State data acquisition method for multi-split cross-platform distributed monitoring system

Abstract

The invention discloses a multi-split cross-platform distributed monitoring system state data acquisition method, which belongs to the field of system operation and maintenance and data monitoring, and comprises the following steps: acquiring node data through a data acquisition module arranged on a distributed monitoring system node, and introducing a self-adaptive acquisition strategy into the data acquisition module, adaptively adjusting the node data acquisition frequency according to the resource utilization condition; a node data monitoring strategy is formulated according to user requirements, and the monitoring strategy is optimized by using a hybrid monitoring algorithm; a Delta compression technology is utilized to compress node data to obtain a compressed data packet, the compressed data packet is transmitted, and a multi-level transmission optimization mechanism is introduced in the transmission process of the compressed data packet. According to the method, the technologies such as message bus publishing and subscribing, a customized data acquisition strategy and data flow optimization are integrated, so that consistent state monitoring can be realized among different platforms and hardware architectures, and the efficiency and reliability of system operation and maintenance are greatly improved.

Description

Technical Field

[0001] The present invention relates to the field of system operation and maintenance and data monitoring, and in particular to a method for acquiring status data of a multi-online cross-platform distributed monitoring system. Background Art

[0002] Distributed monitoring systems often need to perform monitoring tasks under diverse operating systems, hardware architectures, and network conditions. Power system monitoring, financial data center management, and large-scale industrial equipment operation and maintenance applications all rely on intelligent O&M solutions for distributed monitoring systems. With the dynamic changes in network topology and the frequent addition and removal of nodes, existing monitoring methods face multiple challenges in collecting node status data, rapidly responding to failures, and optimizing data transmission. Furthermore, nodes can be highly heterogeneous, such as running different operating systems (e.g., Linux and Windows) or hardware architectures (e.g., x86 and ARM), further increasing the complexity of status data collection and monitoring.

[0003] Distributed monitoring systems play an important role in the operation and maintenance of power system monitoring. However, the diverse equipment and numerous network nodes make it difficult to obtain batch data on system status. Different cluster cross-combinations, differences in operating systems and device architectures, and dispersed operator station locations all increase the complexity and cost of operation and maintenance, making it difficult to detect and resolve system failures in a timely manner.

[0004] For example, the patent document with application number CN201310129236.0 discloses a policy-based unified monitoring and operation and maintenance method and device. This method realizes the centralized management of distributed monitoring systems through policy-driven unified monitoring and operation and maintenance. However, this method lacks the ability to adjust the acquisition frequency and content according to the real-time load and dynamic changes of the system, which will increase the system pressure under high load conditions and affect the responsiveness of monitoring. In addition, the system mainly relies on manually set thresholds to detect faults, lacks automated anomaly analysis and trend prediction, and cannot detect potential faults and performance problems in a timely manner.

[0005] The patent document with application number CN202110308663.X discloses a general counter system operation and maintenance method based on SHELL scripts. The method obtains the system status based on SHELL scripts and calls the built-in basic commands of the Linux system to check the equipment load. It lacks fine-grained status data acquisition: the system only supports batch operations and simple status checks, and lacks comprehensive collection functions for detailed hardware usage such as CPU, GPU, memory, and network, and cannot achieve refined monitoring. Since it is based on batch script execution, the system has a certain delay when monitoring or responding to faults, and the data transmission efficiency is also low, and thus cannot perceive and process dynamic changes in large-scale distributed monitoring systems in real time.

[0006] The patent document with application number CN202411030165.3 discloses an operating system online fault detection method, device and storage medium. The method obtains the operating status information of the terminal operating system online through dynamic capture. The dynamic capture method tracks and monitors the operating status information of the terminal operating system through eBPF (Extended Berkeley Packet Filter) technology. Different execution modules are used for dynamic monitoring corresponding to different operating status information. However, the use of eBPF depends on the operating system's support for this technology, and its monitoring method is relatively cohesive, which is usually more suitable for in-depth monitoring of a single operating system.

[0007] Currently, no effective solutions have been proposed for the problems in related technologies. Summary of the Invention

[0008] In response to the problems in the related art, the present invention proposes a method for acquiring status data of a multi-online cross-platform distributed monitoring system to overcome the above technical problems existing in the existing related art.

[0009] To this end, the specific technical solutions adopted in the present invention are as follows:

[0010] A method for acquiring status data of a multi-online cross-platform distributed monitoring system, the method comprising:

[0011] S1. Node data is collected through a data acquisition module deployed on the nodes of the distributed monitoring system, and an adaptive acquisition strategy is introduced in the data acquisition module to adaptively adjust the node data acquisition frequency according to resource utilization;

[0012] S2. Develop node data monitoring strategies based on user needs and optimize them using hybrid monitoring algorithms based on potential anomalies in the distributed monitoring system.

[0013] S3. Compress the node data using Delta compression technology to obtain a compressed data packet, and transmit the compressed data packet. In the process of transmitting the compressed data packet, a multi-level transmission optimization mechanism is introduced to optimize the data transmission rate.

[0014] S4. Introduce the publish-subscribe model in the distributed monitoring system. Any node subscribes to other node data on demand, and the node data is stored in a predefined time series database.

[0015] Preferably, collecting node data by a data acquisition module arranged on a node of a distributed monitoring system, and introducing an adaptive acquisition strategy in the data acquisition module to adaptively adjust the node data acquisition frequency according to resource utilization include:

[0016] S11, deploying the monitoring tool and executable files on all nodes of the distributed monitoring system, and reading the configuration file of the distributed monitoring system to start the data acquisition module of each node;

[0017] S12. Utilize the data acquisition module to collect node data, match the node data to determine whether it can participate in data interaction with the distributed monitoring system, and perform real-time display of the online status and configuration of each node based on the data interaction results;

[0018] S13. An adaptive collection strategy is introduced into the data acquisition module, and the distributed monitoring system adjusts the node data collection frequency according to the current resource utilization.

[0019] Preferably, an adaptive acquisition strategy is introduced into the data acquisition module, and the distributed monitoring system adjusts the node data acquisition frequency according to the current resource utilization, including:

[0020] S131. Setting a maximum load threshold of the distributed monitoring system, and calculating a sampling adjustment coefficient based on a ratio function of a current load of the distributed monitoring system to the maximum load threshold;

[0021] S132. If the current load of the distributed monitoring system is less than a preset maximum load threshold, the acquisition frequency is maintained at the base frequency. If the current load of the distributed monitoring system increases to the preset maximum load threshold, the sampling adjustment coefficient is set to the maximum value to reduce the acquisition frequency to a safe lower limit under high load.

