Large-scale cloud data center-oriented server collaborative monitoring method and system

Abstract

The invention provides a server collaborative monitoring method and system for a large-scale cloud data center. Through a large-scale intelligent computing cloud data platform, an AI server and an intelligent computing cloud computing power server cluster are constructed through algorithm integration according to data types, and an intelligent computing cloud server cross monitoring network is formed; the intelligent computing cloud computing power server performs server connection state detection and large-scale server cooperative detection according to a keep-alive detection mechanism, and detects a large-scale server cluster state; performing large-scale server cluster state classification on large-scale server cluster states through a cloud data center, judging various cluster state data resource allocation, and performing large-scale server cluster state data resource collaborative configuration; cooperative monitoring is performed on a large-scale server cluster through an intelligent computing cloud data platform and an intelligent computing cloud computing power server, and real-time refined monitoring is performed on large model training data resources of a large-scale cloud data center.

Description

Technical Field

[0001] The present invention relates to the field of intelligent cloud server cluster monitoring technology, and more specifically, to a server collaborative monitoring method and system for a large-scale cloud data center. Background Art

[0002] At present, with the rapid development of large models, how to monitor the computing power cluster of intelligent server in large-scale intelligent cloud data through cloud service monitoring network, how to build a detection mechanism for connection status detection and collaborative detection of cluster status, how to determine the status of large-scale server clusters and allocate various cluster status data resources, how to collaboratively monitor large-scale server clusters and monitor large model training data resources in large-scale cloud data centers, etc., have yet to be solved; therefore, it is necessary to propose a server collaborative monitoring method and system for large-scale cloud data centers to at least partially solve the problems existing in the prior art. Summary of the invention

[0003] A series of simplified concepts are introduced in the content of the invention, which will be further described in detail in the specific implementation method section; the content of the invention of the present invention does not mean to attempt to limit the key features and essential technical features of the technical solution claimed for protection, nor does it mean to attempt to determine the scope of protection of the technical solution claimed for protection.

[0004] To at least partially solve the above problems, the present invention provides a server collaborative monitoring method for a large-scale cloud data center, comprising:

[0005] S100, through the large-scale intelligent cloud data platform, builds AI servers and intelligent cloud computing server clusters through algorithm integration according to data types, forming an intelligent cloud server cross-monitoring network;

[0006] S200, the intelligent cloud computing server performs server connection status detection and large-scale server collaboration detection according to the keep-alive detection mechanism, and detects the status of large-scale server clusters;

[0007] S300, classifying the status of a large-scale server cluster through a cloud data center, determining the allocation of data resources for various types of cluster status, and performing collaborative configuration of data resources for the status of a large-scale server cluster;

[0008] S400, through the intelligent cloud data platform and intelligent cloud computing power server, coordinates the monitoring of large-scale server clusters, and conducts real-time and refined monitoring of large-model training data resources in large-scale cloud data centers.

[0009] Preferably, S100 includes:

[0010] S101, through the large-scale intelligent cloud data platform, builds AI servers through algorithm integration according to data types;

[0011] S102, deploy and schedule AI servers and general servers to build a smart cloud computing server cluster;

[0012] S103, construct a cross-monitoring network topology structure based on the intelligent cloud computing power server cluster to form an intelligent cloud server cross-monitoring network.

[0013] Preferably, S200 includes:

[0014] S201, the Zhisuan cloud computing server performs server connection status detection according to the keep-alive detection mechanism;

[0015] S202: Perform large-scale server collaborative detection based on server connection status detection to detect large-scale server cluster status.

[0016] Preferably, S300 includes:

[0017] S301, classifying the status of a large-scale server cluster through a cloud data center;

[0018] S302, according to the large-scale server cluster status classification, determine the allocation of various types of cluster status data resources, and perform large-scale server cluster status data resource collaborative configuration;

[0019] According to the classification of large-scale server cluster status, the allocation of various types of cluster status data resources is determined, and cluster status data resource configuration is performed, including: Zhisuan Cloud Computing Server classifies the status of large-scale server clusters, determines the allocation of various types of cluster status data resources, performs large-scale server data resource collaboration, collaboratively processes the status of similar clusters, and collaboratively configures cluster status data resources.

[0020] Preferably, S400 includes:

[0021] S401, through the Zhisuan Cloud data platform and Zhisuan Cloud computing power server, large-scale server cluster status data resource collaborative monitoring of large-scale server clusters;

[0022] S402: Based on the collaborative monitoring of the Zhisuan Cloud data platform and the Zhisuan Cloud computing power server, real-time and refined monitoring of large-model training data resources in large-scale cloud data centers is performed.

