatics cluster. Using the above example: assuming that the user chooses to initiate the multi-point data pragmatics request on a Shurui operator digital object warehouse node, and selects a total of 10 Shurui operator digital object warehouse nodes to run the data pragmatics, then the Shurui operator digital object warehouse node that initiates the multi-point data pragmatics request is the master Shurui operator digital object warehouse node, and 9 Shurui operator digital object warehouse nodes are randomly selected from the remaining 19 Shurui operator digital object warehouse nodes as slave Shurui operator digital object warehouse nodes.

[0108] After the multi-point data pragmatics is started and running, the main data operator digital object warehouse node can determine the operation mode of the data pragmatics cluster according to the classification of the multi-point data pragmatics. The classification of multi-point data pragmatics includes: stateless data pragmatics and stateful data pragmatics. The so-called stateless data pragmatics means that the execution result of the data pragmatics is only related to the call, and has nothing to do with the historical state of the data pragmatics, such as: four arithmetic data pragmatics, etc.; the so-called stateful data pragmatics means that the execution result of the data pragmatics is related to its historical execution, such as: counter data pragmatics, transfer data pragmatics, etc.

[0109] If the multi-point data pragmatics is classified as stateless data pragmatics, the main Shurui operator digital object warehouse node determines that the operation mode of the data pragmatics cluster is the normal operation mode. If the multi-point data pragmatics is classified as stateful data pragmatics, the main Shurui operator digital object warehouse node needs to determine whether the operation mode of the data pragmatics cluster is the normal operation mode or the stable operation mode based on the total number of Shurui operator digital object warehouse nodes in the data pragmatics cluster and the total number of Shurui operator digital object warehouse nodes currently operating normally.

[0110] Specifically: when the total number of Shurui operator digital object warehouse nodes in the data pragmatics cluster is between the total number of Shurui operator digital object warehouse nodes currently operating normally, and the first preset condition is met, the main Shurui operator digital object warehouse node determines that the operating mode is the normal operating mode; when the total number of Shurui operator digital object warehouse nodes in the data pragmatics cluster is between the total number of Shurui operator digital object warehouse nodes currently operating normally, and the second preset condition is met, the main Shurui operator digital object warehouse node determines that the operating mode is the stable operating mode.

[0111] Among them, the so-called first preset condition is: the total number of the currently operating normally digital object warehouse nodes of the digital object warehouse of the digital object warehouse of the digital object warehouse of the digital object warehouse of the data pragmatics cluster is greater than the total number after the preset operation is performed on the total number of digital object warehouse nodes of the digital object warehouse of the digital object warehouse of the data pragmatics cluster; the so-called second preset condition is: the total number of the currently operating normally digital object warehouse nodes of the digital object warehouse of the digital object warehouse of the digital object warehouse of the digital object warehouse of the digital object warehouse of the data pragmatics cluster is not greater than the total number after the preset operation is performed on the total number of digital object warehouse nodes of the digital object warehouse of the digital object warehouse of the digital object warehouse of the data pragmatics cluster; the so-called preset operation is: take half of the total number of digital object warehouse nodes of the digital object warehouse of the digital object warehouse of the digital object warehouse of the digital object warehouse of the data pragmatics cluster, obtain the first result, and round up the first result. Assume: the total number of digital object warehouse nodes of the digital object warehouse of the digital object warehouse of the digital object warehouse of the digital object warehouse of the data pragmatics cluster is N, and the total number of digital object warehouse nodes of the digital object warehouse of the digital object warehouse of the digital object warehouse of the data pragmatics cluster that are currently operating normally is M; then the preset operation is: ⌈N / 2⌉, the first preset condition is: M>⌈N / 2⌉, and the second preset condition is: M≤⌈N / 2⌉.

[0112] That is, when M>⌈N / 2⌉, it can be determined that the majority of the Shurui operator digital object warehouse nodes are currently operating normally in the data pragmatics cluster, so the main Shurui operator digital object warehouse node determines that the operating mode of the data pragmatics cluster is the normal operating mode; when M≤⌈N / 2⌉, it can be determined that there are not many Shurui operator digital object warehouse nodes currently operating normally in the data pragmatics cluster, so the main Shurui operator digital object warehouse node determines that the operating mode of the data pragmatics cluster is the stable operating mode.

[0113] After determining the operation mode of the data pragmatics cluster, each digital object warehouse node of the data operator in the data pragmatics cluster can store the memory status data of the multi-point data pragmatics running on itself according to the operation mode. Different operation modes correspond to different storage methods.

[0114] Specifically: when the operation mode is the normal operation mode, each Shurui operator digital object warehouse node stores its own memory status data in a memory persistence manner; when the operation mode is the stable operation mode, each Shurui operator digital object warehouse node stores its own memory status data in a hard disk persistence manner.

[0115] In the embodiment of the present invention, the memory state data of the multi-point data pragmatics includes: full state data and incremental state data. The full state data includes: all global variables in the data pragmatics and their current values, data pragmatics ID, data pragmatics Name, data pragmatics type, and data pragmatics state version number and other information. The incremental state data includes: parameters and method name of the data pragmatics call request, and if the call includes a cross-data pragmatics call, it also includes the execution result of the cross-data pragmatics call initiated by the data pragmatics.

[0116] Among them, all global variables and current values ​​in the data pragmatics are the data pragmatics status of the multi-point data pragmatics. In the data pragmatics based on memory computing, the data pragmatics status refers to the name, type and value of the global variables in the data pragmatics. The change of the data pragmatics status is due to the impact of different data pragmatics calls on the values ​​of global variables in the data pragmatics. These historical states will affect the results of subsequent data pragmatics calls.

[0117] State version number of data pragmatics: For stateful data pragmatics, each data pragmatics call will cause a change in the state of the data pragmatics. Therefore, after each call, a new state version number will be generated. For example, the initial state version number is -1. After the first call request is executed, the state version number becomes 0; after the second call request is executed, the state version number becomes 1. The state version number can be used to know whether the data pragmatics state of a certain data operator digital object warehouse node lags behind other nodes. And when executing a certain data pragmatics call request, it can be judged whether the data operator digital object warehouse node has missed some call requests or received some call requests repeatedly by whether the current data pragmatics version number and the request sequence number are continuous.

[0118] Cross-data pragmatics call refers to the data pragmatics in the index network supporting the call of one data pragmatics to another data pragmatics, i.e. cross-data pragmatics call. It can be a multi-point data pragmatics calling a single-point data pragmatics, or a multi-point data pragmatics calling another multi-point data pragmatics.

[0119] Data pragmatics meta information includes: data pragmatics ID, data pragmatics name, data pragmatics project name, data pragmatics state version number, cross-data pragmatics call random variable seed in data pragmatics that supports cross-data pragmatics calls, and the value of the current cross-data pragmatics call random variable. Whether it is a stateful data pragmatics or a stateless data pragmatics, the data pragmatics meta information needs to be consistent.

[0120] Compared with full state data, incremental state data is a record of incremental changes in the state of multi-point data pragmatics, which cannot reflect all the state information of a multi-point data pragmatics at a certain moment. An incremental state data can only record the change process of multi-point data pragmatics from one state to another.

[0121] The specific process of incremental status data recording is as follows:

[0122] (1) The user initiates a call;

[0123] (2) Start incremental state recording at the same time as starting data pragmatic execution;

[0124] (3) Record the data pragmatics calling method and the incoming parameters;

[0125] (4) If the method called by the data pragmatics this time includes a cross-data pragmatics call, the results of calling other data pragmatics in the cross-data pragmatics call need to be recorded;

[0126] (5) After the data pragmatic call is completed, a complete incremental status data will be formed.

[0127] Based on the above theory, in the normal operation mode, since the number of Shurui operator digital object warehouse nodes currently operating normally in the data pragmatics cluster accounts for the majority, and the situation where all Shurui operator digital object warehouse nodes crash at the same time and cause the loss of memory state data can be ignored, the performance factor of multi-point data pragmatics calls should be given priority. In the normal operation mode, all memory state data of multi-point data pragmatics are in memory. In the normal operation mode, each Shurui operator digital object warehouse node does not need to perform additional operations. During the execution of multi-point data pragmatics, changes in transaction requests will cause changes in memory state data. All state data of multi-point data pragmatics are in memory. This method is the record of multi-point data pragmatics memory state data running in memory persistence mode, which is equivalent to the storage of multi-point data pragmatics memory state data. Even if memory persistence has the problem of easy loss of memory state data, because the number of Shurui operator digital object warehouse nodes currently operating normally in the data pragmatics cluster accounts for the majority, the correctness of the memory state data of multi-point data pragmatics can still be guaranteed. After the currently abnormally operating Shurui Operator Digital Object Warehouse node resumes normal operation, it can obtain the correct memory status data based on the normally operating Shurui Operator Digital Object Warehouse node to ensure the normal operation of multi-point data semantics.

[0128] In the stable operation mode, since there are not many Shurui operator digital object warehouse nodes that are currently operating normally in the data pragmatics cluster, the availability of the data pragmatics cluster is given priority, so each Shurui operator digital object warehouse node uses hard disk persistence for storage. Generally, when the Shurui operator digital object warehouse node is stored, it is necessary to record the memory status data based on other programs or methods to realize the hard disk storage of the memory status data. That is, while the multi-point data pragmatics is executed, the memory status data of the data pragmatics needs to be stored on the hard disk. If a Shurui operator digital object warehouse node loses its memory status data, the Shurui operator digital object warehouse node may recover the memory status data from the hard disk. Of course, it can also obtain the correct memory status data based on the normally operating Shurui operator digital object warehouse node to ensure the normal operation of the multi-point data pragmatics.

[0129] In an embodiment of the present invention, the specific method for each digital object warehouse node of the operator to store data in a hard disk persistent manner is: recording the full status data or incremental status data of the multi-point data language to its own hard disk; when the number of records of the incremental status data of the multi-point data language reaches a second preset number, updating and recording the full status data of the multi-point data language, and deleting the incremental status data records of the multi-point data language at the same time.