[0022] S133. Update the collection frequency of the distributed monitoring system based on the sampling adjustment coefficient.

[0023] Preferably, the calculation formula of the sampling adjustment coefficient is:

[0024]

[0025] In the formula, α represents the sampling adjustment coefficient; max_load represents the preset maximum load threshold; system_load represents the current load of the distributed monitoring system; α min represents the minimum sampling adjustment coefficient; k represents the adjustment coefficient scaling factor.

[0026] Preferably, the node data includes static information and customized information;

[0027] Static information includes the node’s hardware model, firmware version, and operating system;

[0028] Customized information includes hardware usage of CPU, GPU, memory, disk, and network, time series database status, solid-state drive life, system cooling performance, hardware fault logs, and network latency.

[0029] Preferably, a node data monitoring strategy is formulated according to user needs. Based on the potential abnormalities of the distributed monitoring system, the monitoring strategy is optimized using a hybrid monitoring algorithm, including:

[0030] S21. Extract historical monitoring data from a time series database and construct a training set based on key indicators in the historical monitoring data; train a prediction model using the training set and a supervised learning algorithm to obtain a prediction model for predicting short-term trends;

[0031] S22. Setting initial threshold rules and association rules for key indicators and using them as judgment conditions;

[0032] S23, extracting the latest status data obtained in each monitoring cycle during the operation of the distributed monitoring system, inputting the latest status data into the prediction model for prediction, and obtaining the observed value of the performance indicator at the next moment;

[0033] S24. Determine whether the observed value meets the judgment conditions. If the observed value does not meet the judgment conditions, it is determined that a potential abnormality has occurred in the distributed monitoring system, and the collection frequency of key indicators is increased, while the collection frequency of other non-key indicators is reduced. Conversely, if the observed value meets the judgment conditions, the preset collection frequency is gradually restored.

[0034] Preferably, the node data is compressed using Delta compression technology to obtain a compressed data packet, and the compressed data packet is transmitted. A multi-level transmission optimization mechanism is introduced during the transmission of the compressed data packet to optimize the data transmission rate, including:

[0035] S31, extracting the node data collected by the distributed monitoring system in the previous monitoring cycle and the node data collected currently and calculating the data difference. If the data difference is empty, the current data transmission is interrupted; otherwise, the next step is executed;

[0036] S32. Based on the data difference calculation result, the changed data is compressed using the Delta compression technology to obtain a compressed data packet, and the compressed data packet is published and transmitted through the message bus module;

[0037] S33. During the transmission of compressed data packets, a distributed monitoring system is introduced into a multi-level transmission optimization mechanism to perceive the transmission status of compressed data packets in real time;

[0038] S34. After receiving the compressed data packet, the receiving end parses it and updates the local status data.

[0039] Preferably, the multi-level transmission optimization mechanism includes:

[0040] Use data deduplication strategy to mark and filter duplicate data in pre-transmitted data in sequence;

[0041] Calculate the adjustment coefficient based on the current load of the distributed monitoring system and the preset maximum load threshold, and adjust the acquisition frequency according to steps S131 to S133;

[0042] A dynamic bandwidth adjustment mechanism is introduced into the distributed monitoring system. The dynamic bandwidth adjustment mechanism can sense the current network status in real time and adaptively adjust the data transmission rate according to the available bandwidth.

[0043] Preferably, the dynamic bandwidth adjustment mechanism is used to perceive the current network status in real time and to adaptively adjust the data transmission rate according to the available bandwidth, including:

[0044] Detect the transmission network status. If the bandwidth is lower than the preset threshold, the transmission rate parameter will be reduced. Otherwise, the current transmission rate will be maintained or increased.

[0045] The calculation formula of the transmission rate parameter is:

[0046] transmission_rate=g(network_status);

[0047] Wherein, transmission_rate represents the transmission rate parameter; g represents the bandwidth adjustment function; and network_status represents the bandwidth adjustment function.

[0048] Preferably, a publish-subscribe model is introduced in the distributed monitoring system, where any node subscribes to other node data on demand, and the node data is stored in a predefined time series database, including:

[0049] S41. Introducing a publish-subscribe model in a distributed monitoring system. Based on the publish-subscribe model, any node can subscribe to data from other nodes on demand to achieve monitoring of global nodes or specific nodes.

[0050] S42. Storing the node data in a time series database, predicting the future load of the node data based on the historical data pattern in the time series database and combining time series analysis technology, and predicting the capacity requirements of the distributed monitoring system based on the future load of the node data.

[0051] The beneficial effects of the present invention are:

[0052] 1. The present invention integrates advanced technologies such as message bus publishing and subscription, customized data collection strategies, and data flow optimization, thereby enabling consistent status monitoring across different platforms and hardware architectures, significantly improving the efficiency, response speed, and reliability of system operation and maintenance, and solving problems in the prior art such as incomplete monitoring data, low system flexibility, excessive resource utilization, low data transmission efficiency, and poor platform compatibility.

[0053] 2. The present invention comprehensively collects hardware usage including CPU, GPU, memory, disk, network, as well as customized information such as database status, solid-state drive life, system heat dissipation performance, hardware fault log, network delay, etc. Users can dynamically adjust data collection strategies through the interface to avoid excessive collection of invalid data, thereby reducing system load. By introducing an adaptive collection strategy, the system can flexibly adjust the collection frequency according to current resource utilization and reduce the collection density under high load to reduce system pressure.

[0054] 3. The present invention establishes an efficient message bus system to support the publication and subscription of status data between nodes. Based on the publish / subscribe model, any node can subscribe to the data of other nodes on demand, thereby realizing global or specific node monitoring. In addition, through standardized data acquisition and publishing interfaces, data is uniformly collected and processed, enhancing the adaptability and scalability of the system. The present invention introduces Delta compression technology to compress dynamic data in real time, and only transmits the changed parts. It is suitable for most distributed monitoring systems where data changes frequently but the amount of change is relatively small, thereby significantly reducing the amount of transmitted data and reducing network load. BRIEF DESCRIPTION OF THE DRAWINGS

[0055] In order to more clearly illustrate the embodiments of the present invention or the technical solutions in the prior art, the following briefly introduces the drawings required for use in the embodiments. Obviously, the drawings described below are only some embodiments of the present invention. For ordinary technicians in this field, other drawings can be obtained based on these drawings without paying any creative work.

