Real-time event processing method and system based on cloud computing in computer software system

Abstract

The invention discloses a real-time event processing method based on cloud computing in a computer software system, which relates to the technical field of cloud computing and comprises the following steps: constructing an initial interaction graph; optimizing the initial interaction atlas, and updating the optimized interaction atlas to obtain a deep interaction atlas; determining a plurality of main event processing nodes and determining initial behaviors of the main event processing nodes according to the target state information of the deep interaction graph; the secondary event processing nodes under the main event processing node share the initial behavior, the secondary event processing nodes determine the second behavior of the secondary event processing nodes according to the initial behavior and the target state information, output events generated by repeated events are removed, the uniqueness and the processing speed of the output events are ensured, the event processing nodes are dynamically adjusted, and the processing speed of the event processing nodes is improved. When the event processing node fails, a normal node is selected from the corresponding sub-cluster for processing, so that high availability of the whole event processing node group is ensured, and continuity of an upper layer service is ensured.

Description

Technical Field

[0001] The present invention relates to the technical field of cloud computing, and more specifically to a real-time event processing method and system based on cloud computing in a computer software system. Background Art

[0002] The capture, real-time processing, and quasi-real-time analysis of massive events in a cloud computing environment have become an important area of ​​network information technology at present and in the next few years. When each application system sends real-time events in an active push mode, it becomes a difficult technical problem to achieve low-latency processing and high-efficiency feedback under high throughput, so higher requirements are put forward: with the continuous increase in data volume and the rapid growth of data scale, the real-time processing system needs to be able to efficiently process and analyze massive data, which requires the system to have good scalability to adapt to the growth of data volume; the real-time large-scale event processing architecture needs to have high availability and be able to automatically handle node failures, network failures and other problems to ensure continuous processing of events.

[0003] Therefore, how to propose a real-time event processing method and system based on cloud computing in a computer software system to meet the requirements of large data flow, high concurrency and high real-time data analysis, maintain business continuity, realize dynamic adjustment of system processing nodes, and improve work performance and scope of application is an urgent problem that technical personnel in this field need to solve. Summary of the invention

[0004] In view of this, the present invention provides a real-time event processing method and system based on cloud computing in a computer software system, which meets the requirements of large data flow, high concurrency and high real-time data analysis, maintains business continuity, realizes dynamic adjustment of system processing nodes, and improves working performance and scope of application. In order to achieve the above purpose, the present invention adopts the following technical solutions:

[0005] A real-time event processing method based on cloud computing in a computer software system, comprising:

[0006] Distribute the event stream to different event processing nodes, and receive messages about the addition or exit of event processing nodes;

[0007] Determine multiple event processing nodes and interaction information between multiple event processing nodes according to the addition of event processing nodes or the exit message of event processing nodes, and construct an initial interaction graph;

[0008] Optimizing the initial interaction graph to obtain an optimized interaction graph, and updating the optimized interaction graph to obtain a deep interaction graph;

[0009] Determine a plurality of main event processing nodes and determine initial behaviors of the main event processing nodes according to target state information of the deep interaction graph, and execute the initial behaviors;

[0010] The secondary event processing nodes under the main event processing node share the initial behavior, and the secondary event processing nodes determine the second behavior of the secondary event processing nodes based on the initial behavior and the target state information, and execute the second behavior, organize the second behavior, and remove output events generated by duplicate events.

[0011] Optionally, the process further includes determining the secondary event processing node:

[0012] Any node in the secondary event processing node sends a request message containing verification information; determines whether the node sending the request message is in the trusted node list, if so, there is no consensus phase, and this node directly executes the second behavior, otherwise, the second behavior is executed after the consensus phase;

[0013] If there are n identical request messages within a specified time, a consensus is reached; for the trusted node list, the primary event processing node periodically performs message matching to inspect the trusted nodes in the secondary event processing node.

[0014] Optionally, the verification information in the request message containing verification information includes a signature of the main event processing node and a verification code.

[0015] Optionally, the determining of multiple main event processing nodes and determining the initial behavior of the main event processing nodes according to the target state information of the deep interaction graph further includes:

[0016] Set the initial parameters of the cluster; the initial parameters are used to indicate the number of subclusters into which the cluster is divided; for any subcluster, the subcluster is composed of a primary event processing node and a secondary event processing node; the primary event processing node and the secondary event processing node are individual members in the subcluster; the target state information is a fault state or a normal state.