[0130] Considering the working efficiency of the data pragmatics cluster, in general, the data size of the full state data is larger, while the data size of the incremental state data is smaller. And recording the full state data in the hard disk also requires calling the corresponding method to collect the various variables and their values ​​in the data pragmatics memory. Therefore, after the Shurui operator digital object warehouse node records the full state data to its own hard disk storage at a certain moment, when an incremental state data appears later, only the incremental state data is recorded to the hard disk storage, which improves the working efficiency of the data pragmatics cluster. When the number of records of the incremental state data reaches the second preset number, the latest full state data is recorded to the hard disk storage, and the previously stored incremental state data records are deleted. For example: the second preset number is 20. Then, after the Shurui operator digital object warehouse node records the full status data to its own hard disk storage at a certain moment, when the first incremental status data appears subsequently, only the first incremental status data is recorded to the hard disk storage. When the second incremental status data appears, the second incremental status data is recorded to the hard disk storage, and so on, until the 20th incremental data appears, the latest full status data is recorded to the hard disk storage, and the previously stored 20 incremental status data records and the old full status data records are deleted.

[0131] It should also be noted that for stateless data pragmatics, since the result of each call is irrelevant to the historical state, if the data pragmatics cluster runs in stable mode, it is meaningless for the stateless data pragmatics to use full state data plus incremental state data to record the hard disk storage, which will also bring additional overhead to the data pragmatics cluster. Therefore, for stateless data pragmatics, the data pragmatics cluster always runs in normal mode, not in stable mode.

[0132] After all the digital object warehouse nodes of the data pragmatics cluster start running based on the above method, the master digital object warehouse node generally receives the call request sent by the user. After receiving the call request, the master digital object warehouse node will immediately execute the call request. At the same time, the master digital object warehouse node also needs to broadcast the call request to all the slave digital object warehouse nodes in the data pragmatics cluster. After receiving the call request, all the slave digital object warehouse nodes in the data pragmatics cluster will immediately execute it. After the execution is completed, they will feed back their respective call results to the master digital object warehouse node.

[0133] Step 102: The master Shurui operator digital object repository node receives all call results fed back by the slave Shurui operator digital object repository nodes.

[0134] In the embodiment of the present invention, the master digital object warehouse node will receive the call results fed back by all slave nodes, and of course the call results of the master digital object warehouse node itself. The master digital object warehouse node will compare all the call results.

[0135] Step 103: When the call result fed back by the first slave digital object warehouse node among all slave digital object warehouse nodes is erroneous, the master digital object warehouse node sends the full status data of the multi-point data application running on itself to the first slave digital object warehouse node.

[0136] In an embodiment of the present invention, when the master digital object warehouse node compares all call results, if it is found that the call result fed back by the first slave digital object warehouse node is wrong, the reason for the wrong call result may be that the data pragmatics of the multi-point data pragmatics running on the first slave digital object warehouse node crashes, or the sequence number of the cross-data pragmatics call is inconsistent with that of other digital object warehouse nodes, or the current state version number of the multi-point data pragmatics is inconsistent with that of other digital object warehouse nodes. At this time, it means that the memory state data of the multi-point data pragmatics running on the first slave digital object warehouse node is wrong, then the master digital object warehouse node will send the full state data of the multi-point data pragmatics running on itself to the first slave digital object warehouse node. In this way, the memory state data of the multi-point data pragmatics running on the first slave digital object warehouse node is restored.

[0137] Step 104: When the master operator digital object repository node executes a call request and determines that its own call result is wrong, the master operator digital object repository node sends a status data request to the second slave operator digital object repository node among all slave operator digital object repository nodes, and the second slave operator digital object repository node is any slave operator digital object repository node among all slave operator digital object repository nodes whose call results are correct.

[0138] In the embodiment of the present invention, when the master digital object warehouse node compares all the call results, if it is found that its own call result is wrong, the reason for the call result error is the same as that of the first slave digital object warehouse node, which may be caused by the data pragmatics crash of the multi-point data pragmatics running on the master digital object warehouse node, or the sequence number of the cross-data pragmatics call may be inconsistent with that of other digital object warehouse nodes, or the current state version number of the multi-point data pragmatics may be inconsistent with that of other digital object warehouse nodes. At this time, it means that the memory state data of the multi-point data pragmatics running on the master digital object warehouse node is wrong, and the master digital object warehouse node needs to send a state data request to any slave digital object warehouse node with a correct call result. For example: if any slave node with a correct call result is the second slave node, then the master digital object warehouse node sends a state data request to the second slave digital object warehouse node.

[0139] Step 105: The master digital object repository node receives the full state data of the multi-point data application running on the second slave digital object repository node sent by the second slave digital object repository node.

[0140] In the embodiment of the present invention, after receiving the status data request, the second slave operator digital object warehouse node will send the full status data of the multi-point data application running on itself to the master operator digital object warehouse node. In this way, the memory status data of the multi-point data application running on the master operator digital object warehouse node is restored.

[0141] Step 106: When the first slave operator digital object warehouse node goes offline and comes back online, the first slave operator digital object warehouse node sends recovery information to the master operator digital object warehouse node. The master operator digital object warehouse node sends the full status data of its own multi-point data pragmatics to the first slave operator digital object warehouse node, or sends the incremental status data of its own multi-point data pragmatics to the first slave operator digital object warehouse node, or does not send any data to the first slave operator digital object warehouse node, based on the recovery information and the current operation mode of the data pragmatics cluster.

[0142] In the embodiment of the present invention, during the actual operation of multi-point data pragmatics, the master digital object warehouse node or the slave digital object warehouse node may be offline due to various factors. For example, a certain digital object warehouse node may be delayed too long due to network congestion. Although the digital object warehouse node is not actually physically offline, other digital object warehouse nodes may think that the digital object warehouse node is offline. Or the digital object warehouse node may be physically offline due to factors such as downtime or power failure. Or the management process of the digital object warehouse node stops running, but the digital object is still running. Although the digital object warehouse node is not actually physically offline, other digital object warehouse nodes may think that the digital object warehouse node is offline.

[0143] For any slave operator digital object warehouse node in the data pragmatics cluster, for example, the first slave operator digital object warehouse node, when the first slave operator digital object warehouse node goes offline and comes back online, the first slave operator digital object warehouse node will actively send recovery information to the master operator digital object warehouse node, and the master operator digital object warehouse node will send the full status data of the multi-point data pragmatics running on itself to the first slave operator digital object warehouse node, or send the incremental status data of the multi-point data pragmatics running on itself to the first slave operator digital object warehouse node, or not send any data to the first slave operator digital object warehouse node according to the recovery information and the current operation mode of the data pragmatics cluster.

[0144] Specifically: the recovery information sent by the first slave operator digital object warehouse node includes: the operation mode of the data pragmatics cluster before the first slave operator digital object warehouse node goes offline, and the version number of the multi-point data pragmatics running after the first slave operator digital object warehouse node comes back online. The first case is: before the first slave operator digital object warehouse node goes offline, the data pragmatics cluster is in the normal operation mode, and after the first slave operator digital object warehouse node comes back online, the data pragmatics cluster is currently in any operation mode, that is, when the first slave operator digital object warehouse node goes offline, the data pragmatics cluster is in the normal operation mode, and after the first slave operator digital object warehouse node comes back online, the master operator digital object warehouse node determines that the data pragmatics cluster is currently in any operation mode, then the master operator digital object warehouse node only needs to determine whether the version number of the multi-point data pragmatics running after the first slave operator digital object warehouse node comes back online is the same as the current state version number of the data pragmatics cluster.

[0145] If the digital object of the multi-point data application on the first slave digital operator digital object warehouse node has not crashed, but only the management process has crashed, or other digital operator digital object warehouse nodes think that the first slave digital operator digital object warehouse node is offline due to network delay, then the recovery information sent by the first slave digital operator digital object warehouse node to the master digital operator digital object warehouse node carries the version number of the multi-point data application running after the first slave digital operator digital object warehouse node is back online; if the digital object of the multi-point data application on the first slave digital operator digital object warehouse node has crashed, then the recovery information sent by the first slave digital operator digital object warehouse node to the master digital operator digital object warehouse node will not carry the version number of the multi-point data application running after the first slave digital operator digital object warehouse node is back online.

[0146] After the master operator digital object warehouse node receives the recovery information, if the recovery information carries the version number of the multi-point data pragmatics running after the first slave operator digital object warehouse node is back online, it can determine whether the version number of the multi-point data pragmatics running after the first slave operator digital object warehouse node is back online is the same as the current status version number of the data pragmatics cluster; if the recovery information does not carry the version number of the multi-point data pragmatics running after the first slave operator digital object warehouse node is back online, the master operator digital object warehouse node directly determines that the version number of the multi-point data pragmatics running after the first slave operator digital object warehouse node is back online is not the same as the current status version number of the data pragmatics cluster.

[0147] If the version number of the multi-point data pragmatics running after the first slave operator digital object warehouse node is back online is the same as the current state version number of the data pragmatics cluster, the master operator digital object warehouse node does not send any data to the first slave operator digital object warehouse node. This is because if the current state version number of the multi-point data pragmatics running after the first slave operator digital object warehouse node is back online is the same as the current state version number of the data pragmatics cluster, it means that during the period when the first slave operator digital object warehouse node was offline and back online, the multi-point data pragmatics of the data pragmatics cluster did not make any call requests, so the first slave operator digital object warehouse node does not need to restore the multi-point data pragmatics memory state data.

