Hyper-converged server multi-resource integration system and scheduling method

Abstract

The invention discloses a hyper-converged server multi-resource integration system and a scheduling method, belongs to the technical field of computer resource management, and aims at solving the problems that a traditional hyper-converged server is dispersed in resource scheduling, low in storage efficiency, difficult to identify abnormities and the like. Multi-source data is obtained by means of distributed acquisition nodes, a distributed time sequence database architecture is adopted for storage, a standard data set containing a multi-dimensional index is constructed, and efficient storage and rapid retrieval of the data are achieved. Positioning target data to construct a graph model, building a layered and partitioned distributed graph database, and capturing change events in real time to realize dynamic updating. And mining a resource causal relationship by using a graph neural network, and screening an effective causal chain. Meanwhile, according to a graph model, monitoring weights of nodes and connecting edges are calculated, differential monitoring is implemented, potential abnormal points are accurately identified, and a resource scheduling strategy is generated in combination with a causal relationship. Real-time feedback adjustment and database updating are performed during execution, multi-resource deep integration and intelligent scheduling are achieved, and the resource utilization rate and stability of the system are remarkably improved.

Description

Technical Field

[0001] The present invention relates to the technical field of computer resource management, and more particularly to a hyper-converged server multi-resource integration system and a scheduling method. Background Art

[0002] Hyperconverged servers, as a key architecture for achieving intensive resource management in data centers, are facing multi-dimensional technical challenges in responding to rapid business iterations and cluster expansion. With the deep integration of heterogeneous hardware devices and diverse business systems, format differences and semantic ambiguity in data interfaces across different vendors are becoming increasingly prominent. Traditional centralized data processing architectures, lacking standardized adaptation mechanisms, result in inefficient multi-source data integration and struggle to form a unified resource management data foundation. Furthermore, the dynamic evolution of resource topology relationships and business logic dependencies makes it impossible for static modeling methods to capture the temporal changes between physical connections and logical interactions in real time. Correlation analysis results often lag behind the actual system operation, making it difficult to support refined resource scheduling requirements. At the anomaly management level, the existing threshold alarm mechanism cannot effectively distinguish between accidental fluctuations and systemic failures, and lacks the ability to trace the cause and effect of anomaly propagation paths, which frequently leads to false positives, missed reports, or deviations in root cause location. In the process of generating resource scheduling strategies, the optimization model based on empirical rules or single objectives is difficult to achieve a global optimal balance between resource utilization, business continuity, and operating costs, and the policy execution effect deviates significantly from expectations. At the same time, the management system lacks a closed-loop adjustment link from anomaly identification, policy execution, to effect feedback, and is unable to dynamically optimize resource allocation according to real-time operating status. Long-term operation can easily lead to system performance degradation and resource fragmentation. Therefore, in order to overcome these limitations, the present invention proposes a hyper-converged server multi-resource integration system and scheduling method. Summary of the Invention

[0003] In response to the shortcomings of the existing technology, the purpose of the present invention is to provide a hyper-converged server multi-resource integration system and scheduling method, which solves the problems of low efficiency and standardization of heterogeneous multi-source data integration, difficulty in real-time dynamic modeling of resource association relationships, insufficient accuracy in locating the root causes of resource anomalies, lack of systematic optimization of resource scheduling strategies, and the lack of a dynamic closed-loop management mechanism, which leads to the inability of resource allocation to evolve continuously.

[0004] To achieve the above object, the present invention provides the following technical solutions:

[0005] Hyper-converged server multi-resource integration system, including:

[0006] Multi-source datasets are collected through distributed collection nodes. After semantic alignment and format conversion, they are stored in a distributed time series database architecture to build a standard dataset with a multi-dimensional index structure.

[0007] Locate the target data of the graph model based on the multi-dimensional index structure of the standard data set, build the graph model, use hierarchical partitioning to build a distributed graph database, and trigger the update of the graph database by capturing change events;

[0008] The graph neural network algorithm is used to extract the temporal and spatial topological properties of graph nodes, and a sliding window is used to construct a temporal graph. The structural causal model is used to generate directed causal edges, and valid causal chains are selected and stored in the causal relationship library. When the graph database is updated, the causal relationship incremental update process is started, and the valid causal chains in the causal relationship library are updated by dividing the local graph model.

[0009] Based on the monitoring weights of graph nodes and connecting edges in the graph model, the collection cycle is adjusted based on the basic monitoring frequency. Differentiated monitoring is implemented through performance indicator data to identify potential anomalies. The target causal set is constructed through the causal relationship library. The multi-objective optimization function is constructed through the multi-objective optimization engine. The target candidate strategy for resource scheduling is generated and sent to the target node for resource scheduling.

[0010] Monitor execution results in real time, conduct feedback adjustments, and update the graph database and causal relationship library.

[0011] Specifically, the steps to identify potential anomalies include:

[0012] Based on the historical performance indicator data of the graph model, statistical analysis is performed on the performance indicator data of the graph nodes and connected edges to generate a basic threshold range including the mean and standard deviation;

[0013] Based on the frequency of potential anomalies in the graph nodes and edges in the graph model, differentiated threshold adjustments are made to the basic threshold ranges of the graph nodes and edges to construct dynamic threshold ranges for the performance indicator data of the graph nodes and edges.

[0014] Compare the performance indicator data collected in real time by the graph model with the corresponding dynamic threshold range. If it is within the corresponding dynamic threshold range, perform time series mutation monitoring. Based on the change range of the performance indicator data, determine whether to trigger a resource anomaly warning;

[0015] If the performance indicator data collected in real time is outside the corresponding dynamic threshold range, the collection point is marked as a potential anomaly point, and outlier identification is performed on the potential anomaly point. If the potential anomaly point is determined to be an outlier, time series mutation monitoring is performed to determine whether a resource anomaly warning is triggered;

[0016] If a potential outlier is determined to be a non-outlier, resource scheduling path analysis is triggered, and the graph node where the potential outlier is located is marked as an abnormal graph node, and the connection edge where the potential outlier is located is marked as an abnormal connection edge.

[0017] Specifically, the steps for generating a target candidate strategy for resource scheduling include:

[0018] When resource scheduling path analysis is triggered, a double-layer screening mechanism is initiated in the causal relationship library based on the timestamp of the potential abnormal point, screening the target causal chain and constructing the target causal chain set;

[0019] Extract the causal subject and causal object of the target causal chain, map them to the graph nodes in the current graph model using the resource ID of the graph database, trace the root causal subject, and identify the graph model to be scheduled in the graph database as the target graph model;

[0020] For the graph nodes of the root causal subject and the graph nodes associated through the target causal chain in the target graph model, a multi-objective optimization function is constructed through the multi-objective optimization engine to generate a set of candidate strategies. The multi-objective optimization function includes minimizing the resource utilization rate of the graph nodes of the root causal subject and the graph nodes associated through the target causal chain, minimizing the loss of the resource scheduling service link, and minimizing the execution cost of the resource scheduling operation.

[0021] The candidate strategies are verified through the cluster simulation model, the target candidate strategy is selected, and it is decomposed into an execution instruction chain and sent to the scheduling execution module. At the same time, the causal strength of the target causal chain in the causal relationship library is updated according to the execution results.

[0022] Specifically, the specific steps of executing resource scheduling and real-time monitoring of execution results for monitoring feedback and adjustment include:

[0023] The resource orchestration engine parses the execution instruction chain of the target candidate policy into an atomic operation sequence. It generates an instruction chain based on the operation type and target node and maps it into a call sequence. The call sequence is then sent to the target node through a message queue.

[0024] Execute the call sequence for the target node in sequence and set the target node's scheduling cool-down period. Configure the scheduling observation period according to the target node's monitoring weight. During the scheduling observation period of executing resource scheduling, use high-frequency anomaly monitoring on the target node to identify potential anomalies.

[0025] Count the frequency of potential anomalies within the scheduling observation period, configure the scheduling anomaly threshold, and determine whether a scheduling anomaly warning is triggered. If so, block the resource scheduling path, pause the call sequence execution through the message queue, perform a rollback operation, and release the scheduling cool-down period of the target node.

[0026] Eliminate the pseudo causal chain that triggers the scheduling exception warning, re-identify the target causal chain in the causal relationship library, update the target candidate strategy, map the call sequence, and execute the call sequence on the target node again;

[0027] If the frequency of potential anomaly points is less than or equal to the scheduling anomaly threshold, the anomaly is confirmed to be resolved, resource scheduling observation is stopped, and the abnormal status label and scheduling timestamp of the node in the graph database are synchronously updated;

[0028] Locate the graph nodes and connection edges in the resource scheduling path involved in the call sequence, update the causal relationship data volume and graph database, and record the full process log of the call sequence execution.