[0056] Figure 1 This is a flow chart of a method for acquiring status data of a multi-online cross-platform distributed monitoring system according to an embodiment of the present invention;

[0057] Figure 2 The present invention is a flowchart of data flow optimization and compression transmission in a method for acquiring status data of a multi-connected cross-platform distributed monitoring system according to an embodiment of the present invention. DETAILED DESCRIPTION

[0058] To further illustrate each embodiment, the present invention provides drawings, which are part of the disclosure of the present invention. They are mainly used to illustrate the embodiments and can be used in conjunction with the relevant descriptions in the specification to explain the operating principles of the embodiments. By referring to these contents, ordinary technicians in this field should be able to understand other possible implementation methods and the advantages of the present invention.

[0059] According to an embodiment of the present invention, a method for acquiring status data of a multi-online cross-platform distributed monitoring system is provided.

[0060] The present invention will now be further described with reference to the accompanying drawings and specific embodiments. Figure 1 As shown, according to an embodiment of the present invention, a method for acquiring status data of a multi-online cross-platform distributed monitoring system includes:

[0061] S1. Node data is collected through a data acquisition module arranged on the nodes of the distributed monitoring system, and an adaptive acquisition strategy is introduced in the data acquisition module to adaptively adjust the node data acquisition frequency according to resource utilization.

[0062] The node data is collected by a data acquisition module arranged on the nodes of the distributed monitoring system, and an adaptive acquisition strategy is introduced in the data acquisition module to adaptively adjust the node data acquisition frequency according to resource utilization, including:

[0063] S11, deploying the monitoring tool and executable files on all nodes of the distributed monitoring system, and reading the configuration file of the distributed monitoring system to start the data acquisition module of each node;

[0064] S12. Utilize the data acquisition module to collect node data, match the node data to see whether it can participate in the distributed monitoring system data interaction, and perform real-time display of the online status and configuration of each node based on the data interaction results.

[0065] It should be noted that node data includes static information and customized information;

[0066] Static information includes the node’s hardware model, firmware version, and operating system;

[0067] Customized information includes hardware usage of CPU, GPU, memory, disk, and network, time series database status, solid-state drive life, system cooling performance, hardware fault logs, and network latency.

[0068] S13. An adaptive collection strategy is introduced into the data acquisition module, and the distributed monitoring system adjusts the node data collection frequency according to the current resource utilization.

[0069] Among them, the adaptive collection strategy is introduced in the data acquisition module. The distributed monitoring system adjusts the node data collection frequency according to the current resource utilization, including:

[0070] S131. Setting a maximum load threshold of the distributed monitoring system, and calculating a sampling adjustment coefficient based on a ratio function of a current load of the distributed monitoring system to the maximum load threshold;

[0071] S132. If the current load of the distributed monitoring system is less than a preset maximum load threshold, the acquisition frequency is maintained at the base frequency. If the current load of the distributed monitoring system increases to the preset maximum load threshold, the sampling adjustment coefficient is set to the maximum value to reduce the acquisition frequency to a safe lower limit under high load.

[0072] S133. Update the collection frequency of the distributed monitoring system based on the sampling adjustment coefficient.

[0073] It should be noted that comprehensive and customized data collection and deployment methods:

[0074] Deploy syswatch and syswatch.exe on all nodes in the distributed monitoring system. Each node's data acquisition module is activated by reading the distributed monitoring system configuration file (DSC.ini). This automatically collects static information such as the node's hardware model, firmware version, and operating system, and determines whether it can participate in distributed monitoring system data interaction. Through the message bus middleware, each node can directly access this information, displaying its online status and configuration in real time. This process effectively reduces the workload of manual configuration and significantly improves the system's real-time response performance.

[0075] The node data acquisition module comprehensively collects hardware usage including CPU, GPU, memory, disk, network, as well as customized information such as database status, solid-state drive life, system heat dissipation performance, hardware fault logs, network latency, etc. Users can dynamically adjust data collection strategies through the interface to avoid excessive collection of invalid data, thereby reducing system load. The modular data collection agent combined with the policy engine ensures that the frequency and content of data collection can be dynamically adjusted, thereby improving the relevance and effectiveness of the data. By introducing an adaptive collection strategy, the system can flexibly adjust the collection frequency according to the current resource utilization, reducing the collection density under high load to reduce system pressure, and increasing the collection frequency under low load to provide a more detailed data view.

[0076] Node data is transmitted in JSON format, and the instance structure is roughly as follows:

[0077]

[0078] Adaptive collection strategy:

[0079] By setting the maximum load threshold max_load allowed by the system, a function of the ratio of the current load to the threshold is defined to calculate the sampling adjustment coefficient α. The calculation formula of the sampling adjustment coefficient is:

[0080]

[0081] Where: max_load is the preset maximum load threshold; system_load is the current system load; α min is the minimum allowable adjustment coefficient (for example, 0.5, ensuring that the sampling frequency is not lower than the safety lower limit under high load); k is the adjustment coefficient scaling factor (0 < k ≤ 1), which controls the sensitivity of α to load changes. When system_load is much lower than max_load, α approaches 1 (the acquisition frequency approaches the base frequency base_rate); when system_load approaches or exceeds max_load, α drops to α min (the acquisition frequency decreases), and the acquisition frequency is adjusted accordingly: f new = α * base_rate, so as to achieve the dynamic adjustment of the sampling frequency according to the load, while minimizing the system overhead under high load while ensuring the necessary monitoring accuracy.

[0082] It should be noted that in the current single load ratio linear mapping sampling adjustment calculation formula, a multi-dimensional resource weight fusion-driven adaptive sampling mechanism is introduced, which specifically includes:

[0083] Monitoring nodes are usually restricted by multiple bottlenecks such as CPU, memory, disk I / O, and network bandwidth at the same time. Therefore, the single-index load is extended to a comprehensive load index:

[0084]

[0085] In the formula, L(t) represents the comprehensive load index; w i represents the weight of each resource dimension, which can be obtained through expert static configuration; β i represents the non-linear amplification coefficient, which is used to emphasize the sensitivity of the high load interval (load i →cap i ), and remains stable in the low load interval; cpu, men, disk, and net respectively represent the central processing unit, memory, disk, and network bandwidth.