[0023] The present invention provides a server collaborative monitoring system for a large-scale cloud data center, comprising:

[0024] The cross-monitoring network of intelligent cloud servers is formed through the large-scale intelligent cloud data platform, which builds AI servers and intelligent cloud computing server clusters through algorithm integration according to data types, forming a cross-monitoring network of intelligent cloud servers;

[0025] The computing power mechanism collaborative detection subsystem, the Zhisuan cloud computing power server performs server connection status detection and large-scale server collaborative detection according to the keep-alive detection mechanism, and detects the status of large-scale server clusters;

[0026] The cluster classification and collaborative configuration subsystem classifies the status of large-scale server clusters through the cloud data center, determines the allocation of data resources for various types of cluster status, and performs collaborative configuration of data resources for large-scale server cluster status;

[0027] The intelligent cloud server collaborative monitoring subsystem conducts collaborative monitoring of large-scale server clusters through the intelligent cloud data platform and intelligent cloud computing power servers, and conducts real-time and refined monitoring of large-model training data resources in large-scale cloud data centers.

[0028] Preferably, the intelligent cloud server cross-monitoring network includes:

[0029] Algorithm integration AI server, through the large-scale intelligent cloud data platform, build AI server through algorithm integration according to data type;

[0030] Intelligent computing cloud computing service cluster, deploys and schedules AI servers and general servers to build an intelligent computing cloud computing server cluster;

[0031] The cross-monitoring network builds a cross-monitoring network topology based on the intelligent cloud computing server cluster to form an intelligent cloud server cross-monitoring network.

[0032] Preferably, the computing power mechanism cooperates with the detection subsystem, including:

[0033] The detection mechanism builds a subsystem. The Zhisuan Cloud computing server detects the server connection status based on the keep-alive detection mechanism.

[0034] The server cluster status detection subsystem performs large-scale server collaborative detection based on server connection status detection to detect the status of large-scale server clusters;

[0035] Preferably, the cluster classification collaborative configuration subsystem includes:

[0036] The cluster status classification subsystem classifies the status of large-scale server clusters through cloud data centers;

[0037] The data resource coordination subsystem determines the allocation of data resources for various cluster states according to the classification of large-scale server cluster states, and coordinates the configuration of data resources for large-scale server cluster states;

[0038] According to the classification of large-scale server cluster status, the allocation of various types of cluster status data resources is determined, and cluster status data resource configuration is performed, including: Zhisuan Cloud Computing Server classifies the status of large-scale server clusters, determines the allocation of various types of cluster status data resources, performs large-scale server data resource collaboration, collaboratively processes the status of similar clusters, and collaboratively configures cluster status data resources.

[0039] Preferably, the intelligent cloud server collaborative monitoring subsystem includes:

[0040] The server cluster resource collaborative monitoring subsystem uses the Zhisuan Cloud data platform and Zhisuan Cloud computing power servers to conduct collaborative monitoring of large-scale server cluster status data resources.

[0041] The collaborative monitoring subsystem for large model training resources conducts real-time and refined monitoring of large model training data resources in large-scale cloud data centers based on the collaborative monitoring of the Zhisuan Cloud data platform and the Zhisuan Cloud computing power server.

[0042] The beneficial effects of the above technical solution include:

[0043] The present invention provides a server collaborative monitoring method and system for a large-scale cloud data center. Through a large-scale intelligent computing cloud data platform, AI servers and intelligent computing cloud computing power server clusters are constructed through algorithm integration according to data types to form an intelligent computing cloud server cross-monitoring network; the intelligent computing cloud computing power server performs server connection status detection and large-scale server collaborative detection according to a keep-alive detection mechanism to detect the status of a large-scale server cluster; the large-scale server cluster status is classified through a cloud data center to determine the allocation of various cluster status data resources, and the large-scale server cluster status data resources are collaboratively configured; the large-scale server cluster is collaboratively monitored through the intelligent computing cloud data platform and the intelligent computing cloud computing power server, and the large-scale model training data resources of the large-scale cloud data center are monitored in a real-time and refined manner; it can adapt to large-scale data processing monitoring of large models; it can monitor the large-scale intelligent computing cloud data intelligent server computing power cluster through a cloud service monitoring network; it can build a detection mechanism for connection status detection and collaborative detection of cluster status; it can perform large-scale server cluster status determination and various cluster status data resource allocation; it can collaboratively monitor large-scale server clusters and monitor large-model training data resources of large-scale cloud data centers, significantly improving the server collaborative monitoring efficiency and data resource allocation efficiency of large-scale cloud data centers.