[0029] Specifically, the steps for effective causal chain screening include:

[0030] The graph neural network algorithm is used to extract features from the graph models in the distributed graph database. The temporal attributes and spatial topological attributes of the graph nodes are used as input to construct a temporal graph structure of resource state transitions.

[0031] Quantify the causal influence direction between nodes in the time series graph structure and generate directed causal edges containing time lag relationships;

[0032] Statistical significance verification is performed on the extracted directed causal edges, and valid causal chains are screened by setting a causal strength threshold;

[0033] When the graph database triggers an update, the causal relationship incremental update process is automatically started. By dividing the affected local graph model in the graph model, the temporal attributes and spatial topological attributes of the graph nodes in the local graph model are re-extracted, and the directed causal edges are updated through the incremental graph neural network.

[0034] Convert the effective causal chain into causal records according to the standardized structure and store them in the causal relationship database. The causal records include the causal subject, causal object, causal type, causal strength and storage timestamp;

[0035] The aging mechanism is used to delete causal records in the causal relationship library that have not been used for more than a preset time period, and the clustering algorithm is used to merge the causal records based on similarity to optimize the causal relationship library.

[0036] Specifically, the steps for obtaining a graph database include:

[0037] Locate the target data of the graph model based on the multi-dimensional index structure of the standard data set;

[0038] Graph nodes are constructed based on the resource entities in the target data of the graph model. The acquisition timestamp and spatial topology label are reused to assign time and space attributes to the graph nodes respectively, and real-time performance indicator data is obtained as status attributes.

[0039] Generate physical connection edges of graph nodes based on spatial topology labels in standard datasets, and construct a time series graph between graph nodes based on the timestamp sequence of business interaction records of graph nodes. By calculating the logical dependency strength between graph nodes, select graph node pairs and generate logical dependency edges.

[0040] For the connection edges of graph nodes, based on the graph model target data of the standard dataset, the request count and traffic data of the physical connection edges and logical dependency edges are extracted, and edge weights are assigned to the physical connection edges and logical dependency edges respectively according to the weighted priority mapping algorithm;

[0041] A hierarchical partitioning strategy is adopted to build a distributed graph database storage architecture, in which graph nodes and connection edges in the graph model are stored in two-level shards according to resource type and spatial location, and a multi-dimensional composite index system is established.

[0042] Specifically, the steps for obtaining the graph database also include:

[0043] Capture change events of standard datasets in real time. Change events include resource status changes, resource configuration adjustments, or business rule changes. When a change event of a standard dataset is captured, the graph database update process is triggered:

[0044] Based on the change event type and resource ID, the affected target shard area of ​​the graph model is located through the multi-dimensional composite index system of the distributed graph database.

[0045] For the located target shard area, update the graph node attributes of the involved graph model, dynamically adjust the connection edges of the graph nodes, and update the edge weights of the connection edges;

[0046] After the graph database update process is completed, the transaction consistency mechanism of the distributed graph database is used to perform integrity verification on the updated target shard area of ​​the graph model, and the update results are synchronized to the remaining shard areas of the graph database.

[0047] Specifically, the steps for constructing a standard dataset include:

[0048] Receive multi-source data sets through distributed message queues, manage sub-queues, initialize distributed time series databases, and establish data docking channels;

[0049] Build a semantic model based on domain ontology, define standard terms and relationship attributes of resource entities, and form a standardized semantic system;

[0050] The rule engine maps heterogeneous format terms in multi-source datasets to a unified semantic space. A plug-in-based adaptive conversion architecture is used to dynamically load conversion plug-ins to convert heterogeneous format terms into standardized columnar data.

[0051] Record the distributed node acquisition timestamps, obtain network topology information, and dynamically generate spatial topology labels for multi-source datasets;

[0052] LSM tree indexing and columnar storage are used in distributed time series databases. Data from multi-source datasets are classified into hot and cold categories based on their time dimension, access frequency, and business priority to adjust shard storage and the number of replicas.

[0053] Integrate spatial topology labels and resource type identifiers to build a multidimensional index structure that includes time, space, and resource types, and build a metadata index; desensitize and encrypt sensitive data for storage, and record data operation behaviors.

[0054] Specifically, the steps for adjusting the collection period based on the basic monitoring frequency and implementing differentiated monitoring through performance indicator data include:

[0055] Based on the historical activity, business criticality, and data interaction frequency of each graph node in the graph database, multi-dimensional feature fusion is used to configure the weight coefficients of the feature dimensions, and a weighted summation method is used to generate node weights in the range of 0 to 1.

[0056] Configure weight classification thresholds, including upper and lower thresholds, and implement differentiated monitoring based on monitoring weights, which include node weights and edge weights.

[0057] When the monitoring weight is greater than the upper threshold of the weight classification, high-frequency abnormal monitoring is adopted. On the basis of the preset basic monitoring frequency, the collection cycle is shortened according to the configured shortening ratio;

[0058] When the monitoring weight is less than the lower threshold of the weight classification, low-frequency abnormal monitoring is adopted. On the basis of the preset basic monitoring frequency, the collection cycle is increased according to the configured growth ratio;

[0059] When the monitoring weight is less than or equal to the upper threshold of the weight classification, and greater than or equal to the lower threshold of the weight classification, the collection cycle is set according to the basic monitoring frequency to perform abnormal monitoring;

[0060] Differentiated monitoring of graph nodes or connection edges is performed according to the dynamic collection cycle, and performance indicator data of graph nodes and connection edges is collected.

[0061] The multi-resource scheduling method for a hyper-converged server includes the following steps:

[0062] Step S1: Collect multi-source datasets through distributed collection nodes, perform semantic alignment and format conversion on the multi-source datasets, and then use a distributed time series database architecture to store them and build a standard dataset with a multi-dimensional index structure.

[0063] Step S2: Locate the target data of the graph model based on the multidimensional index structure of the standard data set, build the graph model, use hierarchical partitioning to build a distributed graph database, and trigger the update of the graph database by capturing change events;

[0064] Step S3: The graph neural network algorithm is used to extract the temporal attributes and spatial topological attributes of the graph nodes. A temporal graph is constructed by combining the sliding window. The structural causal model is used to generate directed causal edges, and valid causal chains are selected and stored in the causal relationship library. When the graph database is updated, the causal relationship incremental update process is started to update the valid causal chains in the causal relationship library by dividing the local graph model.

[0065] Step S4: Based on the monitoring weights of the graph nodes and connecting edges in the graph model, the collection period is adjusted on the basis of the basic monitoring frequency. Differentiated monitoring is implemented through performance indicator data to identify potential anomalies. The target causal set is constructed through the causal relationship library. The multi-objective optimization function is constructed through the multi-objective optimization engine. The target candidate strategy for resource scheduling is generated and sent to the target node for resource scheduling.

[0066] Step S5: Monitor the execution results in real time, perform monitoring feedback adjustments, and update the graph database and causal relationship library.

[0067] Beneficial effects of the present invention:

[0068] The present invention realizes efficient standardized integration and elastic storage of heterogeneous multi-source data through a distributed time series database with multi-dimensional indexing, effectively solving the problems of data format heterogeneity and semantic ambiguity; with the help of a hierarchical and partitioned dynamic graph database, it captures resource topology evolution and time series dependencies in real time, realizes dynamic modeling of physical connections and logical interactions, and provides an accurate carrier for resource association analysis; uses graph neural networks and structural causal models to deeply mine time series causal chains, constructs a causal relationship library with both statistical significance and business rationality, improves the accuracy of locating the root cause of anomalies and reduces the misjudgment rate; based on a multi-objective optimization engine combined with cluster simulation, it generates a global optimization strategy that takes into account multi-dimensional needs, breaking through the limitations of traditional single-objective scheduling; accurately distinguishes anomaly types through dynamic threshold adjustment and outlier identification mechanism, and realizes continuous evolution of resource management in combination with a closed-loop feedback regulation system, ultimately greatly improving the resource integration efficiency, scheduling strategy effectiveness and system reliability of hyper-converged servers. BRIEF DESCRIPTION OF THE DRAWINGS

[0069] Figure 1 This is a schematic diagram of the structure of the multi-resource integration system of the hyper-converged server of the present invention;

[0070] Figure 2 A flowchart of the specific steps for constructing the standard data set of the present invention;

[0071] Figure 3 A flowchart of the specific steps of constructing a graph model of the present invention;

[0072] Figure 4 Generate a flow chart of resource scheduling strategy for the present invention;

[0073] Figure 5 This is a flow chart of the multi-resource scheduling method for a hyper-converged server according to the present invention. DETAILED DESCRIPTION

[0074] Example 1:

[0075] See also Figure 1 ,This embodiment introduces a hyper-converged server multi-resource integration system,,including an edge acquisition module, a converged storage module, an,association decision module and a scheduling execution module;

[0076] The edge acquisition module is used to collect computing, storage, network and other multi-resource data of each acquisition node in the hyper-converged server in real time through distributed deployment of acquisition nodes, and perform edge preprocessing to build a multi-source data set that can be analyzed by the upper-level system. It builds a standardized driver interface through the hardware abstraction layer, uniformly encapsulates the underlying protocols of ARM, X86 servers and storage and network devices, realizes heterogeneous access device driver adaptation, collects original multi-source data, and uses adaptive data parsing algorithms for each distributed node. Based on the data type and acquisition frequency of the original multi-source data, the preset templates are matched with dynamic rules to complete the unification of the heterogeneous original multi-source data format. The invalid data of the original multi-source data is cleaned through the general rule base and semantic analysis, and the data volume is reduced by combining the hybrid compression algorithm to form a multi-source data set with unified format, streamlined compression and timeliness.