[0017] Optionally, the cluster is divided into n first-level subclusters according to the initial parameters of the cluster; each first-level subcluster includes M nodes, and any one of the M nodes in each first-level subcluster is set as the first-level master node of the first-level subcluster to which it belongs; and n is a positive integer.

[0018] Optionally, in the process of forwarding the request message, each node forwarding the request message adds the signature of the secondary event processing node to the request message when forwarding.

[0019] Optionally, optimizing the initial interaction graph to obtain an optimized interaction graph, and updating the optimized interaction graph to obtain a deep interaction graph includes:

[0020] The initial interaction graph is optimized based on the optimization rule to obtain an optimized interaction graph, wherein the optimized interaction graph includes optimized interaction nodes and optimized connection relationships, wherein the optimization rule is to optimize the initial interaction nodes into the optimized interaction nodes based on the primary fitting relationship, and to optimize the initial connection relationship into the optimized connection relationship based on the secondary fitting relationship;

[0021] In response to a viewing operation on a primary interaction node in the optimized interaction graph, a secondary interaction node corresponding to the primary interaction node is determined in the initial interaction node based on the primary fitting relationship, and a bottom-level interaction node interacting with the secondary interaction node is determined in the initial interaction node based on the initial connection relationship, as well as an initial connection relationship between the secondary interaction node and the bottom-level interaction node;

[0022] A deep interaction map is generated based on the secondary interaction nodes, the bottom-level interaction nodes, and the initial connection relationship between the secondary interaction nodes and the bottom-level interaction nodes, and the displayed optimized interaction map is replaced with the deep interaction map.

[0023] Optionally, also include:

[0024] After receiving the new node message, the main event processing node responds to the new node's consensus message request;

[0025] After receiving the request, the secondary event processing node calculates the information of the newly added node after it is added, and the information after it is added includes at least the state information and interaction information between the nodes;

[0026] After the new node is added, a new master node is generated, and the update of the new master node is verified;

[0027] When the main event processing node periodically performs message matching, if it matches with the newly added node message, the newly added node is added to the trusted nodes.

[0028] Optionally, it also includes deleting nodes; the main event processing node initiates a deletion message, and when the deleted node receives the exit consent message sent by the remaining nodes, it exits automatically, and the remaining nodes update their own status information.

[0029] Optionally, a real-time event processing system based on cloud computing in a computer software system comprises:

[0030] Receiving module: used to distribute event streams to different event processing nodes and receive messages about the addition or exit of event processing nodes;

[0031] Construction module: used to determine multiple event processing nodes and interaction information between multiple event processing nodes according to the addition of event processing nodes or the exit message of event processing nodes, and to construct an initial interaction graph;

[0032] Optimization module: used for optimizing the initial interaction graph to obtain an optimized interaction graph, and updating the optimized interaction graph to obtain a deep interaction graph;

[0033] Pre-execution module: used to determine multiple main event processing nodes and determine the initial behavior of the main event processing nodes according to the target state information of the deep interaction graph, and execute the initial behavior;

[0034] Output module: used for the secondary event processing nodes under the main event processing node to share the initial behavior, the secondary event processing node determines the second behavior of the secondary event processing node according to the initial behavior and the target state information, executes the second behavior, organizes the second behavior, and removes output events generated by duplicate events.

[0035] It can be seen from the above technical solutions that, compared with the prior art, the present invention discloses a real-time event processing method and system based on cloud computing in a computer software system, which has the following beneficial effects:

[0036] The present invention proposes a real-time event processing method based on cloud computing in a computer software system, comprising: distributing event streams to different event processing nodes, and receiving messages of newly added event processing nodes or exit messages of event processing nodes; determining multiple event processing nodes and interaction information between multiple event processing nodes according to messages of newly added event processing nodes or exit messages of event processing nodes, and constructing an initial interaction graph; optimizing the initial interaction graph to obtain an optimized interaction graph, and updating the optimized interaction graph to obtain a deep interaction graph; determining multiple main event processing nodes and determining the initial behavior of the main event processing nodes according to target state information of the deep interaction graph, and executing the initial behavior; the secondary event processing nodes under the main event processing nodes share the initial behavior, the secondary event processing nodes determine the second behavior of the secondary event processing nodes according to the initial behavior and the target state information, and execute the second behavior, sorting the second behavior, and removing output events generated by repeated events. The present invention automatically assigns new events to appropriate event processing nodes. The secondary event processing node determines the secondary event processing node's second behavior according to the initial behavior and the target state information, executes the second behavior, sorts the second behavior, removes output events generated by duplicate events, and ensures the uniqueness and processing speed of output events. The scale of event processing nodes in the system is dynamically adjusted. When a certain event processing node fails, a normal node can be randomly selected from its corresponding event processing node group for processing to ensure the high availability of the entire event processing node group, meet the requirements of large data flow, high concurrency and high real-time data analysis, maintain business continuity, realize dynamic adjustment of system processing nodes, improve work performance and scope of application, and thus ensure the continuity of upper-level business. BRIEF DESCRIPTION OF THE DRAWINGS

[0037] In order to more clearly illustrate the embodiments of the present invention or the technical solutions in the prior art, the drawings required for use in the embodiments or the description of the prior art will be briefly introduced below. Obviously, the drawings described below are only embodiments of the present invention. For ordinary technicians in this field, other drawings can be obtained based on the provided drawings without paying creative work.

[0038] Figure 1 A flowchart of a real-time event processing method based on cloud computing in a computer software system provided by the present invention.

[0039] Figure 2 A structural framework diagram of a real-time event processing system based on cloud computing in a computer software system provided by the present invention. DETAILED DESCRIPTION

[0040] The following will be combined with the drawings in the embodiments of the present invention to clearly and completely describe the technical solutions in the embodiments of the present invention. Obviously, the described embodiments are only part of the embodiments of the present invention, not all of the embodiments. Based on the embodiments of the present invention, all other embodiments obtained by ordinary technicians in this field without creative work are within the scope of protection of the present invention.

[0041] The embodiment of the present invention discloses a real-time event processing method based on cloud computing in a computer software system, such as Figure 1 As shown, including:

[0042] Distribute the event stream to different event processing nodes, and receive messages about the addition or exit of event processing nodes;

[0043] Determine multiple event processing nodes and interaction information between multiple event processing nodes according to the addition of event processing nodes or the exit message of event processing nodes, and construct an initial interaction graph;

[0044] Optimizing the initial interaction graph to obtain an optimized interaction graph, and updating the optimized interaction graph to obtain a deep interaction graph;

[0045] Determine a plurality of main event processing nodes and determine initial behaviors of the main event processing nodes according to target state information of the deep interaction graph, and execute the initial behaviors;

[0046] The secondary event processing nodes under the main event processing node share the initial behavior, and the secondary event processing nodes determine the second behavior of the secondary event processing nodes based on the initial behavior and the target state information, and execute the second behavior, organize the second behavior, and remove output events generated by duplicate events.

[0047] Furthermore, it also includes judging the secondary event processing node:

[0048] Any node in the secondary event processing node sends a request message containing verification information; determines whether the node sending the request message is in the trusted node list, if so, there is no consensus phase, and this node directly executes the second behavior, otherwise, the second behavior is executed after the consensus phase;

[0049] If there are n identical request messages within a specified time, a consensus is reached; for the trusted node list, the primary event processing node periodically performs message matching to inspect the trusted nodes in the secondary event processing node.

[0050] Furthermore, the verification information in the request message containing verification information includes the signature of the main event processing node and the verification code, so that all messages are signed messages, thereby ensuring that the identity of the message sender and the message content cannot be forged or tampered with.

[0051] Furthermore, digital signatures and message verification codes are used in combination, where only view change messages and new view messages use signature messages, and other messages use message verification codes, thus saving a lot of encryption and decryption performance overhead.

[0052] Furthermore, the determining of multiple main event processing nodes and determining the initial behavior of the main event processing nodes according to the target state information of the deep interaction graph also includes:

[0053] Set the initial parameters of the cluster; the initial parameters are used to indicate the number of subclusters into which the cluster is divided; for a subcluster, the subcluster is composed of a primary event processing node and a secondary event processing node; the primary event processing node and the secondary event processing node are individual members in the subcluster; the target state information is a fault state or a normal state.