[0086] Based on the comprehensive load index, the sampling coefficient is improved, and the comprehensive improved sampling coefficient formula is:

[0087] α(t) = α min +(α max [[ID=3s6]]-α min )·f(L(t));

[0088] In the formula, α(t) represents the sampling adjustment coefficient at time t; α min represents the minimum adjustment coefficient; α maxrepresents the maximum adjustment coefficient; f(·) represents the optional function, including Sigmoid function and power function. Among them, the Sigmoid function has the largest gradient in the medium and high load range (around 0.5), achieving smooth and sensitive frequency transition and avoiding oscillation; in the power function L γ In the process, the high-load range is exponentially amplified by γ>1, and the frequency reduction is accelerated to ensure system stability.

[0089] Compared with the existing single-slope linear function, this nonlinear mapping mechanism maintains sampling integrity in the low-medium load stage and quickly converges to the safety lower limit α in the high load stage. min , significantly reducing resource usage and being more friendly to overload.

[0090] S2. Develop node data monitoring strategies based on user needs, and optimize the monitoring strategies using hybrid monitoring algorithms based on potential abnormalities in the distributed monitoring system.

[0091] Among them, the node data monitoring strategy is formulated according to user needs, and the monitoring strategy is optimized using a hybrid monitoring algorithm based on potential abnormal situations in the distributed monitoring system, including:

[0092] S21. Extract historical monitoring data from a time series database and construct a training set based on key indicators in the historical monitoring data; train a prediction model using the training set and a supervised learning algorithm to obtain a prediction model for predicting short-term trends;

[0093] S22. Setting initial threshold rules and association rules for key indicators and using them as judgment conditions;

[0094] S23, extracting the latest status data obtained in each monitoring cycle during the operation of the distributed monitoring system, inputting the latest status data into the prediction model for prediction, and obtaining the observed value of the performance indicator at the next moment;

[0095] S24. Determine whether the observed value meets the judgment conditions. If the observed value does not meet the judgment conditions, it is determined that a potential abnormality has occurred in the distributed monitoring system, and the collection frequency of key indicators is increased, while the collection frequency of other non-key indicators is reduced. Conversely, if the observed value meets the judgment conditions, the preset collection frequency is gradually restored.

[0096] It should be noted that the cross-platform compatibility interface and flexible monitoring strategy configuration:

[0097] The state data acquisition method is compatible with multiple platforms and architectures (such as x86 and ARM) and supports multiple operating systems (such as Linux and Windows). Through standardized data acquisition and publishing interfaces, it enables unified data collection and processing, enhancing the system's adaptability and scalability. The implementation of standardized interfaces includes the use of a common data description format (JSON) and a unified communication protocol, ensuring consistent and efficient data transmission across different platforms. The modules provide standardized interfaces and dynamic library interfaces, enabling developers to develop new functional modules based on these interfaces. These modules can be loaded and unloaded at any time based on business needs without modifying the overall system architecture, significantly reducing maintenance costs. Furthermore, the system supports the dynamic addition or replacement of data acquisition segments during runtime, ensuring high flexibility and maintainability. Cross-platform compatibility and the integration of standardized interfaces enable the system to operate efficiently across diverse hardware architectures, operating systems, and deployment environments. Combined with a dynamic module management mechanism, the system achieves high levels of compatibility, scalability, and deployment flexibility, meeting the diverse needs of intelligent operations and maintenance in complex distributed environments.

[0098] Through configuration files and data acquisition interfaces, users can flexibly select the types of data to monitor. The system dynamically adjusts monitoring strategies based on user needs to optimize system load and response speed. An on-demand subscription mechanism further reduces data redundancy and bandwidth consumption, while improving the system's real-time responsiveness. Furthermore, the system introduces a hybrid monitoring strategy based on rules and machine learning. Through historical data analysis and threshold rule definition, the system automatically identifies key performance indicators and adjusts monitoring strategies in real time to address potential anomalies, improving anomaly detection sensitivity and overall system stability.

[0099] Hybrid monitoring strategy algorithm: To further enhance anomaly detection capabilities, this invention combines preset rule threshold detection with machine learning model prediction to achieve dynamic adjustment of intelligent monitoring strategies. The process is as follows:

[0100] 1. Model training initialization: Historical monitoring data is retrieved from the time series database, and key performance indicators are selected to construct a training set. A supervised learning algorithm (support vector machine (SVM)) is used to train the prediction model, learning the fluctuation range and correlation of each indicator under normal operating conditions. After training is complete, a model M is obtained that can predict short-term trends.

[0101] 2. Rule threshold setting: Based on the experience of operation and maintenance experts or regulatory requirements, initial threshold rules are set for key indicators. For example, a CPU utilization threshold of 90%, a memory usage threshold of 80%, and other related rules (such as the abnormal CPU temperature increase rate) are used as quick judgment conditions.

[0102] 3. Real-time monitoring and judgment: During system operation, the latest status data obtained in each monitoring cycle is simultaneously input into model M for prediction to obtain the expected normal range of each indicator at the next moment. If the observed value exceeds the model prediction range or violates the preset threshold rules, it is determined that a potential anomaly has occurred.

[0103] 4. Dynamic Policy Adjustment: Once a potential anomaly is detected, the system automatically increases the frequency and level of detail of data collection for related indicators (for example, increasing the frequency of data collection for abnormal indicators or increasing the frequency of log capture), and immediately triggers an alarm to notify operations and maintenance personnel. Simultaneously, the frequency of data collection for other non-critical indicators can be reduced to free up resources. If the system continues to operate stably for a period of time and all indicators are within normal ranges, the preset data collection frequency policy will be gradually restored.

[0104] 5. Self-learning and updating: Feedback the abnormal data characteristics to the model training module, and regularly use the newly accumulated operating data to update the model M and threshold settings, so that the monitoring strategy can continuously adapt to the latest system behavior characteristics.

[0105] Through the above process, the monitoring system of the present invention can "self-evolve", utilizing the reliability of fixed rules and giving full play to the adaptive ability of machine learning, to achieve keen capture of anomalies and optimal configuration of monitoring resources, ensuring efficient and stable operation of the system.

[0106] S3. Delta compression technology is used to compress node data to obtain compressed data packets, and the compressed data packets are transmitted. A multi-level transmission optimization mechanism is introduced during the transmission of compressed data packets to optimize the data transmission rate.