[0148] If the master operator digital object warehouse node determines that the current status version number of the multi-point data application running after the first slave operator digital object warehouse node is back online is different from the current status version number of the data application cluster, the master operator digital object warehouse node will send the full status data of the multi-point data application running by itself to the first slave operator digital object warehouse node. This is because if the current status version number of the multi-point data pragmatics running after the first slave operator digital object warehouse node is back online is different from the current status version number of the data pragmatics cluster, it means that during the period when the first slave operator digital object warehouse node was offline and back online, the multi-point data pragmatics of the data pragmatics cluster made at least one call request, and since the first slave operator digital object warehouse node was in normal operation mode before going offline, there is no status data record in the hard disk of the first slave operator digital object warehouse node. Therefore, the master operator digital object warehouse node sends the full status data of the multi-point data pragmatics running by itself to the first slave operator digital object warehouse node, so that the first slave operator digital object warehouse node can run its own multi-point data pragmatics based on the received full status data, and can restore the memory status data of the multi-point data pragmatics running on the first slave operator digital object warehouse node.

[0149] The second situation: before the first slave operator digital object warehouse node goes offline, the data pragmatics cluster is in stable operation mode, and after the first slave operator digital object warehouse node comes back online, the data pragmatics cluster is in normal operation mode. Since the full status data of the multi-point data pragmatics run by the master operator digital object warehouse node is definitely correct at this time, the master operator digital object warehouse node only needs to send the full status data of the multi-point data pragmatics run by itself to the first slave operator digital object warehouse node, so that the first slave operator digital object warehouse node can run its own multi-point data pragmatics based on the received full status data.

[0150] The third situation: before the first slave operator digital object warehouse node goes offline, the data pragmatics cluster is in stable operation mode, and after the first slave operator digital object warehouse node comes back online, the data pragmatics cluster is in stable operation mode, the first slave operator digital object warehouse node obtains the full state data and incremental state data of the multi-point data pragmatics from its own hard disk. At the same time, the master operator digital object warehouse node determines whether the version number of the multi-point data pragmatics running after the first slave operator digital object warehouse node comes back online is the same as the current state version number of the data pragmatics cluster. If they are the same, the master operator digital object The warehouse node does not send any data to the first slave operator digital object warehouse node. If not, the master operator digital object warehouse node sends the newly added incremental status data of the multi-point data pragmatics running by itself to the first slave operator digital object warehouse node, so that the first slave operator digital object warehouse node runs its own multi-point data pragmatics based on the received newly added incremental status data and the full status data and incremental status data obtained from its own hard disk. The so-called newly added incremental status data refers to: the incremental status data that is not in the first slave operator digital object warehouse node's own hard disk due to the first slave operator digital object warehouse node being offline.

[0151] In the third case, since the data pragmatics cluster was in a stable operation mode before the first slave Shu Rui operator digital object warehouse node went offline, the first slave Shu Rui operator digital object warehouse node will store the full status data and incremental status data of the multi-point data pragmatics running on its own hard disk. When the first slave Shu Rui operator digital object warehouse node comes back online, since the hard disk stores the full status data and incremental status data before going offline, the first slave Shu Rui operator digital object warehouse node can first restore the full status data and incremental status data from its own hard disk, and then send the recovery information to the master Shu Rui operator digital object warehouse node. After the master Shurui operator digital object warehouse node receives the recovery information, it uses the same method as in the first case above to determine whether the status version number is the same. If they are the same, it means that during the period when the first slave Shurui operator digital object warehouse node went offline and came back online, the data pragmatics cluster did not have incremental status data. If they are not the same, it means that during the period when the first slave Shurui operator digital object warehouse node went offline and came back online, there was new incremental status data in the data pragmatics cluster. Therefore, the master Shurui operator digital object warehouse node needs to send the new incremental status data that is not stored in the hard disk of the first slave Shurui operator digital object warehouse node to the first slave Shurui operator digital object warehouse node.

[0152] Regarding how to dynamically switch the operation mode of the data pragmatics cluster when the first slave Shu Rui operator digital object warehouse node goes offline and comes back online, or when the master Shu Rui operator digital object warehouse node goes offline and comes back online, it will be explained below and will not be elaborated here.

[0153] Step 107: When the master operator digital object warehouse node goes offline and comes back online, and the master operator digital object warehouse node becomes a slave operator digital object warehouse node, and the second slave operator digital object warehouse node becomes a new master operator digital object warehouse node, the master operator digital object warehouse node sends recovery information to the second slave operator digital object warehouse node, and the second slave operator digital object warehouse node sends the full status data of its own multi-point data pragmatics to the master operator digital object warehouse node, or sends the incremental status data of its own multi-point data pragmatics to the master operator digital object warehouse node, or does not send any data to the master operator digital object warehouse node according to the recovery information and the current operation mode of the data pragmatics cluster.

[0154] In the embodiment of the present invention, the aforementioned step 106 illustrates how to restore the memory status data of multi-point data pragmatics after the first slave digital object warehouse node, which is any slave digital object warehouse node, goes offline and comes back online. After the master digital object warehouse node goes offline and comes back online, the second slave node becomes the new master digital object warehouse node, and the master digital object warehouse node becomes the slave digital object warehouse node. How to restore the memory status data of multi-point data pragmatics is essentially the same as the first slave node.

[0155] When the data pragmatics cluster is in normal operation mode before the master operator digital object warehouse node goes offline, and the data pragmatics cluster is currently in any operation mode after the master operator digital object warehouse node comes back online, the second slave operator digital object warehouse node determines whether the version number of the multi-point data pragmatics running after the master operator digital object warehouse node comes back online is the same as the current status version number of the data pragmatics cluster. If they are the same, the second slave operator digital object warehouse node does not send any data to the master operator digital object warehouse node. If they are not the same, the second slave operator digital object warehouse node sends the full status data of the multi-point data pragmatics running by itself to the master operator digital object warehouse node, so that the master operator digital object warehouse node runs its own multi-point data pragmatics based on the received full status data.

[0156] When the data pragmatics cluster is in stable operation mode before the master Shu Rui operator digital object warehouse node goes offline, and the data pragmatics cluster is in normal operation mode after the master Shu Rui operator digital object warehouse node comes back online, the second slave Shu Rui operator digital object warehouse node sends the full status data of its own multi-point data pragmatics to the master Shu Rui operator digital object warehouse node, so that the master Shu Rui operator digital object warehouse node runs its own multi-point data pragmatics based on the received full status data.

[0157] When the data pragmatics cluster is in stable operation mode before the master Shurui operator digital object warehouse node goes offline, and the data pragmatics cluster is in stable operation mode after the master Shurui operator digital object warehouse node is back online, the master Shurui operator digital object warehouse node obtains the full status data and incremental status data of the multi-point data pragmatics from its own hard disk. At the same time, the second slave Shurui operator digital object warehouse node determines whether the version number of the multi-point data pragmatics running after the master Shurui operator digital object warehouse node is back online is the same as the current status version number of the data pragmatics cluster. If they are the same, the second slave Shurui operator digital object warehouse node does not send any data to the master Shurui operator digital object warehouse node. If they are not the same, the second slave Shurui operator digital object warehouse node sends the newly added incremental status data of the multi-point data pragmatics running by itself to the master Shurui operator digital object warehouse node, so that the master Shurui operator digital object warehouse node runs its own multi-point data pragmatics based on the received newly added incremental status data and the full status data and incremental status data obtained from its own hard disk.

[0158] In the actual operation of the data pragmatics cluster, the master digital object warehouse node may also go offline and come back online. This can be divided into two situations: the first is that the master digital object warehouse node goes offline and comes back online, and the master digital object warehouse node becomes a slave digital object warehouse node, and the second slave digital object warehouse node of any slave digital object warehouse node becomes the new master digital object warehouse node; the second is that the master digital object warehouse node goes offline and comes back online, but the master digital object warehouse node is still the master digital object warehouse node.

[0159] For the first case, the steps are as follows:

[0160] Step S1: When the main Shurui operator digital object warehouse node goes offline and has not come back online after being offline for longer than the preset time, other Shurui operator digital object warehouse nodes in the data pragmatics cluster disconnect from the main Shurui operator digital object warehouse node and send a re-election request to the Shurui digital network identity resolution system in the digital network.

[0161] In an embodiment of the present invention, the master digital object warehouse node and the slave digital object warehouse node maintain a connection with each other based on a heartbeat mechanism, and determine whether the other party is offline or online based on the heartbeat mechanism. When the master digital object warehouse node is offline and the offline time exceeds a preset time, for example, the preset time is 5 seconds, then when the master digital object warehouse node is offline and the offline time exceeds 5 seconds and has not been back online, the other digital object warehouse nodes in the data pragmatics cluster, that is, all the slave digital object warehouse nodes will determine that the master digital object warehouse node is offline, and will automatically disconnect from the master digital object warehouse node. At the same time, all the slave digital object warehouse nodes will send a re-election request to the digital network identity resolution system in the data network. The Shurui Digital Network Identity Resolution System is a management node in the digital network. It does not participate in the consensus, storage and other operations of the digital network. It is only responsible for managing the access of the Shurui operator digital object warehouse nodes and the Shurui storage digital object warehouse nodes in the digital network.

[0162] Step S2: after receiving the re-election request sent by all the digital object repository nodes except the master digital object repository node, the digital network identification resolution system selects a second slave digital object repository node from all the digital object repository nodes except the master digital object repository node as the new master digital object repository node;

[0163] Step S3: After the second slave digital object repository node becomes the new master digital object repository node, a connection request is sent to all digital object repository nodes except the master digital object repository node, so that all digital object repository nodes except the master digital object repository node are connected to the second slave digital object repository node.

[0164] In the embodiment of the invention, after receiving the re-election request sent by all the digital object warehouse nodes except the master digital object warehouse node, the digital network identification resolution system will select the second slave digital object warehouse node from all the digital object warehouse nodes except the master digital object warehouse node as the new master digital object warehouse node. Even if the master digital object warehouse node has been back online during this period, because it has been offline and back online, the digital network identification resolution system believes that the master digital object warehouse node has stability problems. Based on the stability of the digital network operation, it will no longer be based on the computer as the master digital object warehouse node.