[0077] In this embodiment, a lightweight collection agent is deployed on each physical node or rack unit in a hyper-converged server cluster as a collection node. These collection nodes form a distributed collection network via a high-speed backplane or 10 Gigabit Ethernet, enabling parallel data collection. The hardware abstraction layer utilizes a layered driver architecture, consisting of a device adaptation layer, a protocol conversion layer, and a data abstraction layer. Standardized driver plug-ins adapt to ARM and X86 servers, storage, and network devices, uniformly convert proprietary protocols, and provide standard interfaces, enabling unified access to heterogeneous devices. An adaptive data parsing algorithm automatically identifies data types and collection frequencies, selects corresponding data format templates from a predefined template library, and performs standardized data conversion based on these templates, unifying heterogeneous data into a unified format. A general rule library integrates multiple invalid data identification rules and combines semantic analysis with data cleansing. A hybrid compression algorithm employs a hierarchical strategy, using Snappy for fast compression of real-time data and Zstandard for deep compression of non-real-time data. CRC checksums ensure data integrity. Each collection node has a built-in edge cache queue, employing a ring buffer design. This cache prevents data loss during transient network congestion. At the same time, the collection node supports the breakpoint resume function, automatically retransmitting unsent data after the network is restored, ensuring the integrity and timeliness of multi-source data sets.

[0078] The fusion storage module performs semantic alignment and format conversion on the multi-source datasets output by the edge acquisition module. It then uses a distributed time-series database architecture for intelligent storage, constructing a multidimensional index structure encompassing time, space, and resource type dimensions to form a standard dataset that supports high-speed retrieval and cross-domain correlation analysis. By establishing a semantic model to unify the semantic representation of multi-source data, an adaptive conversion mechanism is employed to automatically adapt to data format changes caused by hardware upgrades or business changes. A clock synchronization algorithm is used to eliminate time errors between nodes, and topology monitoring technology is used to update data spatial topology labels in real time, enabling data spatiotemporal attribute annotation. During storage, data reliability is ensured by optimizing the time-series database's write performance and incorporating redundancy strategies. Furthermore, based on access patterns such as data read and write frequency and real-time requirements, an algorithm is employed to dynamically adjust the data shard size and number of replicas, reducing storage costs while ensuring data availability. The final output is a standard dataset that integrates spatiotemporal attributes, boasts efficient storage, and a standardized structure, providing structured data support for upper-level modules such as intelligent correlation analysis and decision-making scheduling.

[0079] In this embodiment, the fusion storage module adopts a distributed time series database architecture and implements multi-source data access buffering through message queues. Based on the semantic model, the data terminology is unified, and the data format changes of the hardware and business systems are dynamically adapted through a plug-in mechanism. A hybrid clock synchronization strategy is adopted to eliminate time errors between nodes, and the data space topology labels are updated in real time by combining software-defined networking and container orchestration technology. Write performance is improved through index optimization and column storage, and data reliability is guaranteed by combining a hybrid redundancy strategy of multiple copies and erasure codes. Based on the online learning algorithm, data access patterns are analyzed, the shard size and the number of copies are dynamically adjusted, and a differentiated storage strategy for hot and cold data is implemented. Finally, data is stored in a standardized columnar format, and a multi-dimensional index structure is constructed to support fast time range queries and cross-resource association retrieval.

[0080] See also Figure 2 , preferably, the specific steps of constructing the standard data set include:

[0081] The data flow in the multi-source datasets output by the edge acquisition module fluctuates, and the real-time requirements for data from different resource types vary significantly. The multi-source datasets output by the edge acquisition module are received through a distributed message queue, and a flow control mechanism is used to achieve dynamic buffering and load balancing of data flow. A message classification mechanism based on resource type priority is adopted, with priorities divided according to real-time requirements. Multi-source datasets are queued and managed by hyper-converged server resource type, including computing resources, storage resources, and network resources, ensuring that high-timeliness data is processed first. At the same time, the distributed time series database architecture is initialized and a data connection channel with the distributed message queue is established, laying the foundation for subsequent data storage.

[0082] Hyper-converged servers involve multiple vendors and multiple types of resources, and the original data terms are heterogeneous. A semantic model is built based on the domain ontology to define the standard terms and relationship attributes of hyper-converged server resource entities. Resource entities include computing resource classes, storage resource classes, network resource classes, security resource classes, etc. Object attributes are used to define the relationship between resource entities, and data attributes are used to define entity characteristics to form a standardized semantic system.

[0083] Hardware vendor upgrades or business system changes can cause dynamic changes in data formats. This requires ensuring data format compatibility with the distributed time series database. Using a rules engine, heterogeneous terms in multi-source datasets are mapped to a unified semantic space. For example, different expressions such as CPU usage and memory utilization can be uniformly mapped to standard concepts such as computing resources, processors, and utilization. Using a plug-in-based adaptive conversion architecture, corresponding conversion plug-ins are dynamically loaded to address data format changes caused by hardware vendor upgrades or business system changes, converting heterogeneous terms in multi-source datasets into standardized columnar data. Ensuring data format compatibility with the distributed time series database architecture provides the prerequisite for subsequent storage and retrieval optimization.

[0084] Hyper-converged server clusters are distributed architectures. Node time errors can cause data time sequence distortion. Spatial topology changes require real-time awareness to support spatiotemporal correlation analysis. High-precision clock servers are deployed to record the acquisition timestamps of distributed nodes. In conjunction with software-defined networking, network topology information is acquired in real time. The container orchestration platform is linked to monitor the spatial coordinates and logical groupings of distributed nodes, dynamically generating spatial topology labels for multi-source datasets. Spatiotemporal encoding is used to integrate acquisition timestamps and spatial coordinates into a composite index key, enabling precise spatiotemporal correlation.

[0085] Hyper-converged server clusters generate a large amount of monitoring data daily. Random writes can cause database performance degradation, and a balance must be struck between data reliability and storage costs. For connected multi-source datasets, LSM tree index optimization and columnar storage layout are used in distributed time-series databases to improve the writing and query performance of massive data. LSM tree indexes convert random writes into sequential writes, improving the writing performance of massive data. Columnar storage is compressed by resource type, encoding and compressing numerical data, and adopting a hybrid redundancy strategy of multiple copies and erasure codes. Specifically, by configuring a dynamic update cycle, copies of multi-source dataset data within the update cycle are retained to ensure high availability, and erasure codes are applied to historical data of multi-source datasets outside the update cycle to reduce storage costs. The dynamic update cycle is determined by the frequency of data access, business real-time requirements, and data lifecycle strategies. A distributed consensus algorithm is used to ensure consistency during the writing process of multi-source dataset data, while supporting data redundancy deployment across availability zones.

[0086] Data access patterns vary significantly, and a unified storage strategy would waste resources. We continuously analyze data access patterns using online learning algorithms, classifying multi-source data sets into hot and cold categories based on their time dimension, access frequency, and business priority. For hot data, such as resource monitoring data from the past seven days, we use small shards for storage and configure replicas to improve real-time query efficiency. For cold data, such as log data from 30 days ago, we merge it into large shards and reduce the number of replicas. Shard merging and replica migration are performed through an asynchronous scheduling mechanism to optimize storage resource utilization.

[0087] Upper-level applications need to support complex queries, and a single index cannot meet cross-dimensional retrieval requirements. Based on time series, we integrate spatial topology tags and resource type identifiers to build a multidimensional index structure that encompasses time, space, and resource types. B+ tree indexes are used in the time dimension to enable millisecond-level range queries. R-tree indexes are used in the spatial dimension to support topological proximity retrieval. An inverted index is constructed in the resource type dimension to accelerate classified and aggregated queries. An automatic index optimization mechanism dynamically adjusts index weights and sharding strategies based on actual query patterns, improving the efficiency of cross-dimensional complex queries.