[0107] Among them, Delta compression technology is used to compress node data to obtain compressed data packets, and the compressed data packets are transmitted. A multi-level transmission optimization mechanism is introduced during the transmission of compressed data packets to optimize the data transmission rate, including:

[0108] S31, extracting the node data collected by the distributed monitoring system in the previous monitoring cycle and the node data collected currently and calculating the data difference. If the data difference is empty, the current data transmission is interrupted; otherwise, the next step is executed;

[0109] S32. Based on the data difference calculation result, the changed data is compressed using the Delta compression technology to obtain a compressed data packet, and the compressed data packet is published and transmitted through the message bus module;

[0110] S33. During the transmission of compressed data packets, a distributed monitoring system is introduced into a multi-level transmission optimization mechanism to perceive the transmission status of the compressed data packets in real time.

[0111] Among them, the multi-level transmission optimization mechanism includes:

[0112] Use data deduplication strategy to mark and filter duplicate data in pre-transmitted data in sequence;

[0113] Calculate the adjustment coefficient based on the current load of the distributed monitoring system and the preset maximum load threshold, and adjust the acquisition frequency according to steps S131 to S133;

[0114] A dynamic bandwidth adjustment mechanism is introduced into the distributed monitoring system. The dynamic bandwidth adjustment mechanism can perceive the current network status in real time and adaptively adjust the data transmission rate according to the available bandwidth.

[0115] The dynamic bandwidth adjustment mechanism senses the current network status in real time and adaptively adjusts the data transmission rate based on the available bandwidth, including:

[0116] Detect the transmission network status. If the bandwidth is lower than the preset threshold, the transmission rate parameter will be reduced. Otherwise, the current transmission rate will be maintained or increased.

[0117] The calculation formula of the transmission rate parameter is:

[0118] transmission_rate=g(network_status);

[0119] Wherein, transmission_rate represents the transmission rate parameter; g represents the bandwidth adjustment function; and network_status represents the bandwidth adjustment function.

[0120] S34. After receiving the compressed data packet, the receiving end parses it and updates the local status data.

[0121] It should be noted that data traffic optimization and compressed transmission include:

[0122] like Figure 2 As shown in the figure, during the data publishing and parsing process, the system adopts a series of data traffic optimization technologies to improve transmission efficiency and reduce the network burden. First, Delta compression technology (incremental compression technology) is used to compress dynamic data in real time, and only the changed parts are transmitted. This method is particularly suitable for situations where some data in multi-connected distributed monitoring systems change frequently but the amount of change is relatively small, thereby significantly reducing the amount of data transmitted and reducing the network load by about 50%. In addition, combined with the data deduplication strategy, in scenarios with high data redundancy, the transmission of duplicate data is effectively reduced, further improving bandwidth utilization efficiency. Data deduplication makes transmission more streamlined by marking and filtering duplicate data, which is particularly effective for monitoring data that needs to be updated frequently.

[0123] The calculation formula for data difference is:

[0124] |ΔD|=D prev -D current ;

[0125] Where |ΔD| is the data difference, and only transmitting this part can achieve the effect of data compression; D prev Indicates the data value collected last time; D current Indicates the currently collected data value;

[0126] The calculation formula for the adjusted acquisition frequency is:

[0127]

[0128] Where, f new Indicates the adjusted acquisition frequency; f base Indicates the basic acquisition frequency; L current Indicates the current system load; L max Indicates the maximum load allowed by the system; α is the adjustment coefficient, which controls the sensitivity of the acquisition frequency to changes in load.

[0129] It should be noted that the adjustment coefficient α changes with the current system load (system_load) relative to the maximum allowable load (max_load). By setting the minimum coefficient α min and the adjustment factor k, which ensures that the sampling frequency decreases proportionally but not too low when the load increases, and gives a new sampling frequency f new =α*base_rate. Therefore, when the system load approaches the upper limit, α drops to the preset lower limit to significantly reduce the acquisition frequency and alleviate system pressure. When the load decreases, α gradually increases to restore the normal sampling frequency to ensure data integrity.

[0130] To adapt to changing network conditions, the system also introduces a dynamic bandwidth adjustment mechanism, which can sense the current network status in real time and adaptively adjust the data transmission rate based on bandwidth availability, ensuring that critical data is transmitted first in congested situations. Through this bandwidth priority mechanism, critical monitoring data can be transmitted in a timely manner, effectively avoiding the problem of important data delays caused by network congestion. This multi-level data optimization method aims to maximize bandwidth utilization and improve the reliable transmission capability of data by compressing differences, removing duplication and redundancy, and dynamically adapting to transmission conditions. Through the above optimization methods, the system significantly reduces network resource consumption while maintaining efficient data synchronization, improving overall operation and maintenance efficiency and monitoring accuracy.

[0131] It should be noted that the bandwidth priority mechanism ensures the timely transmission of critical data by establishing a priority evaluation model and adopting a dynamic scheduling strategy. Specifically, the system sets priority weights for different categories of monitoring data (for example, assigning a higher weight to critical alarm information and a lower weight to general status information), and automatically adjusts the transmission rate of each category of data according to the real-time network bandwidth conditions. When the network bandwidth is sufficient, all types of data are sent normally at the predetermined frequency; when insufficient bandwidth or congestion is detected, the system prioritizes ensuring that high-priority data is sent at the original frequency, while appropriately reducing the transmission frequency of low-priority data. Bandwidth resources can be allocated proportionally according to priority weights. This ensures that when the total bandwidth is limited, high-priority data occupies a larger share of the bandwidth. If the network is severely congested (the available bandwidth is reduced to only enough to transmit high-priority data), the system will temporarily suspend the transmission of low-priority data and transmit it again after the bandwidth is restored. Through the bandwidth priority scheduling mechanism, the present invention prioritizes the transmission of important monitoring data when network resources are limited, effectively avoiding delays in the transmission of critical data under congestion conditions, and fully demonstrates the role of the priority strategy.

[0132] This invention provides a semi-formal algorithm flow to describe the key steps of data flow optimization and compressed transmission, which specifically include:

[0133] Algorithm: Delta compression and dynamic transmission adjustment algorithm

[0134] enter:

[0135] current_data: currently collected system status data

[0136] prev_data: the last collected data

[0137] base_rate: basic acquisition frequency

[0138] system_load: current system load

[0139] network_status: current network bandwidth status

[0140] Output:

[0141] trans_packet: The final compressed and speed-regulated data packet

[0142] The algorithm process includes:

[0143] Step 1: Calculate data differences:

[0144] Call the function ComputeDifference(current_data, prev_data) to get data_diff;

[0145] If data_diff is empty, end this transmission; otherwise, proceed to the next step.