[0165] After the second slave digital object warehouse node becomes the new master digital object warehouse node, a connection request will be sent to all digital object warehouse nodes except the master digital object warehouse node, so that all digital object warehouse nodes except the master digital object warehouse node are connected to the second slave digital object warehouse node.

[0166] Step S4: When the master digital object repository node comes back online, the second slave digital object repository node sends a connection request to the master digital object repository node, so that the master digital object repository node connects to the second slave digital object repository node as a slave digital object repository node.

[0167] After the second slave digital object warehouse node becomes the new master digital object warehouse node, when the master digital object warehouse node comes back online, the master digital object warehouse node will query the current master digital object warehouse node as the second slave digital object warehouse node, and then the master digital object warehouse node will actively connect to the second slave digital object warehouse node. So that the master digital object warehouse node connects to the second slave digital object warehouse node as a slave digital object warehouse node. Of course, if the master digital object warehouse node has come back online when the second slave digital object warehouse node becomes the new master digital object warehouse node, then after the second slave digital object warehouse node becomes the new master digital object warehouse node, while sending connection requests to all digital object warehouse nodes except the master digital object warehouse node, it will also send a connection request to the master digital object warehouse node.

[0168] Step 108: When the master operator digital object warehouse node goes offline and comes back online, and the master operator digital object warehouse node is still the master operator digital object warehouse node, the master operator digital object warehouse node executes: receiving the call request and executing it, and at the same time the master operator digital object warehouse node broadcasts the call request to all slave operator digital object warehouse nodes in the data pragmatics cluster.

[0169] In an embodiment of the present invention, for the second type of situation: when the main Shurui operator digital object warehouse node goes offline and comes back online within a preset period of time, other Shurui operator digital object warehouse nodes in the data pragmatics cluster will not perceive that the main Shurui operator digital object warehouse node is offline, and will still maintain a connection with the main Shurui operator digital object warehouse node, and the main Shurui operator digital object warehouse node will still be the main Shurui operator digital object warehouse node.

[0170] After the master Shurui operator digital object warehouse node comes back online, the following steps are executed: receiving the call request and executing it, while the master Shurui operator digital object warehouse node broadcasts the call request to all slave Shurui operator digital object warehouse nodes in the data pragmatics cluster.

[0171] In the embodiment of the present invention, if the master digital object warehouse node of the operator goes offline and the offline time does not exceed the preset time, it will go back online. Because its offline time is very short, other digital object warehouse nodes of the operator in the data pragmatics cluster will not perceive that the master digital object warehouse node has gone offline, and still maintain the connection with the master digital object warehouse node, and naturally will not send a re-election request to the digital network identification resolution system. Therefore, after the master digital object warehouse node is back online, regardless of whether the multi-point data pragmatics memory status data it runs is normal, the master digital object warehouse node only needs to execute step 101: receive the call request and execute it, and at the same time, the master digital object warehouse node broadcasts the call request to all slave digital object warehouse nodes in the data pragmatics cluster. That is, after the master digital object warehouse node is back online, it needs to wait for the next call request, and continue to follow steps 102 to 105 to restore its own multi-point data pragmatics memory status data.

[0172] In an embodiment of the present invention, for the running state of the data pragmatics cluster, when the data pragmatics cluster starts running, the master digital object warehouse node can determine the current running mode of the data pragmatics cluster based on whether the total number of digital object warehouse nodes in the data pragmatics cluster and the total number of digital object warehouse nodes currently running normally meet the first preset condition or the second preset condition. During the running process of the data pragmatics cluster, due to various factors, any one or more slave digital object warehouse nodes or the master digital object warehouse node may have problems, causing the relationship between the total number of digital object warehouse nodes in the data pragmatics cluster and the total number of digital object warehouse nodes currently running normally to change, or after the problematic digital object warehouse node is restored, the relationship between the total number of digital object warehouse nodes in the data pragmatics cluster and the total number of digital object warehouse nodes currently running normally changes, which will cause the master digital object warehouse node to determine whether the data pragmatics cluster needs to switch the running mode. Specifically:

[0173] Assuming that when the data pragmatics cluster starts running, the total number of the digital object warehouse nodes in the data pragmatics cluster and the total number of digital object warehouse nodes currently running normally meet the first preset condition, then the master digital object warehouse node determines that the current operation mode of the data pragmatics cluster is the normal operation mode, and the master digital object warehouse node and all slave digital object warehouse nodes use memory persistence to store memory status data. Assuming that the first slave digital object warehouse node will go offline, it is equivalent to that before the first slave digital object warehouse node goes offline, the operation mode of the data pragmatics cluster is the normal operation mode, and the first slave digital object warehouse node stores memory status data in a memory persistence manner.

[0174] When the first slave digital object warehouse node is offline and has not yet been back online, the current cluster data pragmatics still meets the conditions of being in normal operation mode, so the master digital object warehouse node maintains the normal operation mode. In another case, when the first slave digital object warehouse node is offline and has not yet been back online, the current cluster data pragmatics does not meet the conditions of being in normal operation mode, then the master digital object warehouse node will switch the operation mode to stable operation mode.

[0175] When the first slave Shurui operator digital object warehouse node is offline and has not yet been back online, and the current operation mode is the stable operation mode, when the first slave Shurui operator digital object warehouse node is back online, if the total number of Shurui operator digital object warehouse nodes in the data pragmatics cluster and the total number of Shurui operator digital object warehouse nodes currently operating normally meet the conditions of being in the normal operation mode, the master Shurui operator digital object warehouse node will switch the operation mode to the normal operation mode. Naturally, if after the first slave Shurui operator digital object warehouse node is back online, the total number of Shurui operator digital object warehouse nodes in the data pragmatics cluster and the total number of Shurui operator digital object warehouse nodes currently operating normally still do not meet the conditions of being in the normal operation mode, then the master Shurui operator digital object warehouse node maintains the operation mode as the stable operation mode.

[0176] In addition, it should be noted that only after the main Shu Rui operator digital object warehouse node switches the operation mode from the stable operation mode to the normal operation mode, since the normal operation mode is a memory persistence mode, it is no longer necessary to record in a hard disk persistence mode. In order to save the hard disk storage space of the Shu Rui operator digital object warehouse node, when switching to the normal operation mode, all Shu Rui operator digital object warehouse nodes in the data pragmatics cluster need to delete the full status data and incremental status data recorded to the hard disk during the stable operation mode.

[0177] To sum up, through the above-mentioned multiple recovery methods, it is ensured that during the multi-point data pragmatic operation life cycle, if the computer node encounters problems such as digital object crash, downtime, power outage, etc. and returns to normal, the data pragmatic memory status data can be automatically restored. The recovery method of the present invention has high practicality.

[0178] Based on the above-mentioned data pragmatics memory state data recovery method, an embodiment of the present invention also provides a data pragmatics memory state data recovery system, the recovery system includes: multiple data Rui operator digital object warehouse nodes in a digital network, the multiple data Rui operator digital object warehouse nodes include: a master data Rui operator digital object warehouse node, a slave data Rui operator digital object warehouse node; data pragmatics is run on each of the multiple data Rui operator digital object warehouse nodes, the data pragmatics is a multi-point data pragmatics, and the multiple data Rui operator digital object warehouse nodes form a data pragmatics cluster, refer to Figure 2 , shows a block diagram of a master operator digital object warehouse node in a data pragmatic memory state data recovery system according to an embodiment of the present invention, the master operator digital object warehouse node includes:

[0179] A receiving and executing broadcast module is used to receive and execute a call request, and broadcast the call request to all slave operator digital object warehouse nodes in the data pragmatics cluster;

[0180] The result receiving module is used to receive all the call results fed back from the Shurui operator digital object warehouse node;

[0181] A first sending module is used to send the full state data of the multi-point data pragmatics running by itself to the first slave digital object warehouse node when the call result fed back by the first slave digital object warehouse node among all slave digital object warehouse nodes is wrong;

[0182] a data request module, configured to, when executing the call request and determining that its own call result is wrong, send a status data request to a second slave operator digital object repository node among all slave operator digital object repository nodes, wherein the second slave operator digital object repository node is any slave operator digital object repository node among all slave operator digital object repository nodes whose call results are correct;

[0183] A data receiving module, configured to receive full state data of a multi-point data application running on the second slave digital object repository node, sent by the second slave digital object repository node;

[0184] an information receiving and data sending module, used for receiving recovery information sent by the first slave digital object warehouse node when the first slave digital object warehouse node goes offline and comes back online, and sending the full state data of the multi-point data pragmatics running by itself to the first slave digital object warehouse node, or sending the incremental state data of the multi-point data pragmatics running by itself to the first slave digital object warehouse node, or not sending any data to the first slave digital object warehouse node according to the recovery information and the current operation mode of the data pragmatics cluster;

[0185] The information sending and data receiving module is used to send the recovery information to the second slave digital object warehouse node when the module goes offline and comes back online, and the module becomes a slave digital object warehouse node, and the second slave digital object warehouse node becomes a new master digital object warehouse node, and receive the full state data or incremental state data of the multi-point data pragmatics of the second slave digital object warehouse node running, which is sent by the second slave digital object warehouse node according to the recovery information and the running mode;

[0186] The execution module is used to execute the steps of receiving a call request and executing it when it goes offline and comes back online, and when it is still the master digital object warehouse node: broadcasting the call request to all slave digital object warehouse nodes in the data pragmatics cluster.