[0088] Data traceability is the foundation of quality control and compliance auditing. Sensitive data must meet privacy protection requirements, and operational behaviors must be traceable. Record the entire chain of processing from the acquisition of original multi-source data sets to standard data sets, generate data lineage metadata, and build metadata indexes to assist upper-level applications in quickly locating data assets. Based on data classification and grading standards, implement field-level desensitization and encrypted storage for sensitive data; adopt role-based access control combined with attribute permission management to dynamically control data access rights based on user identity, operation scenarios, and other dimensions. Deploy a full-chain audit mechanism to record data access, storage, query, and other operational behaviors to ensure that data usage meets privacy protection and compliance requirements. A standard dataset with the following characteristics is formed: at the semantic level, terminology unification and relationship modeling are achieved through ontology models to eliminate cross-source data ambiguity; standardized columnar storage is used in the format to support efficient compression and sharding processing; precise timestamps and dynamic spatial topology labels are integrated in the spatiotemporal dimension to meet the needs of spatiotemporal correlation analysis; millisecond-level time range queries and cross-resource type correlation retrieval are achieved through multi-dimensional indexing and storage strategy optimization; data quality is guaranteed through full-process verification, and security mechanisms ensure controllable data access, providing standardized and highly available data support for upper-level applications such as intelligent analysis, decision-making and scheduling.

[0089] The association decision module is used to build a graph model based on a structured standard data set, apply the graph neural network algorithm, and combine time series and spatial topology information to mine the causal relationship between resources, generate resource scheduling strategies, and improve the response speed of complex queries by optimizing the index of the graph database and reconstructing the query statements; construct a multi-objective optimization function, combine the reinforcement learning scheduling algorithm to generate the resource scheduling strategy, and use simulation technology to build a cluster simulation model to predict and evaluate the resource scheduling strategy; use the rule conflict detection and resolution algorithm to perform logical verification on the strategy, and use the algorithm that combines rule matching and case reasoning to achieve fast and intelligent matching of strategies, completing data value mining and decision strategy generation.

[0090] In this embodiment, the association decision module constructs a graph model containing resource nodes and associated edges, and uses a graph neural network algorithm to fuse time series and spatial topology information to mine implicit causal chains between resources. At the same time, it implements hybrid index optimization and query statement reconstruction on the graph database, significantly improving the response speed of complex queries. The association decision module constructs a multi-objective optimization function, combines it with a reinforcement learning algorithm to generate resource scheduling strategies, and uses a cluster simulation model to predict and evaluate various load scenarios, effectively improving resource utilization efficiency. It detects and resolves policy conflicts through a rule engine, and reuses historical strategies through case-based reasoning to achieve rapid and intelligent matching of strategies. Ultimately, it generates a scheduling strategy that balances resource efficiency and business reliability, supporting dynamic system optimization.

[0091] See also Figure 3 , preferably, the specific steps of constructing the graph model include:

[0092] Standard datasets cover massive amounts of data across all hyperconverged server resources. Direct retrieval is extremely inefficient, and resource scheduling and analysis typically focus on data within specific time periods, regions, or types, requiring precise screening. Based on the multidimensional index structure of standard datasets, the time index is used to filter resource status data within the target time period. The spatial index is used to obtain resource topology information for the target region. The resource type index is used to extract target resource type data to locate the target data for the graph model.

[0093] Hyper-converged server resources are diverse and have varying attributes. Physical and virtual resources need to be abstracted into graph nodes, using a unified model to describe resource status and characteristics for subsequent correlation analysis. Graph nodes are constructed based on resource entities in the graph model's target data, and their unique codes are used as graph node identifiers. The acquisition timestamp and spatial topology labels are reused to assign time and spatial attributes to the graph nodes, respectively. Real-time performance indicator data is obtained as status attributes to complete node construction.

[0094] Resources in a hyperconverged server not only have physical connections but also business logic dependencies. Considering only physical connections cannot fully reflect the essence of resource interactions; logical dependencies must also be explored. Based on the spatial topology labels in a standard dataset, physical connection edges are generated for graph nodes, and edge attributes such as bandwidth and latency are annotated. Based on the timestamp sequence of the graph nodes' business interaction records, a timing diagram is constructed between the graph nodes, such as request initiation time and response completion time. The strength of the logical dependencies between graph nodes is calculated using a Granger causality test. For graph node pairs whose logical dependency strength exceeds a preset logical connection threshold, logical dependency edges are generated to determine resource connectivity. Physical connection edges reflect the physical connectivity between devices and support fault location, while logical dependency edges reflect the implicit dependencies in business operations and support performance bottleneck analysis.

[0095] Different connection edges have varying degrees of impact on resource scheduling and system operation, necessitating quantification of their importance to inform priority decisions. For the connection edges of graph nodes, including physical connection edges and logical dependency edges, we extract the request counts and traffic data for these edges based on the graph model target data of the standard dataset. These edges are then assigned edge weights using a weighted priority mapping algorithm. This algorithm constructs a weight calculation factor library, sets differentiated weight calculation factors for the different characteristics of physical connection edges and logical dependency edges, and calculates edge weights using a weighted summation formula.

[0096] Hyper-converged servers are large in scale, and the amount of graph model data grows rapidly over time. Single-server storage cannot meet performance and scalability requirements, necessitating a distributed architecture. A distributed graph database storage architecture is constructed using a hierarchical partitioning strategy. Graph nodes and connecting edges in the graph model are stored in two-level shards based on resource type and spatial location, achieving complete coverage of all hyper-converged server resource data within the target time period. A multi-dimensional composite index system is established, using resource ID as the primary key. A three-level index structure is constructed using spatial location encoding and resource type tags. A prefix tree algorithm is used to optimize index hierarchical relationships, ensuring millisecond-level node retrieval and relationship query performance even at a scale of tens of millions of nodes.

[0097] The resource status and configuration of hyperconverged servers change dynamically, and graph models need to be updated in real time to reflect the latest situation and avoid invalid analysis results. Change events of standard datasets are captured in real time, including changes in resource status, resource configuration adjustments, or business rule changes. When a change event of a standard dataset is captured, the graph database update process is triggered. Based on the change event type and resource ID, the multi-dimensional composite index system of the distributed graph database is used to quickly locate the affected target shard area of ​​the graph model.

[0098] Resource status changes or resource addition and deletion operations occur frequently, and graph node attributes need to be updated synchronously to ensure model accuracy. For the located target shard area, the graph node attributes of the involved graph model are updated. Specifically, if the change event is a resource status change, such as a sudden change in computing resource load, the system directly extracts the latest real-time performance indicator data from the updated standard data set and replaces the corresponding state attributes of the graph node; if it is a resource addition or deletion event, the corresponding graph node is added or deleted based on the resource entity information of the standard data set, and its associated edge relationship is updated synchronously.

[0099] Network topology reconstruction and changes in business rules can lead to changes in resource connectivity, requiring synchronous updates to edges to accurately depict resource interactions. The system dynamically adjusts the edges connecting graph nodes. If physical connections change due to network topology reconstruction, the system regenerates or deletes the physical edges based on the updated spatial topology labels in the standard dataset, and updates edge attributes such as bandwidth and latency. If business rule changes lead to changes in logical dependencies, the system reconstructs the time series graph based on the updated timestamp sequence of business interaction records. The system recalculates the logical dependency strength using Granger causality tests, and updates the weights or reconstructs the relationships for logical dependency edges whose strength changes exceed a threshold.

[0100] Data such as edge traffic and request counts change in real time, and the original weights no longer reflect their latest importance, requiring recalculation to adapt to scheduling needs. To recalculate the edge weights of graph nodes, the updated request count and traffic data are extracted from the standard dataset. A weighted priority mapping algorithm is then invoked to reweight each edge based on the differential weight calculation factors for physical connection edges and logical dependency edges.

[0101] In a distributed environment, data updates can lead to inconsistencies due to network failures and node anomalies. After the graph database update process is complete, the distributed graph database's transaction consistency mechanism performs an integrity check on the updated target shard area of ​​the graph model and synchronizes the updated results to the remaining shard areas of the graph database, ensuring the consistency and accuracy of the entire graph model data.

[0102] Preferably, the specific steps of mining the causal relationship between resources include:

[0103] The causal relationship between resources in a hyper-converged server is complex and implicit. Traditional methods find it difficult to capture nonlinear and temporal dependencies, and require deep learning to explore potential relationships. The graph neural network algorithm is used to extract features from graph models in distributed graph databases. The temporal attributes and spatial topological attributes of graph nodes are used as input, and a sliding window mechanism is combined to construct a temporal graph structure for resource state changes. The structural causal model in the causal graph neural network is used to quantify the causal influence direction between graph nodes, generate directed causal edges containing time lag relationships, and realize the explicit extraction of implicit causal relationship chains.