[0044] The present invention provides a server collaborative monitoring method and system for a large-scale cloud data center. Other advantages, objectives and features of the present invention will be partially reflected through the following description, and will also be partially understood by technical personnel in the field through research and practice of the present invention. BRIEF DESCRIPTION OF THE DRAWINGS

[0045] The accompanying drawings are used to provide a further understanding of the present invention and constitute a part of the specification. Together with the embodiments of the present invention, they are used to explain the present invention and do not constitute a limitation of the present invention. In the accompanying drawings:

[0046] Figure 1 This is a diagram of an embodiment of a server collaborative monitoring system for a large-scale cloud data center according to the present invention.

[0047] Figure 2 A framework implementation diagram of a server collaborative monitoring method and system for a large-scale cloud data center according to the present invention.

[0048] Figure 3 This is another application embodiment diagram of a server collaborative monitoring method and system for a large-scale cloud data center according to the present invention. DETAILED DESCRIPTION

[0049] The present invention is further described in detail below in conjunction with the accompanying drawings and embodiments so that those skilled in the art can implement it according to the instructions; as shown in the figure, the present invention provides a server collaborative monitoring method for a large-scale cloud data center, including:

[0050] S100, through the large-scale intelligent cloud data platform, builds AI servers and intelligent cloud computing server clusters through algorithm integration according to data types, forming an intelligent cloud server cross-monitoring network;

[0051] S200, the intelligent cloud computing server performs server connection status detection and large-scale server collaboration detection according to the keep-alive detection mechanism, and detects the status of large-scale server clusters;

[0052] S300, classifying the status of a large-scale server cluster through a cloud data center, determining the allocation of data resources for various types of cluster status, and performing collaborative configuration of data resources for the status of a large-scale server cluster;

[0053] S400, through the intelligent cloud data platform and intelligent cloud computing power server, coordinates the monitoring of large-scale server clusters, and conducts real-time and refined monitoring of large-model training data resources in large-scale cloud data centers.

[0054] The principle and effect of the above technical solution are as follows: The present invention provides a server collaborative monitoring method for large-scale cloud data centers, including: through a large-scale intelligent cloud data platform, AI servers and intelligent cloud computing server clusters are constructed through algorithm integration according to data types to form an intelligent cloud server cross-monitoring network; the intelligent cloud computing server performs server connection status detection and large-scale server collaborative detection according to the keep-alive detection mechanism to detect the status of large-scale server clusters; the cloud data center classifies the status of large-scale server clusters, determines the allocation of data resources for various types of cluster status, and performs large-scale server cluster status data resource collaborative configuration; through the intelligent cloud data center, The platform and the intelligent cloud computing power server can collaboratively monitor large-scale server clusters, and conduct real-time and refined monitoring of large-model training data resources in large-scale cloud data centers; it can adapt to large-scale data processing monitoring of large models; it can monitor large-scale intelligent cloud data intelligent server computing power clusters through the cloud service monitoring network; it can build a detection mechanism to detect connection status and collaboratively detect cluster status; it can determine the status of large-scale server clusters and allocate various cluster status data resources; it can collaboratively monitor large-scale server clusters and monitor large-model training data resources in large-scale cloud data centers, significantly improving the collaborative monitoring efficiency of servers in large-scale cloud data centers and the efficiency of data resource allocation.

[0055] In one embodiment, S100 includes:

[0056] S101, through the large-scale intelligent cloud data platform, builds AI servers through algorithm integration according to data types;

[0057] S102, deploy and schedule AI servers and general servers to build a smart cloud computing server cluster;

[0058] S103, construct a cross-monitoring network topology structure based on the intelligent cloud computing power server cluster to form an intelligent cloud server cross-monitoring network.