[0187] Optionally, the main digital operator digital object warehouse node further includes: an operation mode determination module, the operation mode determination module is used to:

[0188] In the case where the multi-point data pragmatics is classified as a stateless data pragmatics, determining the operation mode to be a normal operation mode;

[0189] When the multi-point data pragmatics is classified as stateful data pragmatics, and the total number of the digital object repository nodes of the data pragmatics cluster is between the total number of the digital object repository nodes of the currently normally operating data pragmatics, and a first preset condition is satisfied, determining that the operation mode is the normal operation mode;

[0190] When the multi-point data pragmatics is classified as stateful data pragmatics, and the total number of digital object warehouse nodes of the data pragmatics cluster is between the total number of digital object warehouse nodes of the currently normally operating data pragmatics, and a second preset condition is satisfied, determining that the operation mode is the stable operation mode;

[0191] The master digital object repository node and the slave digital object repository node both include: an operation record module, and the operation record module is used to:

[0192] In the case of the normal operation mode, the system runs in a memory persistence mode;

[0193] In the case of the stable operation mode, the full state data and the incremental state data of the multi-point data language are recorded in their own hard disks respectively;

[0194] When the number of records of the incremental status data of the multi-point data language reaches a preset number, the full status data of the multi-point data language is updated and recorded respectively, and the incremental status data records of the multi-point data language are deleted respectively;

[0195] The first preset condition is that the total number of the currently operating Shurui operator digital object warehouse nodes is greater than the total number of the Shurui operator digital object warehouse nodes in the data pragmatics cluster after a preset operation is performed, and the preset operation is: taking half of the total number of the Shurui operator digital object warehouse nodes in the data pragmatics cluster to obtain a first result, and rounding up the first result;

[0196] The second preset condition is that the total number of the Shurui operator digital object warehouse nodes that are currently operating normally is not greater than the total number after a preset calculation is performed on the total number of the Shurui operator digital object warehouse nodes in the data pragmatics cluster.

[0197] Optionally, the recovery information includes: the operation mode of the data pragmatics cluster before the first slave Shurui operator digital object warehouse node goes offline, and the version number of the multi-point data pragmatics running after the first slave Shurui operator digital object warehouse node comes back online;

[0198] The module for receiving information and sending data is specifically used for:

[0199] In the case that the data pragmatics cluster is in the normal operation mode before the first slave operator digital object warehouse node goes offline, and the data pragmatics cluster is currently in any operation mode after the first slave operator digital object warehouse node comes back online, determine whether the version number of the multi-point data pragmatics running after the first slave operator digital object warehouse node comes back online is the same as the current state version number of the data pragmatics cluster; if they are the same, no data is sent to the first slave operator digital object warehouse node; if they are not the same, the full state data of the multi-point data pragmatics running by itself is sent to the first slave operator digital object warehouse node, so that the first slave operator digital object warehouse node runs its own multi-point data pragmatics based on the received full state data;

[0200] Before the first slave operator digital object warehouse node goes offline, the data pragmatics cluster is in the stable operation mode, and after the first slave operator digital object warehouse node comes back online, the data pragmatics cluster is in the normal operation mode, sending the full state data of the multi-point data pragmatics running by itself to the first slave operator digital object warehouse node, so that the first slave operator digital object warehouse node runs its own multi-point data pragmatics based on the received full state data;

[0201] When the data pragmatics cluster is in the stable operation mode before the first slave operator digital object warehouse node goes offline, and the data pragmatics cluster is in the stable operation mode after the first slave operator digital object warehouse node comes back online, the first slave operator digital object warehouse node obtains the full state data and incremental state data of the multi-point data pragmatics from its own hard disk, and at the same time, determines whether the version number of the multi-point data pragmatics running after the first slave operator digital object warehouse node comes back online is the same as the current state version number of the data pragmatics cluster. If they are the same, no data is sent to the first slave operator digital object warehouse node. If they are not the same, the newly added incremental state data of the multi-point data pragmatics running by itself is sent to the first slave operator digital object warehouse node, so that the first slave operator digital object warehouse node runs its own multi-point data pragmatics based on the received newly added incremental state data and the full state data and incremental state data obtained from its own hard disk. The newly added incremental state data is the incremental state data that is not in the first slave operator digital object warehouse node's own hard disk due to the first slave operator digital object warehouse node being offline.

[0202] Optionally, the recovery information also includes: the operation mode of the data pragmatics cluster before the master data operator digital object warehouse node goes offline, and the version number of the multi-point data pragmatics running after the master data operator digital object warehouse node comes back online;

[0203] The sending information receiving data module is specifically used for:

[0204] Sending the recovery information to the second slave operator digital object repository node;

[0205] receiving the full state data of the multi-point data pragmatics sent by the second slave operator digital object repository node, and running its own multi-point data pragmatics based on the received full state data;

[0206] Alternatively, the full status data and incremental status data of the multi-point data pragmatics are obtained from its own hard disk, and at the same time, the newly added incremental status data of the multi-point data pragmatics sent by the second slave operator digital object warehouse node is received, and based on the received newly added incremental status data and the full status data and incremental status data obtained from its own hard disk, its own multi-point data pragmatics is run.

[0207] Based on the above-mentioned data pragmatics memory state data recovery method, an embodiment of the present invention also provides a data networking system, the data networking system includes: a data networking identification resolution system, multiple data storage digital object warehouse nodes, and multiple data operator digital object warehouse nodes; the data networking identification resolution system is used to manage multiple data operator digital object warehouse nodes and multiple data storage digital object warehouse nodes; for example: annotate any server or user device into the data networking system as a data operator digital object warehouse node; the multiple data operator digital object warehouse nodes are used to execute the data pragmatics memory state data recovery method as described in any one of steps 101 to 108; the at least one data storage digital object warehouse node is used to store the multi-point data pragmatics operation results sent by the multiple data operator digital object warehouse nodes.

[0208] In order to verify the practicability of the recovery method of the present invention, experimental testing and evaluation are performed on the recovery method of multi-point data pragmatics.

[0209] One device is used as a test server, one device is used as the Shurui data networking identification resolution system, and seven devices are used as Shurui operator digital object warehouse nodes to simulate the Peking University Shurui big data networking system. The Peking University Shurui big data networking system is a typical memory-based data pragmatics big data networking system. A simple key-value pair application is used as an example of a multi-point data pragmatics. The simple key-value pair application is a data pragmatics used to create users and transfer money between users. You can also use any currently known multi-point data pragmatics for experiments.

[0210] For testing the availability of multi-point data pragmatics fault recovery, the overall idea is as follows:

[0211] The fault recovery mechanism of the Shurui operator digital object warehouse node is tested to see whether it can operate normally by means of a crash sequence. N represents the number of all Shurui operator digital object warehouse nodes in the data pragmatics cluster, M represents the number of all normally operating Shurui operator digital object warehouse nodes in the data pragmatics cluster, and Q represents the number of consistent results returned to the main Shurui operator digital object warehouse node by a certain data pragmatics call, which is considered to be a correct call result at that time. For each item in the crash sequence, a data pragmatics call request is initiated to any online Shurui operator digital object warehouse node in the data pragmatics cluster at that time, and the execution result is recorded. After the entire "crash sequence" is executed, the execution results of each data pragmatics call request recorded are monitored to determine whether the node has been successfully recovered and is available.

[0212] Take the crash sequence of the 7-Shurui Operator Digital Object Warehouse node cluster "1234567->123->12345->234->2345" as an example, where "1234567" means that all 7 Shurui Operator Digital Object Warehouse nodes are online, and "123" means that only the 3 Shurui Operator Digital Object Warehouse nodes "123" are online. At this time, the Shurui Operator Digital Object Warehouse node "4567" stops working, and then from "123" to "12345", the Shurui Operator Digital Object Warehouse node "45" is restarted.

[0213] According to the experimental requirements, the test environment is set up and the program is deployed. Then, availability tests are carried out on stateful data pragmatics and stateless data pragmatics in the 5-Shurui operator digital object warehouse node cluster and the 7-Shurui operator digital object warehouse node cluster respectively.

[0214] It is also necessary to generate a crash sequence file. Taking the generation of a crash sequence of a data pragmatics cluster of 7 Shurui Operator Digital Object Warehouse nodes as an example, write and generate a crash sequence. When generating each new sequence, randomly select some currently online Shurui Operator Digital Object Warehouse nodes to stop working, or randomly select some currently crashed Shurui Operator Digital Object Warehouse nodes to resume working. Set the sequence length to 700, and finally generate a crash sequence of length 700, of which 665 sequences can execute data pragmatics call requests. This crash sequence covers the paths of different recovery methods after all Shurui Operator Digital Object Warehouse nodes are back online. Some of the sequences are shown in the following table:

[0215]

[0216] It is also necessary to generate a call request file, generate different requests for different data pragmatics and write them. Taking a simple key-value pair application as an example, when generating a request, randomly select methods for creating users, setting initial account balances, transferring money, querying the balance of a certain account, querying the balance of all accounts, querying the total balance, etc. and generate call parameters based on the selected method, separate the method name and parameters with a semicolon, and a request is generated.

[0217] Pragmatic usability testing for stateful data:

[0218] In this section, two stateful data pragmatics will be tested for availability: one is the counter counting data pragmatic, which contains two methods, one for incremental calculation and the other for accumulating the length of the passed parameter. The state volume of this data pragmatic is small; the other is a simple key-value pair application transfer. The state volume of this data pragmatic increases with the increase in the number of accounts, and its state volume is larger than that of the counter data pragmatic.

[0219] 1) Counter counting data pragmatics test

[0220] The tests were carried out in a 5-Shurui operator digital object warehouse node cluster and a 7-Shurui operator digital object warehouse node cluster respectively, and the data pragmatics call request results recorded by the test program were obtained.

[0221] Use other local laptops to start a single-point counter counting data pragmatics, execute it through the same request file, and record the single-point execution results. Use this as a standard to compare the number of correct, incorrect, and unavailable results in the 665 data pragmatics call requests executed by the counter counting data pragmatics in the 5-Shu Rui operator digital object warehouse node cluster and the 7-Shu Rui operator digital object warehouse node cluster.

[0222] By writing a program to compare the results, we can obtain the results of the counter counting data pragmatics execution in the 5-shurui operator digital object warehouse node cluster and the 7-shurui operator digital object warehouse node cluster, as shown in the following table:

[0223]

[0224] 2) Simple key-value pair application test

[0225] The same method is used to test a simple key-value pair application, and the data pragmatics call request results recorded by the test program are obtained.