[0104] Causal edges extracted solely through algorithms may be misjudged and require dual verification using both statistics and business logic to ensure the reliability and practicality of causal relationships. The extracted directed causal edges are statistically significant based on the Granger causality test algorithm. By constructing a vector autoregression model, lagged variables are introduced into the model, and regression analysis is performed on time series data to calculate the causal strength of the directed causal edges. Valid causal chains are then screened by setting a causal strength threshold. For example, for the temporal relationship between a sudden increase in storage IOPS and a database transaction delay, the causal probability value within the lag order is calculated. If it exceeds a preset threshold, it is considered a valid causal relationship. At the same time, valid causal chains are verified in conjunction with the business rule library to eliminate pseudo-causal relationships that contradict actual business logic and ensure the business rationality of the causal chain.

[0105] Hyper-converged server resources change dynamically, and static causal relationships cannot reflect real-time status, requiring real-time updates to prevent analysis failures. When a graph database triggers an update, the causal relationship incremental update process is automatically initiated. The affected local graph model is partitioned, and the temporal and spatial topological properties of the graph nodes in the local graph model are re-extracted. Directed causal edges are then updated using an incremental graph neural network. For example, when a new server is added to the storage cluster, the causal relationship between the server's I / O interactions with the existing storage nodes is recalculated, and the causal chain topology is dynamically corrected to ensure that the causal mining results are consistent with the real-time resource status.

[0106] Massive causal chains require structured storage to support fast queries. Furthermore, causal relationships have time-series characteristics, making single-store storage insufficient for multi-dimensional analysis. Valid causal chains are converted into causal records according to a standardized structure and stored in a causal relationship database. Each causal record contains metadata such as the causal subject, causal object, causal type, causal strength, and storage timestamp.

[0107] Long-term accumulation of causal records consumes a large amount of storage resources, and similar causal chains are stored repeatedly, necessitating optimization of storage efficiency and refinement of general models. An aging mechanism is used to delete causal records from the causal relationship library that have been unused for an extended period of time, freeing up storage resources. A clustering algorithm is then used to merge causal records based on similarity, optimizing the causal relationship library. For example, similar causal chains related to fully loaded CPUs in different racks and response delays in the same storage array can be combined into a generalized model.

[0108] See also Figure 4 Preferably, the specific steps of generating a resource scheduling strategy include:

[0109] The importance and activity of resources in hyper-converged servers vary significantly. A unified monitoring strategy can lead to insufficient monitoring of critical resources or waste of computing resources for non-critical resources. Based on the historical activity, business criticality, and data interaction frequency of each graph node in the graph database, multi-dimensional feature fusion is used to configure the weight coefficient of the feature dimension, and a weighted summation method is used to generate node weights in the range of 0 to 1, providing a quantitative basis for subsequent hierarchical monitoring strategies. Historical activity is used to characterize the frequency of resource usage of a node within a set time period. It is obtained by statistically analyzing the timestamp sequence of node status change records in the graph database, and the number of active times per unit time is calculated using a sliding window algorithm. Business criticality is used to identify the importance of a graph node in the core business link and is determined by the configured graph node priority label and service impact range. Data interaction frequency is used to quantify the intensity of data exchange between graph nodes and other graph nodes. It is obtained by statistically analyzing the traffic logs and request response records recorded in the standard data set, and the frequency index is generated by analyzing the interaction frequency and data volume of physical connection edges and logical dependency edges.

[0110] Resources with different weights have different impacts on the system and require different monitoring frequencies to balance monitoring accuracy and resource consumption. Configure weight classification thresholds, including upper and lower thresholds, and implement differentiated monitoring based on monitoring weights. Monitoring weights include node weights and edge weights.

[0111] When the monitoring weight is greater than the upper threshold of the weight classification, it means that the graph node or connection edge belongs to a key resource in the core business link, or has recently shown high-frequency resource interaction and high activity. In this case, high-frequency anomaly monitoring is adopted. Based on the preset basic monitoring frequency, the collection cycle is shortened according to the configured shortening ratio, and performance indicators of graph nodes and connection edges are collected at high frequency. Performance indicators include CPU load, memory usage, disk IOPS, bandwidth utilization, and transmission delay. The shortening ratio is used to dynamically adjust the monitoring density of key resources and strike a balance between resource risk and system overhead. By constructing a utility function of monitoring benefits and overhead, with the performance indicator fluctuation range, business impact, and resource importance as independent variables, the optimal shortening ratio is solved by the gradient descent method.

[0112] When the monitoring weight is less than the lower threshold of the weight classification, it means that the graph node or connection edge belongs to a non-core business resource, or has been in a low activity state for a long time and the data interaction frequency is low. In this case, low-frequency anomaly monitoring is adopted. On the basis of the preset basic monitoring frequency, the collection cycle is increased according to the configured growth ratio, and the performance indicators of the graph nodes and connection edges are collected at a low frequency. The growth ratio is used to reduce the monitoring frequency of non-critical resources and reduce resource consumption. By calculating the change rate of performance indicators based on historical data and combining it with the resource importance coefficient, a linear regression model is constructed to predict the growth range of the collection cycle.

[0113] When the monitoring weight is less than or equal to the upper threshold of the weight classification and greater than or equal to the lower threshold of the weight classification, it means that the graph node or connection edge belongs to a regular business resource, and the activity and data interaction frequency are within the normal fluctuation range. In this case, abnormal monitoring is performed according to the basic monitoring frequency, and the performance indicators of the graph nodes and connection edges are collected.

[0114] The normal fluctuation range of resource performance indicators is affected by factors such as business load and time cycle. Static thresholds cannot adapt to dynamic changes and are prone to false positives or omissions. Based on the performance indicator data collected from the graph model, the historical performance indicator data of the graph model is obtained by configuring a time series data extractor. Based on the historical performance indicator data of the graph model, the performance indicator data of the graph nodes and connection edges are statistically analyzed to generate a basic threshold range containing the mean and standard deviation. Based on the frequency of potential anomalies of the graph nodes and connection edges in the graph model, the basic threshold range of the graph nodes and connection edges is adjusted through an adaptive threshold regulator to differentiate the thresholds. The basic threshold range is scaled to construct a dynamic threshold range for the performance indicator data of the graph nodes and connection edges, forming a threshold system that dynamically adapts to the graph structure.

[0115] Resource anomalies may manifest as performance indicators exceeding thresholds or changing dramatically within a short period of time. A single judgment method cannot fully identify anomalies. The performance indicator data collected in real time by the graph model is compared with the corresponding dynamic threshold range. If it is within the corresponding dynamic threshold range, time series mutation monitoring is performed. The performance indicator data is monitored for change amplitude through a differential filter. If the change amplitude of the performance indicator data exceeds the preset amplitude threshold, a resource anomaly warning is triggered.

[0116] Potential anomalies may be accidental fluctuations or real anomalies, and need to be treated differently to avoid invalid analysis. At the same time, clarifying the anomaly propagation path will help locate the root cause and formulate effective strategies. If the performance indicator data collected in real time is outside the corresponding dynamic threshold range, the collection point will be marked as a potential anomaly point, and the potential anomaly point will be identified as an outlier. If the potential anomaly point is determined to be an outlier, it indicates that the anomaly may be caused by temporary noise or accidental fluctuations and has not yet formed a systemic impact. In this case, time series mutation monitoring will be performed to determine whether to trigger a resource anomaly warning. If the potential anomaly point is determined to be a non-outlier, it indicates that the anomaly may be an early sign of a real fault or has formed a lasting impact. In this case, resource scheduling path analysis will be triggered, and the graph node where the potential anomaly point is located will be marked as an abnormal graph node, and the connection edge where the potential anomaly point is located will be marked as an abnormal connection edge.

[0117] The causal relationship library contains a large number of causal chains. It is necessary to filter out those related to the current anomaly, narrow the analysis scope, and improve the efficiency of policy generation. When triggering resource scheduling path analysis, a two-layer screening mechanism is initiated in the causal relationship library based on the timestamp of the potential anomaly point. The valid causal chains within the search time period are extracted through the time window index. The target strength threshold is set based on the causal strength to filter the target causal chains and construct a target causal chain set.

[0118] The selected target causal chains need to be mapped to the actual graph model to clarify the scope of affected resources and provide specific objects for formulating scheduling strategies. The causal subject and causal object of the target causal chain are extracted and mapped to specific graph nodes in the current graph model using the resource ID of the graph database. If the causal subject and causal object do not exist in the graph model or are invalid, the target causal chain is removed. Based on the set of removed target causal chains, the root causal subject is traced and the graph model to be scheduled for resources in the graph database is identified as the target graph model.