[0054] Further, the cluster is divided into n first-level subclusters according to the initial parameters of the cluster; each first-level subcluster includes M nodes, and any one of the M nodes in each first-level subcluster is set as the first-level master node of the first-level subcluster to which it belongs; and n is a positive integer.

[0055] In a specific implementation, the first-level master node is a primary event processing node, and M-1 nodes are secondary event processing nodes. When the target state information of the primary event processing node is in a normal state, the first-level master node is used to respond to the received instruction to process the first event stream, obtain the source information of the first event stream, and communicate with other nodes in the sub-cluster to which it belongs to share the first event stream source information and target state information; determine the first event stream interaction information between the current node and the first event stream according to the first event stream source information; communicate with the master nodes of other sub-clusters of the cluster to obtain the status information of the master nodes of other sub-clusters of the cluster; the status information of the master nodes of other sub-clusters of the cluster includes the interaction information and work efficiency vectors of the master nodes of other sub-clusters; determine the work efficiency vector of the current node according to the status information of the master nodes of other sub-clusters of the cluster; determine the first supplementary work efficiency vector of the current node according to the first event stream interaction information and the work efficiency vector of the current node; determine the first behavior of the current node according to the work efficiency vector of the current node and the first supplementary work efficiency vector.

[0056] In a specific implementation, when the target status information of the main event processing node is a fault status, the first-level master node is used to respond to the received instruction to process the second event stream, obtain the source information of the second event stream, and communicate with other nodes in the sub-cluster to which it belongs to share the second event stream source information and target status information; determine the second event stream interaction information between this node and the second event stream according to the second event stream source information; communicate with the master nodes of other sub-clusters of the cluster to obtain the status information of the master nodes of other sub-clusters of the cluster; the status information of the master nodes of other sub-clusters of the cluster includes the interaction information and work efficiency vectors of the master nodes of other sub-clusters; determine the work efficiency vector of this node according to the status information of the master nodes of other sub-clusters of the cluster; determine the second supplementary work efficiency vector of this node according to the second event stream interaction information and the work efficiency vector of this node; determine the first behavior of this node according to the work efficiency vector and the second supplementary work efficiency vector of this node.

[0057] Furthermore, in the process of forwarding the request message, each node forwarding the request message adds the signature of the secondary event processing node to the request message when forwarding.

[0058] Further, optimizing the initial interaction graph to obtain an optimized interaction graph, and updating the optimized interaction graph to obtain a deep interaction graph includes:

[0059] The initial interaction graph is optimized based on the optimization rule to obtain an optimized interaction graph, wherein the optimized interaction graph includes optimized interaction nodes and optimized connection relationships, wherein the optimization rule is to optimize the initial interaction nodes into the optimized interaction nodes based on the primary fitting relationship, and to optimize the initial connection relationship into the optimized connection relationship based on the secondary fitting relationship;

[0060] In response to a viewing operation on a primary interaction node in the optimized interaction graph, a secondary interaction node corresponding to the primary interaction node is determined in the initial interaction node based on the primary fitting relationship, and a bottom-level interaction node interacting with the secondary interaction node is determined in the initial interaction node based on the initial connection relationship, as well as an initial connection relationship between the secondary interaction node and the bottom-level interaction node;

[0061] A deep interaction map is generated based on the secondary interaction nodes, the bottom-level interaction nodes, and the initial connection relationship between the secondary interaction nodes and the bottom-level interaction nodes, and the displayed optimized interaction map is replaced with the deep interaction map.

[0062] In a specific implementation, determining a plurality of event flow information in the overall cloud computing process also includes:

[0063] Determine multiple subclusters with interactive behaviors in the cloud computing process and the interaction information between the multiple subclusters; the subclusters include a receiving end and a processing end; generate an initial interaction graph based on the multiple subclusters and the interaction information, the initial interaction graph includes initial interaction nodes of each processing end and initial interaction nodes of the receiving end, each initial interaction node of the receiving end corresponds to one or more receiving ends, the initial interaction nodes of the processing end and the initial interaction nodes of the processing end are connected by indicator lines, the indicator lines are used to indicate the interaction between the ports of the two initial interaction nodes connected to them, the interaction graph is used to visualize the initial connection relationship between multiple initial interaction nodes, the initial interaction nodes are used to indicate the subclusters, and the initial connection relationship is used to indicate the interaction information.