[0226] By writing a program to compare the results, we can obtain the results of data pragmatic execution of simple key-value pairs in the 5-shurui operator digital object warehouse node cluster and the 7-shurui operator digital object warehouse node cluster, as shown in the following table:

[0227]

[0228] 3) Pragmatic usability testing of stateless data

[0229] In this section, we will test the availability of the data pragmatics mathOperation. This data pragmatics is a stateless data pragmatics responsible for the four arithmetic operations of addition, subtraction, multiplication and division. Its execution result is only related to the input parameters.

[0230] The tests were carried out in a 5-Shurui operator digital object warehouse node cluster and a 7-Shurui operator digital object warehouse node cluster respectively, and the data pragmatics call request results recorded by the test program were obtained.

[0231] By writing a program to compare the results, the results of the mathOperation data pragmatics execution in the 5-shurui operator digital object warehouse node cluster and the 7-shurui operator digital object warehouse node cluster are obtained, as shown in the following table:

[0232]

[0233] The above is a test of the availability of multi-point data pragmatics, and the following is a test of the fault recovery time of multi-point data pragmatics.

[0234] The fault recovery time test of multi-point data pragmatics includes multiple situations: the fault recovery time of data pragmatics with different state sizes under the same network delay through three methods: recovery based on the normal operation mode (recoverFromCommon), recovery based on the stable operation mode (recoverFromStable), and recovery that requires the master digital object warehouse node to send new incremental state data (recoverFromStable (need trans from master)); the fault recovery time of data pragmatics with the same state size under these three methods under different network delays; the election time of a new master digital object warehouse node after the master digital object warehouse node crashes when using different heartbeat packet periods under the same network delay; the election time of a new master digital object warehouse node after the master digital object warehouse node crashes under different network delays under the same heartbeat packet period.

[0235] When testing the fault recovery time of the Shurui operator digital object warehouse node, the test program will repeat the same fault 10 times, record the recovery time of the 10 times, and the final fault recovery time is expressed as the average of the 10 fault recovery times. Use simple key-value pairs with different numbers of accounts to simulate data pragmatics with different state sizes for testing. First, start a single-point simple key-value pair application through a local laptop, send a request to the local simple key-value pair application through a program that automatically executes requests and saves states, and automatically save the state files of the data pragmatics under 2000 users, 4000 users, 6000 users...50000 users. And transmit these state files one by one to the trusted Shurui operator digital object warehouse node of the server.

[0236] Set the multi-point data pragmatics to automatically load the state file after starting. This can simulate multi-point data pragmatics with different state sizes. Similar to the aforementioned cluster availability test, a crash sequence is also generated when testing the fault recovery time. The length of the sequence is 10. For each crash sequence, a data pragmatics call request is sent to determine whether it can be recovered normally. The data pragmatics call requests here are some transfer and balance query requests.

[0237] In the data pragmatics fault recovery mechanism based on memory computing in this paper, there are two main fault recovery methods after the node is back online: one is to dump the full state data of the current multi-point data pragmatics through other Data Rui Operator Digital Object Warehouse nodes and send it to the faulty node for recovery; the other is to recover through the full state data and incremental state data of the multi-point data pragmatics recorded in the local disk. If the Data Rui Operator Digital Object Warehouse node recovers through the full state data and incremental state data of the multi-point data pragmatics recorded in the local disk, if the node still lags after recovering the data through the local hard disk, for example, during the crash, the data pragmatics cluster executed a call request, then after the Data Rui Operator Digital Object Warehouse node recovers through the local state data, it needs to obtain the records of the incremental state data that were missed during the crash from the main Data Rui Operator Digital Object Warehouse node to recover.

[0238] The above method is used to test the three recovery methods for fault recovery under the same network delay and different state size data. The results are as follows: Figure 3 The results shown, Figure 3 The horizontal axis is the number of users using the data, and the vertical axis is the time it takes to recover the data, in milliseconds (ms). Figure 3The darkest color is the recoverFromCommon method, the darker color is the recoverFromStable method, and the lightest color is the recoverFromStable (need trans from master) method. It can be seen that under the same network delay and the same number of data users, the recoverFromStable method takes the least time to recover data, the recoverFromStable (need trans from master) method takes a little longer to recover data, and the recoverFromCommon method takes the longest time to recover data.

[0239] Test the data pragmatic failure recovery time when the network delay is different.

[0240] This section tests the impact of network delay on data pragmatics failure recovery time, using a simple key-value pair application with 10,000 accounts as the test object. The TCP packet delay time of all trusted database operator digital object warehouse nodes is set by writing a script to simulate network delay.

[0241] Through the above method, the data pragmatics with the same state quantity size under different TCP packet delays are tested for fault recovery and the following results are obtained: Figure 4 Results shown. Figure 4 The curve composed of diamonds and solid lines is the recoverFromCommon method, the curve composed of triangles and solid lines is the recoverFromStable method, and the curve composed of squares and solid lines is the recoverFromStable (need trans from master) method. It can be seen that under different TCP packet delays, the recoverFromStable method takes the least time to recover data, the recoverFromStable (need trans from master) method takes a little longer to recover data, and the recoverFromCommon method takes the longest time to recover data.

[0242] Test the election time of the main operator digital object warehouse node when the heartbeat packet cycle is different.

[0243] Because different heartbeat packet periods will cause different times for the Shurui operator digital object warehouse node to detect the crash of the main Shurui operator digital object warehouse node, this section will test the impact of different heartbeat packet periods on the time it takes to elect a new main Shurui operator digital object warehouse node and the time it takes for the new main Shurui operator digital object warehouse node to complete cluster information data recovery after the main Shurui operator digital object warehouse node crashes. Change the heartbeat packet period by setting the heartbeat packet and fault detection period. After setting the new heartbeat packet period, repeat the aforementioned deployment steps for the trusted Shurui operator digital object warehouse node data pragmatics engine. Then run the corresponding test program for testing, and get the results shown in the following table:

[0244]

[0245] A test on the election time of the digital object warehouse node of the master operator when the network delay is different.

[0246] Because the monitoring of the crash by the slave digital object warehouse node after the master digital object warehouse node crashes is affected by the network environment, and the election process of the new master digital object warehouse node and the cluster data recovery of the new master digital object warehouse node require multiple network communications to interact, so this section will test the effect of network delay on the recovery time of the new master digital object warehouse node. Set the heartbeat packet period to 2000ms, use the above method to simulate different network environments, and get the results shown in the following table:

[0247]

[0248] In summary, in the same network environment, the fault recovery time of data pragmatics with different state sizes increases with the increase of state size. Among them, the recoverFromCommon method takes the longest time, and the time increase is the largest with the increase of state size. This is because the recoverFromCommon method requires the master digital object warehouse node to dump the full state data of the current data pragmatics and send it. When the state size continues to increase, the time to save the full state data of the current data pragmatics will gradually increase. Due to the increase of the full state data, the time to send the full state data file will also increase. The recoverFromStable method recovers the state from the local disk, and its performance is better than the recoverFromCommon method. Since recoverFromStable (need trans from master) needs to obtain the incremental state data that was not recorded during the crash from the master digital object warehouse node compared to the recoverFromStable method, its performance is slightly worse than the recoverFromStable method, but better than the recoverFromCommon method.

[0249] For data with the same state size, the recoverFromStable method still has the best recovery performance under different network delay environments. Because the time to send the full state data increases with the increase of network delay, the time increase of recoverFromCommon is the most obvious.

[0250] In the same network environment, different heartbeat packet periods will affect the time it takes for the slave digital object warehouse node to detect the crash of the master digital object warehouse node, and thus affect the time it takes for a new master digital object warehouse node to be elected and restored after the master digital object warehouse node crashes. As the heartbeat packet period increases, the time it takes to elect a new master digital object warehouse node gradually increases.

[0251] For data pragmatics with the same heartbeat packet period, the time it takes for the new master digital object warehouse node to recover is different under different network delays. Network delays will affect the sending of heartbeat packets, and the election and recovery of the new master digital object warehouse node will involve interactions between multiple digital object warehouse nodes. As network delays increase, the recovery time of the new master digital object warehouse node gradually increases.

[0252] For performance testing:

[0253] In this section, we will compare the execution performance of single-point data pragmatics and multi-point data pragmatics through the execution time of data pragmatics requests. Taking a simple key-value application with 30,000 accounts at the beginning as an example, we send requests to query all accounts and balances and to query the balance of a certain account to the single-point data pragmatics and multi-point data pragmatics respectively. The test results are shown in the following table:

[0254]

[0255] Since multi-point data pragmatics needs to broadcast data pragmatics call requests and verify results on the main data warehouse node, the execution performance of single-point data pragmatics is better than that of cluster data pragmatics, and the data pragmatics cluster of 5 data warehouse nodes is better than that of 7 data warehouse nodes. The method of viewing all users and balances takes longer because it needs to view the balances of 30,000 accounts. In summary, the call execution time of multi-point data pragmatics is acceptable.

[0256] Through the above experiments and tests, the practicability of the recovery method of the present invention has been confirmed and is completely feasible.

[0257] The embodiments of the present invention are described with reference to the flowcharts and / or block diagrams of the methods, terminal devices (systems), and computer program products according to the embodiments of the present invention. It should be understood that each process and / or block in the flowchart and / or block diagram, as well as the combination of the processes and / or blocks in the flowchart and / or block diagram, can be implemented by computer program instructions. These computer program instructions can be provided to a processor of a general-purpose computer, a special-purpose computer, an embedded processor, or other programmable data processing terminal device to generate a machine, so that the instructions executed by the processor of the computer or other programmable data processing terminal device generate instructions for implementing the processes in the flowchart and / or block diagram. Figure 1 A process or multiple processes and / or boxes Figure 1 A device that provides the functions specified in a block or multiple blocks.