[0146] Step 2: Delta compression processing:

[0147] Apply DeltaCompression(data_diff) to data_diff to obtain compressed_packet.

[0148] Step 3: Data deduplication processing:

[0149] Use Deduplicate(compressed_packet) to remove duplicate information and update it to unique_packet.

[0150] Step 4: Dynamic acquisition frequency adjustment:

[0151] Calculate the adjustment coefficient α. Calculate the adjustment coefficient α based on the current system load (system_load) and the preset maximum load threshold (max_load), so as to specifically implement the load adjustment function f(system_load):

[0152]

[0153] In the formula, α represents the sampling adjustment coefficient; max_load represents the preset maximum load threshold; system_load represents the current load of the distributed monitoring system; α min represents the minimum allowable adjustment coefficient (for example, set to 0.5) to ensure that the acquisition frequency will not drop too low when the system load is extremely high; k represents the adjustment coefficient scaling factor (0 < k ≤ 1), which is used to adjust the speed of the adjustment coefficient decrease.

[0154] When system_load is much lower than max_load, at this time α is close to 1, and the acquisition frequency basically remains at the base frequency; when system_load increases to max_load, α takes the larger value, so as to ensure that the acquisition frequency drops to a safe lower limit under high load.

[0155] Update the acquisition frequency:

[0156] According to the calculated adjustment coefficient α, the formula for the updated acquisition frequency (adjusted_rate) is:

[0157] adjusted_rate = base_rate × α;

[0158] Where base_rate represents the base acquisition frequency; adjusted_rate represents the updated acquisition frequency; and α represents the adjustment coefficient reflecting the impact of the current system load, which is the specific embodiment of the load adjustment function f(system_load).

[0159] Step 5: Dynamic bandwidth adjustment:

[0160] Check the current network_status. If the bandwidth is lower than the preset threshold, reduce the transmission rate; otherwise, maintain or increase the transmission rate.

[0161] Set the transmission rate parameter transmission_rate=g(network_status), where g is the bandwidth adjustment function.

[0162] Step 6: Data transmission:

[0163] Publish unique_packet through the message bus module at transmission_rate.

[0164] Step 7: Receiver processing:

[0165] After receiving the data packet, the receiver calls DeltaDecompression(unique_packet) to decompress the data;

[0166] Update local status data and record this data as prev_data to prepare for the next collection cycle.

[0167] Step 8. End or loop:

[0168] According to the system schedule, return to step 1 to collect data for the next cycle.

[0169] It should be noted that ComputeDifference means calculating the difference between the current data and the previous data, removing rows and fields that are exactly the same as prev_data.

[0170] Deduplicate means removing duplicate and redundant data content before data transmission, deduplication at the byte level.

[0171] compressed_packet represents the compressed data packet, ensuring that there are no repeated data blocks.

[0172] network_status indicates the current bandwidth usage.

[0173] DeltaDecompression decompresses incremental data at the receiving end.

[0174] prev_data represents a copy of the historical data recorded during the last transmission.

[0175] S4. Introduce the publish-subscribe model in the distributed monitoring system. Any node subscribes to other node data on demand, and the node data is stored in a predefined time series database.

[0176] Among them, the publish-subscribe model is introduced in the distributed monitoring system, where any node subscribes to other node data on demand, and the node data is stored in a predefined time series database, including:

[0177] S41. Introduce the publish-subscribe model in the distributed monitoring system. Based on the publish-subscribe model, any node subscribes to other node data on demand to achieve monitoring of global nodes or specific nodes.

[0178] It should be noted that the status data publishing and subscription of the message bus:

[0179] Establish an efficient message bus system to support the publication and subscription of status data between nodes. Based on the publish / subscribe model, any node can subscribe to data from other nodes on demand, thereby enabling global or specific node monitoring. Using the message bus module, a messaging middleware with high throughput and low latency, significantly improves the system's flexibility and scalability. Furthermore, the message bus module provides a data transmission reliability guarantee mechanism, ensuring data integrity and consistency during transmission through various redundancy check methods. This mechanism significantly reduces operational costs caused by data loss or errors, particularly in large-scale distributed monitoring systems.

[0180] S42. Storing the node data in a time series database, predicting the future load of the node data based on the historical data pattern in the time series database and combining time series analysis technology, and predicting the capacity requirements of the distributed monitoring system based on the future load of the node data.

[0181] It should be noted that state data persistence and timing analysis:

[0182] Node status data is persistently stored in a time-series database to support historical review and in-depth analysis. High-performance systems (such as InfluxDB) are used for this database, enabling efficient and structured storage of large amounts of time-series data. Combined with time-series analysis techniques, the system can predict future load conditions and performance bottlenecks based on patterns and trends in historical data, providing data support for capacity planning and performance optimization.

[0183] Specifically, time series analysis helps the system predict capacity demands before they increase, take expansion or optimization measures in advance, and avoid potential performance issues. Through data aggregation technology, the system can effectively reduce the storage requirements of historical data while ensuring the accuracy and timeliness of queries, which is especially important when accumulating large-scale data.

[0184] It should be noted that, by using data aggregation technology, the storage requirements of historical data are reduced. In terms of historical data storage optimization, the present invention adopts a sliding window data aggregation compression technology (such as the "revolving door" algorithm) to reduce the storage capacity of long-term accumulated data. Based on the trend approximation of time series, the present invention represents continuous data segments with similar change trends with a small number of key points: when a group of continuous data points has an approximately linear change trend, the system only saves the starting point and the end point of the segment, thereby merging the intermediate data. For example, set the slope tolerance threshold θ (such as 0.1), initialize the starting point P0 (x0, y0) for the new data point P i (x i ,y i ), calculate the maximum allowable deviation from the current line segment:

[0185] Upper Bound=y prev +θ·(x i -x prev );

[0186] Lower Bound=y prev -θ·(x i -x prev );

[0187] If y i If the data in the window is within the range of Lower Bound and Upper Bound, it is considered that the trend of change of the window data is consistent, and only the starting point and the end point are used to indicate that the data in this period is merged into the current segment; otherwise, the previous point is stored as the inflection point, and the segment is refitted. By sliding the window in sequence to perform the above judgment and segmentation processing, a large amount of data with a stable trend in a long time series can be compressed into several key trend lines. In particular, for historical data with a long storage period (such as more than one month), the present invention can select a looser error threshold for low-precision compression storage to significantly reduce the amount of data storage while still retaining the overall trend information. As a result, the system greatly reduces the storage overhead of historical data during long-term operation, while ensuring the accuracy and timeliness of historical trend queries. At the same time, in the present invention, the algorithm only uses this method to process historical data with a storage time of more than one year and then re-saves it into the time series database.