[0119] Resource scheduling requires comprehensive consideration of multiple factors, including resource utilization, business continuity, and cost. A single strategy is unlikely to meet complex needs. For the graph nodes of the root causal subject and the graph nodes associated through the target causal chain in the target graph model, a multi-objective optimization function is constructed through a multi-objective optimization engine. Through a multi-dimensional resource load balancing algorithm, combined with the business criticality weights of the graph nodes, the resource unloading path of the root cause node and the associated graph nodes is dynamically planned to achieve a gradient decrease in resource utilization, thereby minimizing the resource utilization of the graph nodes of the root causal subject and the graph nodes associated through the target causal chain; through the business link priority map, the interruption duration and recovery cost of each business link are calculated according to the resource scheduling path, giving priority to ensuring the continuity of high-priority businesses and minimizing the loss of resource scheduling business links to minimize the business interruption loss caused by abnormal propagation; through historical scheduling case similarity matching, combined with the real-time bandwidth of the physical connection edge and the node topological location, the migration path and the execution overhead of expansion deployment are pre-evaluated, and the cost-optimal scheduling scheme is selected to minimize the execution cost of resource scheduling operations; considering resource utilization, business link loss, scheduling cost and other constraints, a candidate strategy set is generated, including resource migration, load balancing, link switching and other operations. For example, if the root cause is storage device overload, the generated strategy may include migrating some IO-intensive services to other storage nodes and adjusting the storage cache strategy to improve performance.

[0120] Before actual implementation, candidate strategies must be evaluated for feasibility and effectiveness to avoid negative impacts on the system. Candidate strategies are verified using a cluster simulation model, evaluating metrics such as anomaly repair time and resource utilization improvement. The optimal candidate strategy is selected based on overall performance. During the simulation, the graph model dynamically updates data, such as newly added node topology and link bandwidth changes, to ensure that the verification results are consistent with the current system state.

[0121] The target candidate strategy is decomposed into an execution instruction chain and sent to the scheduling execution module. At the same time, the causal strength of the target causal chain in the causal relationship library is updated according to the execution results, the causal strength of the effective causal chain is strengthened, and the invalid association is weakened.

[0122] The scheduling execution module is used to send the received execution instruction chain to the target node through the resource orchestration engine, execute resource scheduling, monitor feedback and adjust the resource scheduling, monitor the execution status and effect in real time, and automatically trigger parameter adjustment or policy rollback when the deviation between the actual indicator and the expected indicator exceeds the threshold. It also continues to track and monitor after the adjustment, and feeds back the execution results to the graph database and causal relationship library to form a scheduling closed loop.

[0123] In this embodiment, the scheduling execution module uses the resource orchestration engine to break down the execution instruction chain into a sequence of executable operations, which are then sent to the target node. Simultaneously, probes are deployed at each node to collect resource status in real time and synchronize it with the graph database to update node attributes. When deviations between the actual execution results and expectations are detected, parameter adjustments or policy rollbacks are automatically triggered. After the adjustments are made, continuous tracking and monitoring are maintained, and node dynamic weights and anomaly thresholds are regularly recalculated. This mechanism effectively improves the success rate of scheduling policy execution, shortens anomaly response time, optimizes resource utilization, and continuously updates the causal relationship library through execution results, ensuring intelligent and efficient resource scheduling for hyperconverged server clusters.

[0124] Preferably, the specific steps of executing resource scheduling and performing monitoring, feedback and adjustment on resource scheduling include:

[0125] Target candidate policies are typically abstract logic that needs to be converted into underlying executable operations. The heterogeneous resources of a hyperconverged server must be processed step by step based on type and dependency relationships. The resource orchestration engine parses the target candidate policy's execution instruction chain into a sequence of executable atomic operations. Instruction chains are generated based on operation type and target node. For example, a storage overload policy can be broken down into subtasks such as source node data migration and target node resource pre-allocation, sorted by dependency relationships. A workflow engine maps the execution instruction chain into a call sequence, which is then dispatched to the target node's execution agent via a message queue to ensure operational order and transaction consistency.

[0126] Frequent scheduling of the same node within a short period of time can easily cause resource fluctuations. Nodes of different importance require differentiated monitoring cycles. Execute the call sequence on the target node in sequence and set a scheduling cool-down period for the target node to avoid system fluctuations caused by frequent scheduling of the same node within a short period of time. Configure the scheduling observation period based on the monitoring weight of the target node. During the scheduling observation period for resource scheduling, high-frequency anomaly monitoring is performed on the target node, and performance indicators are collected at high frequency to identify potential anomalies. The frequency of potential anomalies within the scheduling observation period is counted, and a scheduling anomaly threshold is configured. If the frequency of potential anomalies exceeds the scheduling anomaly threshold, a scheduling anomaly warning is triggered, blocking the resource scheduling path, pausing the call sequence execution through the message queue, performing a rollback operation, and terminating the scheduling cool-down period for the target node.

[0127] Eliminate the pseudo causal chain that triggers the scheduling anomaly warning, re-identify the target causal chain of the causal relationship library, update the target candidate strategy, map the call sequence, execute the call sequence on the target node again, and count the frequency of potential anomalies during the scheduling observation period. If the frequency of potential anomalies is less than or equal to the scheduling anomaly threshold, it indicates that the resource scheduling strategy has effectively alleviated the abnormal state of the target node and has not caused new resource fluctuations. Then confirm that the anomaly is resolved, stop resource scheduling observation, and synchronously update the abnormal state label and scheduling timestamp of the node in the graph database; count the number of updates of the target candidate strategy. If it is greater than the preset update threshold, a scheduling update warning is issued to prompt that the current resource scheduling is in a frequent adjustment state. There may be problems such as unreasonable system configuration, causal relationship identification deviation, or the continuous effect of external interference factors. Manual intervention is required to check and optimize the scheduling strategy or adjust system parameters to avoid resource scheduling shocks leading to system performance degradation.

[0128] Through graph database transaction operations, the causal strength of valid causal chains is updated in batches to enhance their effectiveness. The graph nodes and connection edges in the resource scheduling path involved in the call sequence are located, the causal relationship data volume and graph database are updated, and the monitoring weight recalculation is triggered.

[0129] Complete logs are recorded throughout the execution of call sequences, including instruction issuance time, operation sequence, and indicator change curves. Composite indexes are created based on timestamps and graph node IDs. Audit trails are generated for sensitive operations, and logs are stored in a distributed logging system to ensure that the index supports second-level retrieval and meets security audit requirements.

[0130] Example 2:

[0131] See also Figure 5 This embodiment introduces a multi-resource scheduling method for a hyper-converged server, including the following steps:

[0132] Step S1: Collect multi-source datasets through distributed collection nodes, perform semantic alignment and format conversion on the multi-source datasets, and then use a distributed time series database architecture to store them and build a standard dataset with a multi-dimensional index structure.

[0133] Step S2: Locate the target data of the graph model based on the multidimensional index structure of the standard data set, build the graph model, use hierarchical partitioning to build a distributed graph database, and trigger the update of the graph database by capturing change events;

[0134] Step S3: The graph neural network algorithm is used to extract the temporal attributes and spatial topological attributes of the graph nodes. A temporal graph is constructed by combining the sliding window. The structural causal model is used to generate directed causal edges, and valid causal chains are selected and stored in the causal relationship library. When the graph database is updated, the causal relationship incremental update process is started to update the valid causal chains in the causal relationship library by dividing the local graph model.

[0135] Step S4: Based on the monitoring weights of the graph nodes and connecting edges in the graph model, the collection period is adjusted on the basis of the basic monitoring frequency. Differentiated monitoring is implemented through performance indicator data to identify potential anomalies. The target causal set is constructed through the causal relationship library. The multi-objective optimization function is constructed through the multi-objective optimization engine. The target candidate strategy for resource scheduling is generated and sent to the target node for resource scheduling.

[0136] Step S5: Monitor the execution results in real time, perform monitoring feedback adjustments, and update the graph database and causal relationship library.

[0137] Specifically, the steps to identify potential anomalies include:

[0138] Based on the historical performance indicator data of the graph model, statistical analysis is performed on the performance indicator data of the graph nodes and connected edges to generate a basic threshold range including the mean and standard deviation;

[0139] Based on the frequency of potential anomalies in the graph nodes and edges in the graph model, differentiated threshold adjustments are made to the basic threshold ranges of the graph nodes and edges to construct dynamic threshold ranges for the performance indicator data of the graph nodes and edges.