[0258] Finally, it should be noted that, in this article, relational terms such as first and second, etc. are only used to distinguish one entity or operation from another entity or operation, and do not necessarily require or imply any such actual relationship or order between these entities or operations. Moreover, the terms "include", "comprise" or any other variants thereof are intended to cover non-exclusive inclusion, so that a process, method, article or terminal device including a series of elements includes not only those elements, but also other elements not explicitly listed, or also includes elements inherent to such process, method, article or terminal device. In the absence of further restrictions, the elements defined by the sentence "comprise a ..." do not exclude the existence of other identical elements in the process, method, article or terminal device including the elements.

[0259] The embodiments of the present invention are described above in conjunction with the accompanying drawings. Specific examples are used herein to illustrate the principles and implementation methods of the present invention. The description of the above embodiments is only used to help understand the method of the present invention and its core idea. At the same time, for those skilled in the art, according to the idea of ​​the present invention, there will be changes in the specific implementation methods and application scopes. In summary, the content of this specification should not be understood as limiting the present invention.

Claims

1. A method for recovering data in a data pragmatic memory state, characterized in that: The recovery method is applied to a plurality of operator digital object warehouse nodes in a data network, each of the plurality of operator digital object warehouse nodes runs a data pragmatics, the data pragmatics is a multi-point data pragmatics, and the plurality of operator digital object warehouse nodes form a data pragmatics cluster, and the recovery method includes: The master operator digital object repository node in the data pragmatics cluster receives the call request and executes it, and at the same time, the master operator digital object repository node broadcasts the call request to all slave operator digital object repository nodes in the data pragmatics cluster; The master operator digital object repository node receives the call results fed back by all slave operator digital object repository nodes; In the case that the calling result fed back by the first slave operator digital object repository node among all slave operator digital object repository nodes is wrong, the master operator digital object repository node sends the full state data of the multi-point data pragmatics running by itself to the first slave operator digital object repository node; When the master operator digital object repository node executes the call request and determines that its own call result is wrong, the master operator digital object repository node sends a status data request to a second slave operator digital object repository node among all slave operator digital object repository nodes, where the second slave operator digital object repository node is any slave operator digital object repository node among all slave operator digital object repository nodes whose call results are correct; The master operator digital object repository node receives full state data of a multi-point data application running on the second slave operator digital object repository node sent by the second slave operator digital object repository node; In the case where the first slave operator digital object warehouse node goes offline and comes back online, the first slave operator digital object warehouse node sends recovery information to the master operator digital object warehouse node, and the master operator digital object warehouse node sends the full state data of the multi-point data pragmatics running on itself to the first slave operator digital object warehouse node, or sends the incremental state data of the multi-point data pragmatics running on itself to the first slave operator digital object warehouse node, or does not send any data to the first slave operator digital object warehouse node according to the recovery information and the current operation mode of the data pragmatics cluster; When the master operator digital object warehouse node goes offline and comes back online, and the master operator digital object warehouse node becomes a slave operator digital object warehouse node, and the second slave operator digital object warehouse node becomes a new master operator digital object warehouse node, the master operator digital object warehouse node sends the recovery information to the second slave operator digital object warehouse node, and the second slave operator digital object warehouse node sends the full state data of the multi-point data pragmatics running by itself to the master operator digital object warehouse node, or sends the incremental state data of the multi-point data pragmatics running by itself to the master operator digital object warehouse node, or does not send any data to the master operator digital object warehouse node according to the recovery information and the current operation mode of the data pragmatics cluster; When the master operator digital object warehouse node goes offline and comes back online, and the master operator digital object warehouse node is still the master operator digital object warehouse node, the master operator digital object warehouse node performs the steps of: receiving a call request and executing it, and at the same time, the master operator digital object warehouse node broadcasts the call request to all slave operator digital object warehouse nodes in the data pragmatics cluster.

2. The recovery method according to claim 1, characterized in that: Before the main operator digital object warehouse node in the data pragmatics cluster receives and executes the call request, it also includes: When the classification of the multi-point data pragmatics is a stateless data pragmatics, the master operator digital object repository node determines that the operation mode is a normal operation mode; When the classification of the multi-point data pragmatics by the master operator digital object repository node is stateful data pragmatics, and the total number of operator digital object repository nodes in the data pragmatics cluster is between the total number of operator digital object repository nodes currently operating normally, and a first preset condition is satisfied, the master operator digital object repository node determines that the operating mode is the normal operating mode; When the classification of the multi-point data pragmatics by the master operator digital object repository node is stateful data pragmatics, and the total number of operator digital object repository nodes in the data pragmatics cluster is between the total number of operator digital object repository nodes currently operating normally, and a second preset condition is satisfied, the master operator digital object repository node determines that the operating mode is a stable operating mode; In the case of the normal operation mode, each operator digital object warehouse node in the data pragmatics cluster runs in a memory persistence mode; In the case of the stable operation mode, each operator digital object warehouse node records the full state data and incremental state data of the multi-point data pragmatics to its own hard disk; When the number of records of the incremental state data of the multi-point data pragmatics reaches a preset number, each of the operator digital object repository nodes updates and records the full state data of the multi-point data pragmatics, and deletes the incremental state data records of the multi-point data pragmatics; The first preset condition is that the total number of the currently operating operator digital object warehouse nodes is greater than the total number of the operator digital object warehouse nodes in the data pragmatics cluster after a preset operation is performed, and the preset operation is: taking half of the total number of the operator digital object warehouse nodes in the data pragmatics cluster to obtain a first result, and rounding up the first result; The second preset condition is that the total number of the operator digital object warehouse nodes that are currently operating normally is not greater than the total number after a preset calculation is performed on the total number of the operator digital object warehouse nodes in the data pragmatics cluster.

3. The recovery method according to claim 2, characterized in that: The recovery information includes: the operation mode of the data pragmatics cluster before the first slave operator digital object repository node goes offline, and the version number of the multi-point data pragmatics running after the first slave operator digital object repository node comes back online; In the case where the first slave operator digital object warehouse node goes offline and comes back online, the first slave operator digital object warehouse node sends recovery information to the master operator digital object warehouse node, and the master operator digital object warehouse node sends the full state data of the multi-point data pragmatics running by itself to the first slave operator digital object warehouse node, or sends the incremental state data of the multi-point data pragmatics running by itself to the first slave operator digital object warehouse node, or does not send any data to the first slave operator digital object warehouse node, according to the recovery information and the current operation mode of the data pragmatics cluster, including: In the case that the data pragmatics cluster is in the normal operation mode before the first slave operator digital object warehouse node goes offline, and the data pragmatics cluster is currently in any operation mode after the first slave operator digital object warehouse node comes back online, the master operator digital object warehouse node determines whether the version number of the multi-point data pragmatics running after the first slave operator digital object warehouse node comes back online is the same as the current state version number of the data pragmatics cluster; if they are the same, the master operator digital object warehouse node does not send any data to the first slave operator digital object warehouse node; if they are not the same, the master operator digital object warehouse node sends the full state data of the multi-point data pragmatics running by itself to the first slave operator digital object warehouse node, so that the first slave operator digital object warehouse node runs its own multi-point data pragmatics based on the received full state data; In a case where the data pragmatics cluster is in the stable operation mode before the first slave operator digital object warehouse node goes offline, and the data pragmatics cluster is in the normal operation mode after the first slave operator digital object warehouse node comes back online, the master operator digital object warehouse node sends the full state data of the multi-point data pragmatics running by itself to the first slave operator digital object warehouse node, so that the first slave operator digital object warehouse node runs its own multi-point data pragmatics based on the received full state data; In the case where the data pragmatics cluster is in the stable operation mode before the first slave operator digital object warehouse node goes offline, and the data pragmatics cluster is in the stable operation mode after the first slave operator digital object warehouse node comes back online, the first slave operator digital object warehouse node obtains the full state data and incremental state data of the multi-point data pragmatics from its own hard disk. At the same time, the master operator digital object warehouse node determines whether the version number of the multi-point data pragmatics running after the first slave operator digital object warehouse node comes back online is the same as the current state version number of the data pragmatics cluster. If they are the same, the master operator digital object warehouse node The child digital object warehouse node does not send any data to the first slave operator digital object warehouse node. Otherwise, the master operator digital object warehouse node sends the newly added incremental status data of the multi-point data application running by itself to the first slave operator digital object warehouse node, so that the first slave operator digital object warehouse node runs its own multi-point data application based on the received newly added incremental status data and the full status data and incremental status data obtained from its own hard disk. The newly added incremental status data is the incremental status data that is not in the first slave operator digital object warehouse node's own hard disk due to the first slave operator digital object warehouse node being offline.