[0064] Specifically, the receiving end is divided and summarized according to the first division rule to obtain multiple main divisions, and the second division is summarized in each main division to obtain a secondary division; each main division corresponds to a main division visualization set, and each secondary division corresponds to a secondary division identifier. Each main division visualization set is used to visualize the secondary division identifier belonging to the main division, and the receiving end corresponding to the secondary division identifier is folded and displayed in the secondary division identifier; the correspondence between different port information and the main division and the secondary division is predetermined; for the newly added port information, feature extraction is performed according to the port information, and then the Euclidean distance calculation is performed according to the extracted features to obtain the distance between each port, determine the center corresponding to each secondary division, and calculate the corresponding relationship according to the new port information. The added port that is closest to the center of the secondary partition is taken as the secondary partition to which it belongs, and the primary partition of the newly added port is determined according to the subordinate relationship between the primary partition and the secondary partition; wherein a first distance threshold is set in advance, and when the distances between the newly added port and the center of the secondary partition are all greater than the first distance threshold, a new secondary partition can be determined based on the newly added port; the center of the primary partition is determined according to all port information corresponding to the primary partition, and the primary partition to which the new secondary partition belongs is determined according to the distance between the center of the primary partition and the center of the new secondary partition; when it is determined that a new primary partition has been added, the sets of each main partition are reallocated according to the displayable set of the main partition in the interface and the predetermined allocation rules.

[0065] Furthermore, it also includes:

[0066] After receiving the new node message, the main event processing node responds to the new node's consensus message request;

[0067] After receiving the request, the secondary event processing node calculates the information of the newly added node after it is added, and the information after it is added includes at least the state information and interaction information between the nodes;

[0068] After the new node is added, a new master node is generated, and the new master node is updated and verified;

[0069] When the main event processing node periodically performs message matching, if it matches with the newly added node message, the newly added node is added to the trusted nodes.

[0070] In a specific implementation manner, any node sends a request message containing verification information; determines whether the node sending the request message is in the trusted node list, if so, there is no consensus stage, and this node directly executes the request message, otherwise the request message is executed after the consensus stage; if there are n+1 identical request messages within a specified time, consensus is reached; for the trusted node list, the master node periodically performs message matching to inspect the trusted nodes in the candidate nodes, and the message matching content is <request, k, o, g>, wherein g is the master node signature mechanism, t is the time interval for each master node to perform the consensus stage, and k is the execution operation.

[0071] Furthermore, it also includes deleting a node; the main event processing node initiates a deletion message, and when the deleted node receives a consent exit message from the remaining nodes, it exits automatically, and the remaining nodes update their own status information automatically.

[0072] In a specific implementation, a real-time event processing system based on cloud computing in a computer software system, such as Figure 2 As shown, including:

[0073] Receiving module: used to distribute event streams to different event processing nodes and receive messages about the addition or exit of event processing nodes;

[0074] Construction module: used to determine multiple event processing nodes and interaction information between multiple event processing nodes according to the addition of event processing nodes or the exit message of event processing nodes, and to construct an initial interaction graph;

[0075] Optimization module: used for optimizing the initial interaction graph to obtain an optimized interaction graph, and updating the optimized interaction graph to obtain a deep interaction graph;

[0076] Pre-execution module: used to determine multiple main event processing nodes and determine the initial behavior of the main event processing nodes according to the target state information of the deep interaction graph, and execute the initial behavior;

[0077] Output module: used for the secondary event processing nodes under the main event processing node to share the initial behavior, the secondary event processing node determines the second behavior of the secondary event processing node according to the initial behavior and the target state information, executes the second behavior, organizes the second behavior, and removes output events generated by duplicate events.