[0188] This method uses a standardized, cross-platform interface to acquire status data for a distributed monitoring system. This method comprehensively collects hardware utilization information, including CPU, GPU, memory, disk, and network usage, as well as customized information such as database status, solid-state drive lifespan, system cooling performance, hardware fault logs, and network latency. Users can dynamically adjust data collection strategies through the interface to avoid excessive collection of invalid data and reduce system load.

[0189] Secondly, by introducing an adaptive acquisition strategy, the system can flexibly adjust the acquisition frequency according to the current resource utilization and reduce the acquisition density under high load to reduce system pressure.

[0190] Thirdly, the present invention establishes an efficient message bus system to support the publication and subscription of status data between nodes. Based on the publish / subscribe model, any node can subscribe to the data of other nodes on demand, thereby realizing global or specific node monitoring.

[0191] In addition, through standardized data acquisition and publishing interfaces, data is collected and processed in a unified manner, enhancing the adaptability and scalability of the system. The system is also compatible with multiple platforms and architectures and supports multiple operating systems.

[0192] Finally, the present invention introduces Delta compression technology to compress dynamic data in real time, and only transmits the changed parts. It is suitable for most distributed monitoring systems where data changes frequently but the amount of change is relatively small, thereby significantly reducing the amount of transmitted data and reducing the network load by about 50%.

[0193] Comprehensive and customized data collection capabilities: Existing technical solutions lack comprehensiveness and customization in data collection, failing to meet the multi-dimensional data requirements of complex distributed monitoring systems. By deploying syswatch and its extension modules, this invention enables comprehensive data collection, including hardware utilization rates for CPU, GPU, memory, disk, and network, as well as customized information such as database status, SSD lifespan, system cooling performance, hardware fault logs, and network latency. This not only improves the comprehensiveness of data collection but also enables highly customized data collection strategies through a dynamically adjustable user interface, avoiding the excessive collection of invalid data and effectively reducing system load.

[0194] Introduction of an Adaptive Collection Strategy: Existing technologies often rely on static data collection strategies and lack dynamic adjustment capabilities, resulting in poor performance under varying system loads. This invention introduces an adaptive collection strategy, allowing the system to flexibly adjust the collection frequency based on current resource utilization. This reduces the collection frequency during periods of high load to reduce system pressure, while increasing the collection frequency during periods of low load to provide a more detailed data view. This mechanism significantly improves data collection efficiency and overall system stability.

[0195] Efficient message bus system: Existing messaging mechanisms suffer from high latency and low throughput, limiting the system's real-time monitoring capabilities. This invention establishes an efficient message bus system that utilizes a publish / subscribe model, enabling any node to subscribe to data from other nodes on demand, enabling real-time monitoring of both global and specific nodes. The message bus module offers high throughput and low latency, and employs multiple redundancy checks to ensure the integrity and consistency of data transmission, significantly enhancing the system's flexibility and scalability.

[0196] Cross-platform compatibility and standardized interfaces: Existing technologies lack cross-platform compatibility, making it difficult to unify data collection and processing across different operating systems and hardware architectures. This invention achieves unified data collection and processing through standardized data acquisition and publishing interfaces that support multiple platforms and architectures (e.g., x86, ARM) and operating systems (e.g., Linux, Windows). The use of standardized interfaces not only enhances the adaptability and scalability of the system, but also enables the flexible loading and unloading of functional modules through a dynamic module management mechanism, significantly reducing maintenance costs.

[0197] Data flow optimization and compression transmission technology: Low data transmission efficiency is a major bottleneck of existing technologies, especially in large-scale distributed monitoring systems. This invention introduces Delta compression technology to compress dynamic data in real time, transmitting only the changed parts. This technology is suitable for scenarios where data changes frequently but the amount of change is small, thereby significantly reducing the amount of data transmitted and reducing the network load by approximately 50%. In addition, combined with data deduplication strategies and dynamic bandwidth adjustment mechanisms, data transmission efficiency is further optimized, ensuring priority transmission of critical data in the event of network congestion, and improving the system's data synchronization capabilities and real-time response capabilities.

[0198] Intelligent state data persistence and time series analysis: This invention not only realizes efficient data collection and transmission, but also realizes persistent storage and in-depth analysis of node state data through the application of time series databases (such as InfluxDB).

[0199] To sum up, the present invention constructs an efficient, flexible and intelligent multi-online cross-platform distributed monitoring system status data acquisition solution through comprehensive and customized data collection, adaptive collection strategy, efficient message bus system, cross-platform compatibility and standardized interface, data flow optimization and compression transmission technology, and intelligent status data persistence and timing analysis. Compared with the existing technology, the present invention has significant advantages in the comprehensiveness of data collection, system flexibility, high efficiency of data transmission and intelligent level of system monitoring, which greatly improves the operation and maintenance efficiency and reliability of the distributed monitoring system.

[0200] The above description is only a preferred embodiment of the present invention and is not intended to limit the present invention. Any modifications, equivalent substitutions, improvements, etc. made within the spirit and principles of the present invention should be included in the scope of protection of the present invention.

Claims

1. A method for acquiring status data of a multi-online cross-platform distributed monitoring system, characterized in that: The method includes: S1. Node data is collected through a data acquisition module deployed on the nodes of the distributed monitoring system, and an adaptive acquisition strategy is introduced in the data acquisition module to adaptively adjust the node data acquisition frequency according to resource utilization; S2. Develop node data monitoring strategies based on user needs and optimize them using hybrid monitoring algorithms based on potential anomalies in the distributed monitoring system. S3. Compress the node data using Delta compression technology to obtain a compressed data packet, and transmit the compressed data packet. In the process of transmitting the compressed data packet, a multi-level transmission optimization mechanism is introduced to optimize the data transmission rate. S4. Introduce the publish-subscribe model in the distributed monitoring system. Any node subscribes to other node data on demand, and the node data is stored in a predefined time series database.