4. The recovery method according to claim 2, characterized in that: The recovery information also includes: the operation mode of the data pragmatics cluster before the master operator digital object warehouse node goes offline, and the version number of the multi-point data pragmatics running after the master operator digital object warehouse node comes back online; In the case where the master operator digital object warehouse node goes offline and comes back online, and the master operator digital object warehouse node becomes a slave operator digital object warehouse node, and the second slave operator digital object warehouse node becomes a new master operator digital object warehouse node, the master operator digital object warehouse node sends the recovery information to the second slave operator digital object warehouse node, and the second slave operator digital object warehouse node sends the full state data of the multi-point data pragmatics running by itself to the master operator digital object warehouse node, or sends the incremental state data of the multi-point data pragmatics running by itself to the master operator digital object warehouse node, or does not send any data to the master operator digital object warehouse node according to the recovery information and the current operation mode of the data pragmatics cluster, including: In the case where the data pragmatics cluster is in the normal operation mode before the master operator digital object warehouse node goes offline, and the data pragmatics cluster is currently in any operation mode after the master operator digital object warehouse node comes back online, the second slave operator digital object warehouse node determines whether the version number of the multi-point data pragmatics running after the master operator digital object warehouse node comes back online is the same as the current state version number of the data pragmatics cluster. If they are the same, the second slave operator digital object warehouse node does not send any data to the master operator digital object warehouse node. If they are not the same, the second slave operator digital object warehouse node sends the full state data of the multi-point data pragmatics running by itself to the master operator digital object warehouse node, so that the master operator digital object warehouse node runs its own multi-point data pragmatics based on the received full state data; In a case where the data pragmatics cluster is in the stable operation mode before the master operator digital object warehouse node goes offline, and the data pragmatics cluster is in the normal operation mode after the master operator digital object warehouse node comes back online, the second slave operator digital object warehouse node sends the full state data of the multi-point data pragmatics running by itself to the master operator digital object warehouse node, so that the master operator digital object warehouse node runs its own multi-point data pragmatics based on the received full state data; When the data pragmatics cluster is in the stable operation mode before the master operator digital object warehouse node goes offline, and when the data pragmatics cluster is in the stable operation mode after the master operator digital object warehouse node comes back online, the master operator digital object warehouse node obtains the full state data and incremental state data of the multi-point data pragmatics from its own hard disk. At the same time, the second slave operator digital object warehouse node determines whether the version number of the multi-point data pragmatics running after the master operator digital object warehouse node comes back online is the same as the current state version number of the data pragmatics cluster. If they are the same, the second slave operator digital object warehouse node does not send any data to the master operator digital object warehouse node. If they are not the same, the second slave operator digital object warehouse node sends the newly added incremental state data of the multi-point data pragmatics running by itself to the master operator digital object warehouse node, so that the master operator digital object warehouse node runs its own multi-point data pragmatics based on the received newly added incremental state data and the full state data and incremental state data obtained from its own hard disk.

5. The recovery method according to claim 3, characterized in that: The master operator digital object repository node determines whether the version number of the multipoint data pragmatics running after the first slave operator digital object repository node comes back online is the same as the current state version number of the data pragmatics cluster, including: In the case where the digital object of the multi-point data pragmatics on the first slave operator digital object repository node is not corrupted, the recovery information sent by the first slave operator digital object repository node to the master operator digital object repository node carries the version number of the multi-point data pragmatics running after the first slave operator digital object repository node comes back online; The master operator digital object repository node determines, based on the received version number of the multipoint data pragmatics running after the first slave operator digital object repository node is back online, whether the version number of the multipoint data pragmatics running after the first slave operator digital object repository node is back online is the same as the current state version number of the data pragmatics cluster; In the case where the digital object of the multi-point data pragmatics on the first slave operator digital object repository node crashes, the recovery information sent by the first slave operator digital object repository node to the master operator digital object repository node does not carry the version number of the multi-point data pragmatics running after the first slave operator digital object repository node comes back online; The master operator digital object repository node does not receive the version number of the multi-point data pragmatics running after the first slave operator digital object repository node comes back online, and determines that the version number of the multi-point data pragmatics running after the first slave operator digital object repository node comes back online is different from the current state version number of the data pragmatics cluster.

6. The recovery method according to claim 2, characterized in that: The recovery method further comprises: Before the first slave operator digital object repository node goes offline, if the total number of operator digital object repository nodes in the data pragmatics cluster and the total number of operator digital object repository nodes currently operating normally meet the first preset condition, the master operator digital object repository node determines that the operating mode is the normal operating mode; After the first slave operator digital object repository node goes offline and when the first slave operator digital object repository node has not come back online, if the total number of operator digital object repository nodes in the data pragmatics cluster and the total number of operator digital object repository nodes currently operating normally still meet the first preset condition, then the master operator digital object repository node maintains the operating mode as the normal operating mode; or If the total number of operator digital object warehouse nodes in the data pragmatics cluster and the total number of operator digital object warehouse nodes currently operating normally meet the second preset condition, the master operator digital object warehouse node switches the operating mode to the stable operating mode; When the first slave operator digital object repository node is offline and has not yet come back online, and the operation mode is currently the normal operation mode, after the first slave operator digital object repository node comes back online, if the total number of operator digital object repository nodes in the data pragmatics cluster and the total number of operator digital object repository nodes currently operating normally still meet the first preset condition, then the master operator digital object repository node maintains the operation mode as the normal operation mode; or If the total number of operator digital object warehouse nodes in the data pragmatics cluster and the total number of operator digital object warehouse nodes currently operating normally meet the second preset condition, the master operator digital object warehouse node switches the operating mode to the stable operating mode; When the first slave operator digital object warehouse node goes offline and has not yet come back online, and the operation mode is currently the stable operation mode, when the first slave operator digital object warehouse node comes back online, if the total number of operator digital object warehouse nodes in the data pragmatics cluster and the total number of operator digital object warehouse nodes currently operating normally meet the first preset condition, then the master operator digital object warehouse node switches the operation mode to the normal operation mode; or If the total number of operator digital object warehouse nodes in the data pragmatics cluster and the total number of operator digital object warehouse nodes currently operating normally still meet the second preset condition, the main operator digital object warehouse node maintains the operating mode as the stable operating mode; after the main operator digital object warehouse node switches the operating mode from the stable operating mode to the normal operating mode, all operator digital object warehouse nodes in the data pragmatics cluster delete the full status data and incremental status data recorded to the hard disk during the stable operating mode.

7. The recovery method according to claim 1, characterized in that: The master operator digital object warehouse node goes offline and comes back online, and the master operator digital object warehouse node becomes a slave operator digital object warehouse node, and the second slave operator digital object warehouse node becomes the master operator digital object warehouse node, including: When the master operator digital object warehouse node is offline and has not come back online after a preset period of time, all other operator digital object warehouse nodes in the data pragmatics cluster disconnect from the master operator digital object warehouse node and send a re-election request to the data network identity resolution system in the data network; After receiving the re-election request sent by all operator digital object repository nodes except the master operator digital object repository node, the digital network identity resolution system selects the second slave operator digital object repository node as the new master operator digital object repository node from all operator digital object repository nodes except the master operator digital object repository node; After the second slave operator digital object repository node becomes the new master operator digital object repository node, sending a connection request to all operator digital object repository nodes except the master operator digital object repository node, so that all operator digital object repository nodes except the master operator digital object repository node are connected to the second slave operator digital object repository node; When the master operator digital object repository node comes back online, the second slave operator digital object repository node sends the connection request to the master operator digital object repository node, so that the master operator digital object repository node connects to the second slave operator digital object repository node as a slave operator digital object repository node.

8. The recovery method according to claim 1, characterized in that: The master operator digital object warehouse node goes offline and comes back online, and the master operator digital object warehouse node is still the master operator digital object warehouse node, including: When the master operator digital object warehouse node goes offline and comes back online within a preset period of time, other operator digital object warehouse nodes in the data pragmatics cluster will not sense that the master operator digital object warehouse node is offline, and still maintain connection with the master operator digital object warehouse node; After the master operator digital object warehouse node comes back online, the following steps are performed: receiving a call request and executing it, and at the same time, the master operator digital object warehouse node broadcasts the call request to all slave operator digital object warehouse nodes in the data pragmatics cluster.

9. A data pragmatic memory state data recovery system, characterized in that: The recovery system includes: a plurality of operator digital object warehouse nodes in a digital network, wherein the plurality of operator digital object warehouse nodes include: a master operator digital object warehouse node and a slave operator digital object warehouse node; each of the plurality of operator digital object warehouse nodes runs a data pragmatics, wherein the data pragmatics is a multi-point data pragmatics, and the plurality of operator digital object warehouse nodes form a data pragmatics cluster, wherein the master operator digital object warehouse node includes: A receiving and executing broadcast module, used for receiving and executing a call request, and broadcasting the call request to all slave operator digital object warehouse nodes in the data pragmatics cluster; The result receiving module is used to receive all the call results fed back from the operator digital object warehouse node; A first sending module is used to send the full state data of the multi-point data pragmatics running by itself to the first slave operator digital object repository node when the call result fed back by the first slave operator digital object repository node among all the slave operator digital object repository nodes is wrong; a data request module, configured to, when executing the call request and determining that its own call result is wrong, send a status data request to a second slave operator digital object repository node among all slave operator digital object repository nodes, wherein the second slave operator digital object repository node is any slave operator digital object repository node among all slave operator digital object repository nodes whose call results are correct; A data receiving module, configured to receive full state data of a multi-point data application running on the second slave operator digital object repository node, sent by the second slave operator digital object repository node; an information receiving and data sending module, configured to receive recovery information sent by the first slave operator digital object warehouse node when the first slave operator digital object warehouse node goes offline and comes back online, and according to the recovery information and the current operation mode of the data pragmatics cluster, send the full state data of the multi-point data pragmatics running by itself to the first slave operator digital object warehouse node, or send the incremental state data of the multi-point data pragmatics running by itself to the first slave operator digital object warehouse node, or not send any data to the first slave operator digital object warehouse node; The information sending and data receiving module is used to send the recovery information to the second slave operator digital object warehouse node when the second slave operator digital object warehouse node goes offline and comes back online, and the second slave operator digital object warehouse node becomes the new master operator digital object warehouse node, and receive the full state data or incremental state data of the multi-point data language running by the second slave operator digital object warehouse node sent by the second slave operator digital object warehouse node according to the recovery information and the running mode; The execution module is used to execute the steps of receiving a call request and executing it when it goes offline and comes back online, and when it is still a master operator digital object warehouse node, and broadcasting the call request to all slave operator digital object warehouse nodes in the data pragmatics cluster.

10. A digital networking system, characterized in that: The digital networking system comprises: a digital networking identification resolution system, a plurality of storage digital object warehouse nodes, and a plurality of operator digital object warehouse nodes; The digital network identification resolution system is used to manage the plurality of operator digital object repository nodes and whether the plurality of operator digital object repository nodes join the digital network system; The plurality of operator digital object repository nodes are used to execute the method for recovering data pragmatic memory state data as described in any one of claims 1 to 8; The plurality of operator digital object repository nodes are used to store at least one of the following information: The operating result of the multi-point data pragmatics; Status data of the multi-point data pragmatics; The multiple data points include execution branch information.

Images