[0078] In a specific implementation, an embodiment of the present invention further provides an electronic device, including a processor, a communication interface, a memory and a communication bus, wherein the processor, the communication interface and the memory communicate with each other through the communication bus. The memory is used to store computer programs; the processor is used to execute the programs stored in the memory. The communication bus can be a Peripheral Component Interconnect (PCI) bus or an Extended Industry Standard Architecture (EISA) bus, etc. The communication bus can be divided into an address bus, a data bus, a control bus, etc. The communication interface is used for communication between the above-mentioned electronic device and other devices. The memory can include a random access memory (RAM) or a non-volatile memory (NVM), such as at least one disk storage. Optionally, the memory can also be at least one storage device located away from the aforementioned processor. The above-mentioned processor can be a general-purpose processor, including a central processing unit (CPU), a network processor (NP), etc.; it can also be a digital signal processor (DSP), an application specific integrated circuit (ASIC), a field programmable gate array (FPGA) or other programmable logic devices, discrete gate or transistor logic devices, discrete hardware components.

[0079] In another embodiment provided by the present invention, a computer-readable storage medium is also provided, in which instructions are stored. When the computer-readable storage medium is executed on a computer, the computer executes a real-time event processing method based on cloud computing in a computer software system in any one of the above embodiments.

[0080] In another embodiment provided by the present invention, a computer program product including instructions is also provided, which, when executed on a computer, enables the computer to execute a real-time event processing method based on cloud computing in a computer software system in any one of the above embodiments.

[0081] In a specific implementation, a real-time event processing structure based on cloud computing in a computer software system is also included, specifically including:

[0082] Event distribution node: distributes high-speed event streams to different event processing nodes and receives notification messages from the node manager; the notification message is the addition of an event processing node or the exit of an event processing node.

[0083] Event processing node: contains the existing general event processing engine, receives events from the event distribution node, and passes the processed new events to the event deduplication node. The event processing process is completed by the general event processing engine.

[0084] Event processing node group: contains at least two event processing nodes, and each event processing node in the group completes the same complex event processing business logic.

[0085] Event deduplication node: Sorts new events and removes output events generated by duplicate events.

[0086] Node Manager: Controls the event processing nodes and notifies the event distribution nodes to handle the corresponding notification messages.

[0087] The present invention automatically distributes new events to appropriate event processing nodes through event distribution nodes, and ensures the uniqueness of output events through event deduplication nodes. The node manager can dynamically adjust the scale of event processing nodes in the system. When a certain event processing node fails, a normal node can be randomly selected from its corresponding event processing node group for processing to ensure the high availability of the entire event processing node group, thereby ensuring the continuity of upper-level services.

[0088] In this specification, each embodiment is described in a progressive manner, and each embodiment focuses on the differences from other embodiments. The same or similar parts between the embodiments can be referred to each other. For the device disclosed in the embodiment, since it corresponds to the method disclosed in the embodiment, the description is relatively simple, and the relevant parts can be referred to the method part.

[0089] The above description of the disclosed embodiments enables one skilled in the art to implement or use the present invention. Various modifications to these embodiments will be apparent to one skilled in the art, and the general principles defined herein may be implemented in other embodiments without departing from the spirit or scope of the present invention. Therefore, the present invention will not be limited to the embodiments shown herein, but rather to the widest scope consistent with the principles and novel features disclosed herein.

Claims

1. A real-time event processing method based on cloud computing in a computer software system, characterized in that: include: Distribute the event stream to different event processing nodes, and receive messages about the addition or exit of event processing nodes; Determine multiple event processing nodes and interaction information between multiple event processing nodes according to the addition of event processing nodes or the exit message of event processing nodes, and construct an initial interaction graph; Optimizing the initial interaction graph to obtain an optimized interaction graph, and updating the optimized interaction graph to obtain a deep interaction graph; Determine a plurality of main event processing nodes and determine initial behaviors of the main event processing nodes according to target state information of the deep interaction graph, and execute the initial behaviors; The secondary event processing nodes under the main event processing node share the initial behavior, and the secondary event processing nodes determine the second behavior of the secondary event processing nodes based on the initial behavior and the target state information, and execute the second behavior, organize the second behavior, and remove output events generated by duplicate events.

2. The real-time event processing method based on cloud computing in a computer software system according to claim 1, characterized in that: It also includes judging the secondary event processing node: Any node in the secondary event processing node sends a request message containing verification information; determines whether the node sending the request message is in the trusted node list, if so, there is no consensus phase, and this node directly executes the second behavior, otherwise, the second behavior is executed after the consensus phase; If there are n identical request messages within a specified time, a consensus is reached; for the trusted node list, the primary event processing node periodically performs message matching to inspect the trusted nodes in the secondary event processing node.