2. The method for acquiring status data of a multi-connected cross-platform distributed monitoring system according to claim 1, characterized in that: The method of collecting node data by using a data acquisition module arranged on a node of the distributed monitoring system and introducing an adaptive acquisition strategy in the data acquisition module to adaptively adjust the node data acquisition frequency according to resource utilization includes: S11, deploying the monitoring tool and executable files on all nodes of the distributed monitoring system, and reading the configuration file of the distributed monitoring system to start the data acquisition module of each node; S12. Utilize the data acquisition module to collect node data, match the node data to determine whether it can participate in data interaction with the distributed monitoring system, and perform real-time display of the online status and configuration of each node based on the data interaction results; S13. An adaptive collection strategy is introduced into the data acquisition module, and the distributed monitoring system adjusts the node data collection frequency according to the current resource utilization.

3. The method for acquiring status data of a multi-online cross-platform distributed monitoring system according to claim 2, characterized in that: The adaptive acquisition strategy is introduced into the data acquisition module, and the distributed monitoring system adjusts the node data acquisition frequency according to the current resource utilization, including: S131. Setting a maximum load threshold of the distributed monitoring system, and calculating a sampling adjustment coefficient based on a ratio function of a current load of the distributed monitoring system to the maximum load threshold; S132. If the current load of the distributed monitoring system is less than a preset maximum load threshold, the acquisition frequency is maintained at the base frequency. If the current load of the distributed monitoring system increases to the preset maximum load threshold, the sampling adjustment coefficient is set to the maximum value to reduce the acquisition frequency to a safe lower limit under high load. S133. Update the collection frequency of the distributed monitoring system based on the sampling adjustment coefficient.

4. The method for acquiring status data of a multi-connected cross-platform distributed monitoring system according to claim 3, characterized in that: The calculation formula of the sampling adjustment coefficient is: In the formula, α represents the sampling adjustment coefficient; max_load represents the preset maximum load threshold; system_load represents the current load of the distributed monitoring system; α min represents the minimum sampling adjustment coefficient; k represents the adjustment coefficient scaling factor.

5. The method for acquiring status data of a multi-connected cross-platform distributed monitoring system according to claim 4, characterized in that: The node data includes static information and customized information; The static information includes the hardware model, firmware version and operating system of the node; The customized information includes hardware usage of CPU, GPU, memory, disk, and network, time series database status, solid-state drive life, system heat dissipation performance, hardware fault logs, and network latency.

6. The method for acquiring status data of a multi-connected cross-platform distributed monitoring system according to claim 1, characterized in that: The node data monitoring strategy is formulated according to user needs, and the monitoring strategy is optimized using a hybrid monitoring algorithm based on potential abnormal conditions of the distributed monitoring system, including: S21. Extract historical monitoring data from a time series database and construct a training set based on key indicators in the historical monitoring data; train a prediction model using the training set and a supervised learning algorithm to obtain a prediction model for predicting short-term trends; S22. Setting initial threshold rules and association rules for key indicators and using them as judgment conditions; S23, extracting the latest status data obtained in each monitoring cycle during the operation of the distributed monitoring system, inputting the latest status data into the prediction model for prediction, and obtaining the observed value of the performance indicator at the next moment; S24. Determine whether the observed value meets the judgment conditions. If the observed value does not meet the judgment conditions, it is determined that a potential abnormality has occurred in the distributed monitoring system, and the collection frequency of key indicators is increased, while the collection frequency of other non-key indicators is reduced. Conversely, if the observed value meets the judgment conditions, the preset collection frequency is gradually restored.

7. The method for acquiring status data of a multi-connected cross-platform distributed monitoring system according to claim 1, characterized in that: The method of compressing node data using Delta compression technology to obtain a compressed data packet and transmitting the compressed data packet, and introducing a multi-level transmission optimization mechanism during the transmission of the compressed data packet to optimize the data transmission rate includes: S31, extracting the node data collected by the distributed monitoring system in the previous monitoring cycle and the node data collected currently and calculating the data difference. If the data difference is empty, the current data transmission is interrupted; otherwise, the next step is executed; S32. Based on the data difference calculation result, the changed data is compressed using the Delta compression technology to obtain a compressed data packet, and the compressed data packet is published and transmitted through the message bus module; S33. During the transmission of compressed data packets, a distributed monitoring system is introduced into a multi-level transmission optimization mechanism to perceive the transmission status of compressed data packets in real time; S34. After receiving the compressed data packet, the receiving end parses it and updates the local status data.

8. The method for acquiring status data of a multi-connected cross-platform distributed monitoring system according to claim 7, characterized in that: The multi-level transmission optimization mechanism includes: Use data deduplication strategy to mark and filter duplicate data in pre-transmitted data in sequence; Calculate the adjustment coefficient based on the current load of the distributed monitoring system and the preset maximum load threshold, and adjust the acquisition frequency according to steps S131 to S133; A dynamic bandwidth adjustment mechanism is introduced into the distributed monitoring system. The dynamic bandwidth adjustment mechanism can sense the current network status in real time and adaptively adjust the data transmission rate according to the available bandwidth.

9. The method for acquiring status data of a multi-connected cross-platform distributed monitoring system according to claim 8, characterized in that: The method of sensing the current network status in real time through the dynamic bandwidth adjustment mechanism and adaptively adjusting the data transmission rate according to the available bandwidth includes: Detect the transmission network status. If the bandwidth is lower than the preset threshold, the transmission rate parameter will be reduced. Otherwise, the current transmission rate will be maintained or increased. The calculation formula of the transmission rate parameter is: transmission_rate=g(network_status); Wherein, transmission_rate represents the transmission rate parameter; g represents the bandwidth adjustment function; and network_status represents the bandwidth adjustment function.

10. The method for acquiring status data of a multi-connected cross-platform distributed monitoring system according to claim 1, characterized in that: The introduction of the publish-subscribe model in the distributed monitoring system, where any node subscribes to other node data on demand, and the node data is stored in a predefined time series database, includes: S41. Introducing a publish-subscribe model in a distributed monitoring system. Based on the publish-subscribe model, any node can subscribe to data from other nodes on demand to achieve monitoring of global nodes or specific nodes. S42. Storing the node data in a time series database, predicting the future load of the node data based on the historical data pattern in the time series database and combining time series analysis technology, and predicting the capacity requirements of the distributed monitoring system based on the future load of the node data.

Images