[0140] Compare the performance indicator data collected in real time by the graph model with the corresponding dynamic threshold range. If it is within the corresponding dynamic threshold range, perform time series mutation monitoring. Based on the change range of the performance indicator data, determine whether to trigger a resource anomaly warning;

[0141] If the performance indicator data collected in real time is outside the corresponding dynamic threshold range, the collection point is marked as a potential anomaly point, and outlier identification is performed on the potential anomaly point. If the potential anomaly point is determined to be an outlier, time series mutation monitoring is performed to determine whether a resource anomaly warning is triggered;

[0142] If a potential outlier is determined to be a non-outlier, resource scheduling path analysis is triggered, and the graph node where the potential outlier is located is marked as an abnormal graph node, and the connection edge where the potential outlier is located is marked as an abnormal connection edge.

[0143] Specifically, the steps for generating a target candidate strategy for resource scheduling include:

[0144] When resource scheduling path analysis is triggered, a double-layer screening mechanism is initiated in the causal relationship library based on the timestamp of the potential abnormal point, screening the target causal chain and constructing the target causal chain set;

[0145] Extract the causal subject and causal object of the target causal chain, map them to the graph nodes in the current graph model using the resource ID of the graph database, trace the root causal subject, and identify the graph model to be scheduled in the graph database as the target graph model;

[0146] For the graph nodes of the root causal subject and the graph nodes associated through the target causal chain in the target graph model, a multi-objective optimization function is constructed through the multi-objective optimization engine to generate a set of candidate strategies. The multi-objective optimization function includes minimizing the resource utilization rate of the graph nodes of the root causal subject and the graph nodes associated through the target causal chain, minimizing the loss of the resource scheduling service link, and minimizing the execution cost of the resource scheduling operation.

[0147] The candidate strategies are verified through the cluster simulation model, the target candidate strategy is selected, and it is decomposed into an execution instruction chain and sent to the scheduling execution module. At the same time, the causal strength of the target causal chain in the causal relationship library is updated according to the execution results.

[0148] Specifically, the specific steps of executing resource scheduling and real-time monitoring of execution results for monitoring feedback and adjustment include:

[0149] The resource orchestration engine parses the execution instruction chain of the target candidate policy into an atomic operation sequence. It generates an instruction chain based on the operation type and target node and maps it into a call sequence. The call sequence is then sent to the target node through a message queue.

[0150] Execute the call sequence for the target node in sequence and set the target node's scheduling cool-down period. Configure the scheduling observation period according to the target node's monitoring weight. During the scheduling observation period of executing resource scheduling, use high-frequency anomaly monitoring on the target node to identify potential anomalies.

[0151] Count the frequency of potential anomalies within the scheduling observation period, configure the scheduling anomaly threshold, and determine whether a scheduling anomaly warning is triggered. If so, block the resource scheduling path, pause the call sequence execution through the message queue, perform a rollback operation, and release the scheduling cool-down period of the target node.

[0152] Eliminate the pseudo causal chain that triggers the scheduling exception warning, re-identify the target causal chain in the causal relationship library, update the target candidate strategy, map the call sequence, and execute the call sequence on the target node again;

[0153] If the frequency of potential anomaly points is less than or equal to the scheduling anomaly threshold, the anomaly is confirmed to be resolved, resource scheduling observation is stopped, and the abnormal status label and scheduling timestamp of the node in the graph database are synchronously updated;

[0154] Locate the graph nodes and connection edges in the resource scheduling path involved in the call sequence, update the causal relationship data volume and graph database, and fully record the entire process log of the call sequence execution.

[0155] Working principle and its effect:

[0156] This application uses a distributed time series database with multi-dimensional indexing to standardize and integrate heterogeneous multi-source data and perform elastic storage, solving the problems of data format heterogeneity and semantic ambiguity, forming a unified resource data base, and laying a data foundation for precise scheduling; with the help of a hierarchical and partitioned dynamic graph database, it captures the resource topology evolution and temporal dependencies in real time, transforms physical connections and logical interactions into a computable graph model, and realizes dynamic modeling of resource associations, so that scheduling strategies can accurately match the actual status of resources; uses graph neural networks and structural causal models to deeply mine temporal causal chains, and builds a causal relationship library through dual verification of statistical significance and business rules, accurately locates the root cause of anomalies, and avoids Resource waste caused by misjudgment of scheduling strategies; based on the multi-objective optimization engine combined with cluster simulation, a global optimization strategy that takes into account resource utilization, business continuity and scheduling costs is generated, breaking through the limitations of traditional single-objective scheduling and achieving global optimization of resource scheduling; accurately distinguishing abnormal types through dynamic threshold adjustment and outlier identification mechanism, combining closed-loop regulation to continuously optimize resource allocation, so that the scheduling strategy can evolve dynamically with the business scenario, and ultimately achieve efficient normalization of heterogeneous data at the resource integration level, significantly improving the accuracy of the strategy, global optimization capabilities and adaptive evolution capabilities at the scheduling level, and greatly enhancing the collaborative efficiency and system reliability of multi-resource management of hyper-converged servers.

[0157] The above description is merely a preferred embodiment of the present invention. The scope of protection of the present invention is not limited to the above embodiment. All technical solutions based on the concept of the present invention are within the scope of protection of the present invention. It should be noted that for those skilled in the art, various improvements and modifications that do not depart from the principles of the present invention should also be considered within the scope of protection of the present invention.

Claims

1. Hyper-converged server multi-resource integration system, characterized by: include: Collect multi-source data sets through distributed collection nodes, perform semantic alignment and format conversion on the multi-source data sets, and then use a distributed time series database architecture to store them and build a standard data set with a multi-dimensional index structure; Locating the target data of the graph model based on the multidimensional index structure of the standard data set, building a graph model, building a distributed graph database using hierarchical partitioning, and triggering the update of the graph database by capturing change events; The graph neural network algorithm is used to extract the temporal attributes and spatial topological attributes of the graph nodes, and a sliding window is used to construct a temporal graph. The structural causal model is used to generate directed causal edges, and valid causal chains are screened and stored in the causal relationship library. When the graph database is updated, the causal relationship incremental update process is started, and the valid causal chains in the causal relationship library are updated by dividing the local graph model. Based on the monitoring weights of the graph nodes and connecting edges in the graph model, the collection period is adjusted on the basis of the basic monitoring frequency, differentiated monitoring is implemented through performance indicator data, potential anomalies are identified, a target causal set is constructed through the causal relationship library, a multi-objective optimization function is constructed through the multi-objective optimization engine, and a target candidate strategy for resource scheduling is generated and sent to the target node for resource scheduling. Monitor execution results in real time, conduct feedback adjustments, and update the graph database and causal relationship library.

2. The hyper-converged server multi-resource integration system according to claim 1, wherein: The specific steps of identifying potential abnormal points include: Based on the historical performance indicator data of the graph model, statistical analysis is performed on the performance indicator data of the graph nodes and connected edges to generate a basic threshold range including the mean and standard deviation; Based on the frequency of potential anomalies in the graph nodes and edges in the graph model, differentiated threshold adjustments are made to the basic threshold ranges of the graph nodes and edges to construct dynamic threshold ranges for the performance indicator data of the graph nodes and edges. Compare the performance indicator data collected in real time by the graph model with the corresponding dynamic threshold range. If it is within the corresponding dynamic threshold range, perform time series mutation monitoring. Based on the change range of the performance indicator data, determine whether to trigger a resource anomaly warning; If the performance indicator data collected in real time is outside the corresponding dynamic threshold range, the collection point is marked as a potential anomaly point, and outlier identification is performed on the potential anomaly point. If the potential anomaly point is determined to be an outlier, time series mutation monitoring is performed to determine whether a resource anomaly warning is triggered; If a potential outlier is determined to be a non-outlier, resource scheduling path analysis is triggered, and the graph node where the potential outlier is located is marked as an abnormal graph node, and the connection edge where the potential outlier is located is marked as an abnormal connection edge.

3. The hyper-converged server multi-resource integration system according to claim 2, wherein: The specific steps of generating a target candidate strategy for resource scheduling include: When resource scheduling path analysis is triggered, a double-layer screening mechanism is initiated in the causal relationship library based on the timestamp of the potential abnormal point, screening the target causal chain and constructing the target causal chain set; Extract the causal subject and causal object of the target causal chain, map them to the graph nodes in the current graph model using the resource ID of the graph database, trace the root causal subject, and identify the graph model to be scheduled in the graph database as the target graph model; For the graph nodes of the root causal subject and the graph nodes associated through the target causal chain in the target graph model, a multi-objective optimization function is constructed through a multi-objective optimization engine to generate a set of candidate strategies; the multi-objective optimization function includes minimizing the resource utilization rate of the graph nodes of the root causal subject and the graph nodes associated through the target causal chain, minimizing the resource scheduling service link loss, and minimizing the execution cost of the resource scheduling operation; The candidate strategies are verified through the cluster simulation model, the target candidate strategy is selected, and it is decomposed into an execution instruction chain and sent to the scheduling execution module. At the same time, the causal strength of the target causal chain in the causal relationship library is updated according to the execution results.