[0059] The principle and effect of the above technical solution are: through a large-scale intelligent computing cloud data platform, AI servers are built through algorithm integration according to data types; AI servers and general servers are deployed and scheduled to build an intelligent computing cloud computing power server cluster; based on the intelligent computing cloud computing power server cluster, a cross-monitoring network topology is built to form an intelligent computing cloud server cross-monitoring network; based on the intelligent computing cloud computing power server, a cross-monitoring network topology is built, including: based on the intelligent computing cloud computing power server, the AI ​​servers and general servers in the intelligent computing cloud computing power server form a cross-monitoring network, and a cross-monitoring network topology is built, and multiple AI servers and multiple general servers form cross-monitoring network nodes; multiple cross-monitoring network nodes correspond to multiple virtual three-dimensional monitoring data arrays of the large-scale intelligent computing cloud data platform respectively; the virtual three-dimensional monitoring data array corresponds to the AI ​​server monitoring data and the general server monitoring data; the dimension of the virtual three-dimensional monitoring data array corresponds to the server monitoring network level; the monitoring data is aggregated to the intelligent computing cloud computing power server of the cross-monitoring network node, and transmitted to the large-scale intelligent computing cloud data platform through a secure channel to form a cross-monitoring network of intelligent computing cloud servers.

[0060] In one embodiment, S200 includes:

[0061] S201, the Zhisuan cloud computing server performs server connection status detection according to the keep-alive detection mechanism;

[0062] S202: Perform large-scale server collaborative detection based on server connection status detection to detect large-scale server cluster status.

[0063] The principle and effect of the above technical solution are as follows: the intelligent cloud computing server performs server connection status detection according to the keep-alive detection mechanism; performs large-scale server collaborative detection according to the server connection status detection, and detects the status of large-scale server clusters; performs large-scale server collaborative detection according to the server connection status detection, and the detection of large-scale server cluster status includes: obtaining multiple server cluster connection states according to keep-alive timed long connection and keep-alive short connection detection; classifying multiple server cluster connection states to obtain classified server cluster connection states; respectively, through network hardware probes and network software probes, perform large-scale batch comparative analysis of similar behavior patterns of similar server cluster connection states in the classified server cluster connection states, replacing the single-node probe behavior pattern comparative analysis, the network hardware probe and the network software probe collaboratively analyze the source address of the detection data packet, the destination address of the detection data packet and the protocol type of the detection data packet, and if any of the network hardware probes or network software probes or any type of server cluster analyzes an abnormal behavior pattern, the detection information of the same connection state behavior pattern characteristics or a type of server cluster is collaboratively shared, and subsequent similar server clusters do not need to repeat a large number of single-node analyses, and large-scale server collaborative detection is performed to detect the status of large-scale server clusters.

[0064] In one embodiment, S300 includes:

[0065] S301, classifying the status of a large-scale server cluster through a cloud data center;

[0066] S302, according to the large-scale server cluster status classification, determine the allocation of various types of cluster status data resources, and perform large-scale server cluster status data resource collaborative configuration;

[0067] According to the classification of large-scale server cluster status, the allocation of various types of cluster status data resources is determined, and cluster status data resource configuration is performed, including: Zhisuan Cloud Computing Server classifies the status of large-scale server clusters, determines the allocation of various types of cluster status data resources, performs large-scale server data resource collaboration, collaboratively processes the status of similar clusters, and collaboratively configures cluster status data resources.

[0068] The principle and effect of the above technical solution are: to classify the status of large-scale server clusters through the cloud data center; to determine the allocation of various types of cluster status data resources based on the classification of large-scale server cluster status, and to coordinate the large-scale server cluster status data resources; to determine the allocation of various types of cluster status data resources based on the classification of large-scale server cluster status, and to configure cluster status data resources, including: the intelligent cloud computing server determines the allocation of various types of cluster status data resources based on the classification of large-scale server cluster status, coordinates large-scale server data resources, coordinates the status of similar clusters, and coordinates the configuration of cluster status data resources.

[0069] In one embodiment, S400 includes:

[0070] S401, through the Zhisuan Cloud data platform and Zhisuan Cloud computing power server, large-scale server cluster status data resource collaborative monitoring of large-scale server clusters;

[0071] S402: Based on the collaborative monitoring of the Zhisuan Cloud data platform and the Zhisuan Cloud computing power server, real-time and refined monitoring of large-model training data resources in large-scale cloud data centers is performed.