3. The real-time event processing method based on cloud computing in a computer software system according to claim 2, characterized in that: The verification information in the request message containing verification information includes the signature of the main event processing node and a verification code.

4. The real-time event processing method based on cloud computing in a computer software system according to claim 1, characterized in that: The determining of a plurality of main event processing nodes and determining the initial behavior of the main event processing nodes according to the target state information of the deep interaction graph further includes: Set the initial parameters of the cluster; the initial parameters are used to indicate the number of subclusters into which the cluster is divided; for any subcluster, the subcluster is composed of a primary event processing node and a secondary event processing node; the primary event processing node and the secondary event processing node are individual members in the subcluster; the target state information is a fault state or a normal state.

5. The real-time event processing method based on cloud computing in a computer software system according to claim 4, characterized in that: The cluster is divided into n first-level subclusters according to the initial parameters of the cluster; each first-level subcluster includes M nodes, and any one of the M nodes in each first-level subcluster is set as the first-level master node of the first-level subcluster to which it belongs; n is a positive integer.

6. The real-time event processing method based on cloud computing in a computer software system according to claim 3, characterized in that: In the process of forwarding the request message, each node forwarding the request message adds the signature of the secondary event processing node in the request message when forwarding.

7. The real-time event processing method based on cloud computing in a computer software system according to claim 1, characterized in that: Optimizing the initial interaction graph to obtain an optimized interaction graph, and updating the optimized interaction graph to obtain a deep interaction graph includes: The initial interaction graph is optimized based on the optimization rule to obtain an optimized interaction graph, wherein the optimized interaction graph includes optimized interaction nodes and optimized connection relationships, wherein the optimization rule is to optimize the initial interaction nodes into the optimized interaction nodes based on the primary fitting relationship, and to optimize the initial connection relationship into the optimized connection relationship based on the secondary fitting relationship; In response to a viewing operation on a primary interaction node in the optimized interaction graph, a secondary interaction node corresponding to the primary interaction node is determined in the initial interaction node based on the primary fitting relationship, and a bottom-level interaction node interacting with the secondary interaction node is determined in the initial interaction node based on the initial connection relationship, as well as an initial connection relationship between the secondary interaction node and the bottom-level interaction node; A deep interaction map is generated based on the secondary interaction nodes, the bottom-level interaction nodes, and the initial connection relationship between the secondary interaction nodes and the bottom-level interaction nodes, and the displayed optimized interaction map is replaced with the deep interaction map.

8. The real-time event processing method based on cloud computing in a computer software system according to claim 2, characterized in that: Also includes: After receiving the new node message, the main event processing node responds to the new node's consensus message request; After receiving the request, the secondary event processing node calculates the information of the newly added node after it is added, and the information after it is added includes at least the state information and interaction information between the nodes; After the new node is added, a new master node is generated, and the update of the new master node is verified; When the main event processing node periodically performs message matching, if it matches with the newly added node message, the newly added node is added to the trusted nodes.

9. The real-time event processing method based on cloud computing in a computer software system according to claim 1, characterized in that: It also includes deleting nodes; the main event processing node initiates a deletion message, and when the deleted node receives the consent exit message sent by the remaining nodes, it exits automatically, and the remaining nodes update their own status information.

10. A real-time event processing system based on cloud computing in a computer software system, characterized in that: include: Receiving module: used to distribute event streams to different event processing nodes and receive messages about the addition or exit of event processing nodes; Construction module: used to determine multiple event processing nodes and interaction information between multiple event processing nodes according to the addition of event processing nodes or the exit message of event processing nodes, and to construct an initial interaction graph; Optimization module: used for optimizing the initial interaction graph to obtain an optimized interaction graph, and updating the optimized interaction graph to obtain a deep interaction graph; Pre-execution module: used to determine multiple main event processing nodes and determine the initial behavior of the main event processing nodes according to the target state information of the deep interaction graph, and execute the initial behavior; Output module: used for the secondary event processing nodes under the main event processing node to share the initial behavior, the secondary event processing node determines the second behavior of the secondary event processing node according to the initial behavior and the target state information, executes the second behavior, organizes the second behavior, and removes output events generated by duplicate events.

Images