4. The hyper-converged server multi-resource integration system according to claim 3, wherein: The specific steps of executing resource scheduling and real-time monitoring of execution results and performing monitoring feedback adjustment include: The resource orchestration engine parses the execution instruction chain of the target candidate policy into an atomic operation sequence. It generates an instruction chain based on the operation type and target node and maps it into a call sequence. The call sequence is then sent to the target node through a message queue. Execute the call sequence for the target node in sequence and set the target node's scheduling cool-down period. Configure the scheduling observation period according to the target node's monitoring weight. During the scheduling observation period of executing resource scheduling, use high-frequency anomaly monitoring on the target node to identify potential anomalies. Count the frequency of potential anomalies within the scheduling observation period, configure the scheduling anomaly threshold, and determine whether a scheduling anomaly warning is triggered. If so, block the resource scheduling path, pause the call sequence execution through the message queue, perform a rollback operation, and release the scheduling cool-down period of the target node. Eliminate the pseudo causal chain that triggers the scheduling exception warning, re-identify the target causal chain in the causal relationship library, update the target candidate strategy, map the call sequence, and execute the call sequence on the target node again; If the frequency of potential anomaly points is less than or equal to the scheduling anomaly threshold, the anomaly is confirmed to be resolved, resource scheduling observation is stopped, and the abnormal status label and scheduling timestamp of the node in the graph database are synchronously updated; Locate the graph nodes and connection edges in the resource scheduling path involved in the call sequence, update the causal relationship data volume and graph database, and record the full process log of the call sequence execution.

5. The hyper-converged server multi-resource integration system according to claim 1, wherein: The specific steps of screening the effective causal chain include: The graph neural network algorithm is used to extract features from the graph models in the distributed graph database. The temporal attributes and spatial topological attributes of the graph nodes are used as input to construct a temporal graph structure of resource state transitions. Quantify the causal influence direction between nodes in the time series graph structure and generate directed causal edges containing time lag relationships; Statistical significance verification is performed on the extracted directed causal edges, and valid causal chains are screened by setting a causal strength threshold; When the graph database triggers an update, the causal relationship incremental update process is automatically started. By dividing the affected local graph model in the graph model, the temporal attributes and spatial topological attributes of the graph nodes in the local graph model are re-extracted, and the directed causal edges are updated through the incremental graph neural network. Convert the effective causal chain into a causal record according to the standardized structure and store it in the causal relationship library. The causal record includes the causal subject, causal object, causal type, causal strength and storage timestamp; The aging mechanism is used to delete causal records in the causal relationship library that have not been used for more than a preset time period, and the clustering algorithm is used to merge the causal records based on similarity to optimize the causal relationship library.

6. The hyper-converged server multi-resource integration system according to claim 1, wherein: The specific steps of obtaining the graph database include: Locate the target data of the graph model based on the multi-dimensional index structure of the standard data set; Graph nodes are constructed based on the resource entities in the target data of the graph model. The acquisition timestamp and spatial topology label are reused to assign time and space attributes to the graph nodes respectively, and real-time performance indicator data is obtained as status attributes. Generate physical connection edges of graph nodes based on spatial topology labels in standard datasets, and construct a time series graph between graph nodes based on the timestamp sequence of business interaction records of graph nodes. By calculating the logical dependency strength between graph nodes, select graph node pairs and generate logical dependency edges. For the connection edges of graph nodes, based on the graph model target data of the standard dataset, the request count and traffic data of the physical connection edges and logical dependency edges are extracted, and edge weights are assigned to the physical connection edges and logical dependency edges respectively according to the weighted priority mapping algorithm; A hierarchical partitioning strategy is adopted to build a distributed graph database storage architecture, in which graph nodes and connection edges in the graph model are stored in two-level shards according to resource type and spatial location, and a multi-dimensional composite index system is established.

7. The hyper-converged server multi-resource integration system according to claim 6, wherein: The specific steps of obtaining the graph database further include: Capture change events of standard datasets in real time. These change events include resource status changes, resource configuration adjustments, or business rule changes. When a change event of a standard dataset is captured, the graph database update process is triggered: Based on the change event type and resource ID, the affected target shard area of ​​the graph model is located through the multi-dimensional composite index system of the distributed graph database. For the located target shard area, update the graph node attributes of the involved graph model, dynamically adjust the connection edges of the graph nodes, and update the edge weights of the connection edges; After the graph database update process is completed, the transaction consistency mechanism of the distributed graph database is used to perform integrity verification on the updated target shard area of ​​the graph model, and the update results are synchronized to the remaining shard areas of the graph database.

8. The hyper-converged server multi-resource integration system according to claim 1, wherein: The specific steps of constructing the standard dataset include: Receive multi-source data sets through distributed message queues, manage sub-queues, initialize distributed time series databases, and establish data docking channels; Build a semantic model based on domain ontology, define standard terms and relationship attributes of resource entities, and form a standardized semantic system; The heterogeneous format terms of the multi-source data set are mapped to a unified semantic space through a rule engine, and a plug-in adaptive conversion architecture is adopted to dynamically load conversion plug-ins to convert the heterogeneous format terms into standardized columnar data; Record the distributed node acquisition timestamps, obtain network topology information, and dynamically generate spatial topology labels for multi-source datasets; LSM tree indexing and columnar storage are used in distributed time series databases. Data from multi-source datasets are classified into hot and cold categories based on their time dimension, access frequency, and business priority to adjust shard storage and the number of replicas. Integrate spatial topology labels and resource type identifiers to build a multidimensional index structure that includes time, space, and resource types, and build a metadata index; desensitize and encrypt sensitive data for storage, and record data operation behaviors.

9. The hyper-converged server multi-resource integration system according to claim 1, wherein: The specific steps of adjusting the collection period based on the basic monitoring frequency and implementing differentiated monitoring through performance indicator data include: Based on the historical activity, business criticality, and data interaction frequency of each graph node in the graph database, multi-dimensional feature fusion is used to configure the weight coefficients of the feature dimensions, and a weighted summation method is used to generate node weights in the range of 0 to 1. Configure weight classification thresholds, including upper and lower thresholds, and implement differentiated monitoring based on monitoring weights, including node weights and edge weights. When the monitoring weight is greater than the upper threshold of the weight classification, high-frequency abnormal monitoring is adopted. On the basis of the preset basic monitoring frequency, the collection cycle is shortened according to the configured shortening ratio; When the monitoring weight is less than the lower threshold of the weight classification, low-frequency abnormal monitoring is adopted. On the basis of the preset basic monitoring frequency, the collection cycle is increased according to the configured growth ratio; When the monitoring weight is less than or equal to the upper threshold of the weight classification, and greater than or equal to the lower threshold of the weight classification, the collection cycle is set according to the basic monitoring frequency to perform abnormal monitoring; Differentiated monitoring of graph nodes or connection edges is performed according to the dynamic collection cycle, and performance indicator data of graph nodes and connection edges is collected.

10. A method for scheduling multiple resources on a hyper-converged server, which is implemented based on the hyper-converged server multiple resource integration system according to any one of claims 1 to 9, characterized in that: The following steps are involved: Step S1: Collect multi-source datasets through distributed collection nodes, perform semantic alignment and format conversion on the multi-source datasets, and then use a distributed time series database architecture to store them and construct a standard dataset with a multi-dimensional index structure; Step S2: Locate the graph model target data based on the multidimensional index structure of the standard data set, build a graph model, construct a distributed graph database using hierarchical partitioning, and trigger the update of the graph database by capturing change events; Step S3: Extract the temporal attributes and spatial topological attributes of the graph nodes through the graph neural network algorithm, build a temporal graph in combination with the sliding window, generate directed causal edges using the structural causal model, and select valid causal chains to store in the causal relationship library; when the graph database is updated, start the causal relationship incremental update process, and update the valid causal chains in the causal relationship library by dividing the local graph model; Step S4: Based on the monitoring weights of the graph nodes and connecting edges in the graph model, the collection period is adjusted on the basis of the basic monitoring frequency, differentiated monitoring is implemented through performance indicator data, potential anomalies are identified, a target causal set is constructed through the causal relationship library, a multi-objective optimization function is constructed through the multi-objective optimization engine, a target candidate strategy for resource scheduling is generated, and the strategy is sent to the target node for resource scheduling. Step S5: Monitor the execution results in real time, perform monitoring feedback adjustments, and update the graph database and causal relationship library.

Images