[0072] The principles and effects of the above technical solutions are as follows: Through the intelligent cloud data platform and the intelligent cloud computing power server, large-scale server clusters are collaboratively monitored for large-scale server cluster status data resources; based on the collaborative monitoring of the intelligent cloud data platform and the intelligent cloud computing power server, the large-scale cloud data center's large-model training data resources are monitored in real time and in detail; based on the collaborative monitoring of the intelligent cloud data platform and the intelligent cloud computing power server, the large-scale cloud data center's large-model training data resources are monitored in real time and in detail, including: based on the collaborative monitoring of the intelligent cloud data platform and the intelligent cloud computing power server, the large-model training process data resources, the large-model training environment data resources, the large-model training GPU performance and bandwidth consumption indicators are monitored, and ... Based on the GPU performance and bandwidth consumption indicators, the cluster status data resources are coordinated and configured, and the large-scale cloud data center's large-scale model training data resources are monitored in real time and in a refined manner. According to the coordinated monitoring of the Zhisuan Cloud Data Platform and the Zhisuan Cloud Computing Server, the coordinated log information of the Zhisuan Cloud Data Platform and the Zhisuan Cloud Computing Server is collected through the network monitoring log probe, and the data traffic is obtained through the mirror data stream. The network monitoring log probe side pre-processing is performed and sent to the situation awareness center of the Zhisuan Cloud Data Platform and the Zhisuan Cloud Computing Server for large-model analysis association, log generation source tracking and log content analysis, coordinated monitoring of network traffic abnormal events or virus attack events, and coordinated monitoring of abnormal recovery or virus isolation and blocking; collaborative monitoring accurately traces the source and generates detailed monitoring reports, and real-time and refined monitoring of large model training data resources is performed.

[0073] The present invention provides a server collaborative monitoring system for a large-scale cloud data center, comprising:

[0074] The cross-monitoring network of intelligent cloud servers is formed by integrating AI servers and intelligent cloud computing servers based on data types through the large-scale intelligent cloud data platform, thus forming a cross-monitoring network of intelligent cloud servers.

[0075] The computing power mechanism collaborative detection subsystem, the Zhisuan cloud computing power server performs server connection status detection and large-scale server collaborative detection based on the keep-alive detection mechanism, and detects the status of large-scale server clusters.

[0076] The cluster classification and collaborative configuration subsystem classifies the status of large-scale server clusters through the cloud data center, determines the allocation of data resources for various types of cluster status, and performs collaborative configuration of data resources for large-scale server cluster status;

[0077] The intelligent cloud server collaborative monitoring subsystem conducts collaborative monitoring of large-scale server clusters through the intelligent cloud data platform and intelligent cloud computing power servers, and conducts real-time and refined monitoring of large-model training data resources in large-scale cloud data centers.

[0078] The principles and effects of the above technical solution are as follows: The present invention provides a server collaborative monitoring system for large-scale cloud data centers, including: a cross-monitoring network of intelligent cloud servers, which uses a large-scale intelligent cloud data platform to build AI servers and intelligent cloud computing power server clusters through algorithm integration according to data types, forming a cross-monitoring network of intelligent cloud servers; a computing power mechanism collaborative detection subsystem, in which the intelligent cloud computing power server performs server connection status detection and large-scale server collaborative detection according to the keep-alive detection mechanism, and detects the status of large-scale server clusters; a cluster classification collaborative configuration subsystem, which classifies the status of large-scale server clusters through the cloud data center, determines the allocation of data resources for various types of cluster status, and performs large-scale server cluster status data resource allocation. Collaborative configuration; the intelligent cloud server collaborative monitoring subsystem can collaboratively monitor large-scale server clusters through the intelligent cloud data platform and the intelligent cloud computing power server, and conduct real-time and refined monitoring of large-model training data resources in large-scale cloud data centers; it can adapt to large-scale data processing monitoring of large models; it can monitor large-scale intelligent cloud data server computing power clusters through the cloud service monitoring network; it can build a detection mechanism for connection status detection and collaborative detection of cluster status; it can determine the status of large-scale server clusters and allocate various cluster status data resources; it can collaboratively monitor large-scale server clusters and monitor large-model training data resources in large-scale cloud data centers, significantly improving the collaborative monitoring efficiency of servers and data resource allocation efficiency in large-scale cloud data centers.

[0079] In one embodiment, the intelligent computing cloud server cross-monitors the network, including:

[0080] Algorithm integration AI server, through the large-scale intelligent cloud data platform, build AI server through algorithm integration according to data type;

[0081] Intelligent computing cloud computing service cluster, deploys and schedules AI servers and general servers to build an intelligent computing cloud computing server cluster;

[0082] The cross-monitoring network builds a cross-monitoring network topology based on the intelligent cloud computing server cluster to form an intelligent cloud server cross-monitoring network.

[0083] The principle and effect of the above technical solution are: the cross-monitoring network of intelligent cloud servers includes: algorithm integrated AI servers, which build AI servers through algorithm integration according to data types through a large-scale intelligent cloud data platform; intelligent cloud computing service clusters, which deploy and schedule AI servers and general servers to build intelligent cloud computing server clusters; cross-monitoring network, which builds a cross-monitoring network topology according to the intelligent cloud computing server cluster to form an intelligent cloud server cross-monitoring network; according to the intelligent cloud computing server, building a cross-monitoring network topology includes: according to the intelligent cloud computing server, the AI ​​servers and general servers in the intelligent cloud computing server form a cross-monitoring network, and build a cross-monitoring network topology, and multiple AI servers and multiple general servers form cross-monitoring network nodes; multiple cross-monitoring network nodes correspond to multiple virtual three-dimensional monitoring data arrays of the large-scale intelligent cloud data platform respectively; the virtual three-dimensional monitoring data array corresponds to the AI ​​server monitoring data and the general server monitoring data; the dimension of the virtual three-dimensional monitoring data array corresponds to the server monitoring network level; the monitoring data is aggregated to the intelligent cloud computing server of the cross-monitoring network node, and transmitted to the large-scale intelligent cloud data platform through a secure channel to form a cross-monitoring network of intelligent cloud servers.

[0084] In one embodiment, the computing power mechanism cooperates with the detection subsystem, including:

[0085] The detection mechanism builds a subsystem. The Zhisuan Cloud computing server detects the server connection status based on the keep-alive detection mechanism.

[0086] The server cluster status detection subsystem performs large-scale server collaborative detection based on server connection status detection to detect the status of large-scale server clusters.

[0087] The principle and effect of the above technical solution are as follows: the computing power mechanism collaborative detection subsystem includes: the detection mechanism construction subsystem, the intelligent cloud computing power server performs server connection status detection according to the keep-alive detection mechanism; the server cluster status detection subsystem performs large-scale server collaborative detection according to the server connection status detection, and detects the status of large-scale server clusters; performs large-scale server collaborative detection according to the server connection status detection, and the detection of large-scale server cluster status includes: obtaining multiple server cluster connection status according to the keep-alive timed long connection and keep-alive short connection detection; classifying multiple server cluster connection status to obtain classified server cluster connection status; respectively through network hardware probes and network software probes, to conduct large-scale batch comparative analysis of similar behavior patterns of similar server cluster connection states in classified server cluster connection states, replacing the comparative analysis of single-node probe behavior patterns. Network hardware probes and network software probes collaborate to analyze the source address of detection packets, the destination address of detection packets and the protocol type of detection packets. If any of the network hardware probes or network software probes or any type of server cluster analyzes an abnormal behavior pattern, the detection information of the same connection state behavior pattern characteristics or a type of server cluster will be collaboratively shared. Subsequently, similar server clusters will no longer need to repeat a large number of single-node analyses, and large-scale server collaborative detection will be performed to detect the status of large-scale server clusters.

[0088] In one embodiment, the cluster classification collaborative configuration subsystem includes:

[0089] The cluster status classification subsystem classifies the status of large-scale server clusters through cloud data centers;

[0090] The data resource coordination subsystem determines the allocation of data resources for various cluster states according to the classification of large-scale server cluster states, and coordinates the configuration of data resources for large-scale server cluster states;

[0091] According to the classification of large-scale server cluster status, the allocation of various types of cluster status data resources is determined, and cluster status data resource configuration is performed, including: Zhisuan Cloud Computing Server classifies the status of large-scale server clusters, determines the allocation of various types of cluster status data resources, performs large-scale server data resource collaboration, collaboratively processes the status of similar clusters, and collaboratively configures cluster status data resources.

[0092] The principles and effects of the above technical solution are: cluster classification and collaborative configuration subsystem, including: cluster status classification subsystem, which classifies large-scale server cluster status through cloud data center; data resource collaborative subsystem, which determines the allocation of various types of cluster status data resources according to large-scale server cluster status classification, and performs large-scale server cluster status data resource collaborative configuration; determines the allocation of various types of cluster status data resources according to large-scale server cluster status classification, and performs cluster status data resource configuration, including: intelligent cloud computing power server determines the allocation of various types of cluster status data resources according to large-scale server cluster status classification, performs large-scale server data resource collaboration, collaboratively processes similar cluster status, and collaboratively configures cluster status data resources.

[0093] In one embodiment, the intelligent computing cloud server cooperates with the monitoring subsystem, including:

[0094] The server cluster resource collaborative monitoring subsystem uses the Zhisuan Cloud data platform and Zhisuan Cloud computing power servers to conduct collaborative monitoring of large-scale server cluster status data resources.

[0095] The collaborative monitoring subsystem for large model training resources conducts real-time and refined monitoring of large model training data resources in large-scale cloud data centers based on the collaborative monitoring of the Zhisuan Cloud data platform and the Zhisuan Cloud computing power server.

[0096] The principles and effects of the above technical solutions are as follows: the intelligent cloud server collaborative monitoring subsystem includes: the server cluster resource collaborative monitoring subsystem, which performs large-scale server cluster status data resource collaborative monitoring of large-scale server clusters through the intelligent cloud data platform and the intelligent cloud computing power server; the collaborative monitoring of large model training resource subsystem, based on the collaborative monitoring of the intelligent cloud data platform and the intelligent cloud computing power server, performs real-time and refined monitoring of large model training data resources in large-scale cloud data centers; based on the collaborative monitoring of the intelligent cloud data platform and the intelligent cloud computing power server, performs real-time and refined monitoring of large model training data resources in large-scale cloud data centers, including: monitoring large model training process data resources, large model training environment data resources, large model training GPU performance and bandwidth consumption indicators according to the collaborative monitoring of the intelligent cloud data platform and the intelligent cloud computing power server, and The data resources of the program, the data resources of the large model training environment, the GPU performance and bandwidth consumption indicators of the large model training are coordinated and configured in the cluster status data resources, and the large model training data resources of the large-scale cloud data center are monitored in real time and finely; according to the coordinated monitoring of the Zhisuan Cloud Data Platform and the Zhisuan Cloud Computing Server, the coordinated log information of the Zhisuan Cloud Data Platform and the Zhisuan Cloud Computing Server is collected through the network monitoring log probe, and the data traffic is obtained through the mirror data stream. The network monitoring log probe side pre-processing is performed and sent to the situation awareness center of the Zhisuan Cloud Data Platform and the Zhisuan Cloud Computing Server for large model analysis association, log generation source tracking and log content analysis, coordinated monitoring of network traffic abnormal events or virus attack events, and coordinated monitoring abnormal recovery or virus isolation and blocking; collaborative monitoring accurately traces the source and generates detailed monitoring reports, and real-time and fine-grained monitoring of large model training data resources is performed.

[0097] Although the embodiments of the present invention have been disclosed as above, they are not limited to the applications listed in the specification and the implementation methods. They can be fully applied to various fields suitable for the present invention. For those familiar with the art, additional modifications can be easily implemented. Therefore, without departing from the general concept defined by the claims and the scope of equivalents, the present invention is not limited to the specific details and the illustrations shown and described herein.

Claims

1. A server collaborative monitoring method for a large-scale cloud data center, characterized in that: include: S100, through the large-scale intelligent cloud data platform, builds AI servers and intelligent cloud computing server clusters through algorithm integration according to data types, forming an intelligent cloud server cross-monitoring network; S200, the intelligent cloud computing server performs server connection status detection and large-scale server collaboration detection according to the keep-alive detection mechanism, and detects the status of large-scale server clusters; S300, classifying the status of a large-scale server cluster through a cloud data center, determining the allocation of data resources for various types of cluster status, and performing collaborative configuration of data resources for the status of a large-scale server cluster; S400, through the intelligent cloud data platform and intelligent cloud computing power server, coordinates the monitoring of large-scale server clusters, and conducts real-time and refined monitoring of large-model training data resources in large-scale cloud data centers.

2. A server collaborative monitoring method for a large-scale cloud data center according to claim 1, characterized in that: S100 includes: S101, through the large-scale intelligent cloud data platform, builds AI servers through algorithm integration according to data types; S102, deploy and schedule AI servers and general servers to build a smart cloud computing server cluster; S103, construct a cross-monitoring network topology structure based on the intelligent cloud computing power server cluster to form an intelligent cloud server cross-monitoring network.

3. The method for collaborative monitoring of servers in a large-scale cloud data center according to claim 1, characterized in that: S200 includes: S201, the Zhisuan cloud computing server performs server connection status detection according to the keep-alive detection mechanism; S202: Perform large-scale server collaborative detection based on server connection status detection to detect large-scale server cluster status.

4. The method for collaborative monitoring of servers in a large-scale cloud data center according to claim 1, characterized in that: S300 includes: S301, classifying the status of a large-scale server cluster through a cloud data center; S302, according to the large-scale server cluster status classification, determine the allocation of various types of cluster status data resources, and perform large-scale server cluster status data resource collaborative configuration; According to the classification of large-scale server cluster status, the allocation of various types of cluster status data resources is determined, and cluster status data resource configuration is performed, including: Zhisuan Cloud Computing Server classifies the status of large-scale server clusters, determines the allocation of various types of cluster status data resources, performs large-scale server data resource collaboration, collaboratively processes the status of similar clusters, and collaboratively configures cluster status data resources.

5. The method for collaborative monitoring of servers in a large-scale cloud data center according to claim 1, characterized in that: S400 includes: S401, through the Zhisuan Cloud data platform and Zhisuan Cloud computing power server, large-scale server cluster status data resource collaborative monitoring of large-scale server clusters; S402: Based on the collaborative monitoring of the Zhisuan Cloud data platform and the Zhisuan Cloud computing power server, real-time and refined monitoring of large-model training data resources in large-scale cloud data centers is performed.

6. A server collaborative monitoring system for large-scale cloud data centers, characterized in that: include: The cross-monitoring network of intelligent cloud servers is formed through the large-scale intelligent cloud data platform, which builds AI servers and intelligent cloud computing server clusters through algorithm integration according to data types, forming a cross-monitoring network of intelligent cloud servers; The computing power mechanism collaborative detection subsystem, the Zhisuan cloud computing power server performs server connection status detection and large-scale server collaborative detection according to the keep-alive detection mechanism, and detects the status of large-scale server clusters; The cluster classification and collaborative configuration subsystem classifies the status of large-scale server clusters through the cloud data center, determines the allocation of data resources for various types of cluster status, and performs collaborative configuration of data resources for large-scale server cluster status; The intelligent cloud server collaborative monitoring subsystem conducts collaborative monitoring of large-scale server clusters through the intelligent cloud data platform and intelligent cloud computing power servers, and conducts real-time and refined monitoring of large-model training data resources in large-scale cloud data centers.

7. A server collaborative monitoring system for a large-scale cloud data center according to claim 6, characterized in that: The cross-monitoring network of Zhisuan Cloud Server includes: Algorithm integration AI server, through the large-scale intelligent cloud data platform, build AI server through algorithm integration according to data type; Intelligent computing cloud computing service cluster, deploys and schedules AI servers and general servers to build an intelligent computing cloud computing server cluster; The cross-monitoring network builds a cross-monitoring network topology based on the intelligent cloud computing server cluster to form an intelligent cloud server cross-monitoring network.

8. The server collaborative monitoring system for a large-scale cloud data center according to claim 6, characterized in that: The computing power mechanism cooperates with the detection subsystem, including: The detection mechanism builds a subsystem. The Zhisuan Cloud computing server detects the server connection status based on the keep-alive detection mechanism. The server cluster status detection subsystem performs large-scale server collaborative detection based on server connection status detection to detect the status of large-scale server clusters.

9. The server collaborative monitoring system for large-scale cloud data centers according to claim 6, characterized in that: Cluster classification collaborative configuration subsystem, including: The cluster status classification subsystem classifies the status of large-scale server clusters through cloud data centers; The data resource coordination subsystem determines the allocation of data resources for various cluster states according to the classification of large-scale server cluster states, and coordinates the configuration of data resources for large-scale server cluster states; According to the classification of large-scale server cluster status, the allocation of various types of cluster status data resources is determined, and cluster status data resource configuration is performed, including: Zhisuan Cloud Computing Server classifies the status of large-scale server clusters, determines the allocation of various types of cluster status data resources, performs large-scale server data resource collaboration, collaboratively processes the status of similar clusters, and collaboratively configures cluster status data resources.

10. The server collaborative monitoring system for large-scale cloud data centers according to claim 6, characterized in that: The intelligent cloud server collaborative monitoring subsystem includes: The server cluster resource collaborative monitoring subsystem uses the Zhisuan Cloud data platform and Zhisuan Cloud computing power servers to conduct collaborative monitoring of large-scale server cluster status data resources. The collaborative monitoring subsystem for large model training resources conducts real-time and refined monitoring of large model training data resources in large-scale cloud data centers based on the collaborative monitoring of the Zhisuan Cloud data platform and the Zhisuan Cloud computing power server.

Images