Network data processing method, terminal, storage medium and processor

Abstract

This invention discloses a network data processing method, terminal, storage medium, and processor, belonging to the field of cloud network technology. The network data processing method includes: acquiring at least one network component corresponding to a target instance; collecting time-series data from the at least one network component; determining the attribute information of the target instance based on the collected time-series data and / or the geographical location of the network component from which the time-series data originates, wherein the attribute information includes: time attributes corresponding to indicators; acquiring at least one of the following: configuration information of the target instance itself, configuration information of the network components, connection relationship information between at least one network component, and connection relationship information between the target instance and at least one network component; and establishing a model of the target instance based on the attribute information and configuration information, wherein the model is used to display the network information of the target instance. This invention solves the technical problem of the difficulty in acquiring and understanding network data in the prior art.

Description

Technical Field

[0001] This invention relates to the field of cloud network technology, and in particular to a network data processing method, terminal, storage medium, and processor. Background Technology

[0002] Currently, cloud networks are primarily implemented using SDN (Software-Defined Networking). SDN separates the network control plane and forwarding plane, with the control plane responsible for distributing network configurations and the forwarding plane carrying network data traffic. On the same underlay network, one or more virtual overlay logical networks can be constructed through network virtualization technology. In this model, cloud networks can abstract various network infrastructures and basic configurations, extracting diverse product forms, making the network more flexible and reliable.

[0003] However, cloud networks are typically large in scale, with scattered data distribution, diverse data storage, and numerous dimensions, making them difficult to operate or maintain. For example, when analyzing cloud network instances, since instances involve a series of underlying physical devices, operators need to accurately select and understand useful information from the large amount of data generated by these underlying physical devices across various dimensions, resulting in a huge workload and low efficiency.

[0004] Therefore, existing technologies present significant technical challenges in acquiring and understanding cloud network data. Currently, no effective solution has been proposed to address these issues. Summary of the Invention

[0005] This invention provides a network data processing method, terminal, storage medium, and processor to at least solve the technical problem of the difficulty in acquiring and understanding cloud network data in the prior art.

[0006] According to a first aspect of the present invention, a network data processing method is provided, the method comprising: acquiring at least one network component corresponding to a target instance, wherein the network component is hardware and / or software supporting the operation of the target instance on a network; collecting time-series data from the at least one network component, wherein the time-series data is data collected for indicators on the network component; determining attribute information of the target instance based on the collected time-series data and / or the geographical location of the network component from which the time-series data originates, wherein the attribute information includes: time attributes corresponding to the indicators; acquiring configuration information related to the target instance, wherein the related configuration information includes at least one of the following: configuration information of the target instance itself, configuration information of the network component, connection relationship information between at least one network component, and connection relationship information between the target instance and at least one network component; and establishing a description model of the target instance based on the attribute information and the configuration information, wherein the description model is used to display the network information of the target instance.

[0007] Furthermore, determining the attribute information of the target instance based on the collected time-series data and / or the geographical location of the network component from which the time-series data originates includes: filtering out the time-series data belonging to the target instance from the time-series data collected from at least one network component; aggregating the filtered time-series data according to the required dimensions; and determining the attribute information of the target instance based on the aggregated time-series data.

[0008] Furthermore, obtaining configuration information related to the target instance includes: obtaining the dimension table corresponding to the target instance, which is obtained from the network's management layer; and obtaining configuration information related to the target instance based on the dimension table.

[0009] Furthermore, the configuration information of the network component includes at least one of the following: the geographical location of the network component, the specification information of the network component, and the ID information of the network component.

[0010] Furthermore, the network data processing method also includes: determining target network information according to the requirements of the task, wherein the target network information is information related to the task; obtaining a description model of at least one target instance based on the target network information; and establishing a task model based on at least one description model and additional data, wherein the task model is used to display the network information corresponding to the task.

[0011] Furthermore, the network data processing method also includes: receiving model request information from the target terminal; and in response to the model request information, sending the description model or task model of the target instance to the target terminal.

[0012] According to a second aspect of the present invention, a non-volatile storage medium is also provided, wherein the non-volatile storage medium includes a stored program, wherein the program controls the device where the non-volatile storage medium is located to execute the above-described network data processing method during runtime.

[0013] According to a third aspect of the present invention, a processor is also provided, wherein the processor is configured to run a program, wherein the program executes the network data processing method described above during runtime.

[0014] According to a fourth aspect of the present invention, a network data processing terminal is also provided, including a memory, a processor, and a computer program stored in the memory and executable on the processor, wherein the processor executes the computer program to implement the network data processing method described above.

[0015] According to a fifth aspect of the present invention, a cloud network operation and maintenance system is also provided, comprising: a backend layer, the backend layer including multiple network devices on which the cloud network operation depends; a middle platform layer, the middle platform layer being communicatively connected to the backend layer, the middle platform layer obtaining data related to a target instance from the backend layer, the middle platform layer generating a description model of the target instance based on the data, the description model being used to display the network information of the target instance; and a frontend layer, the frontend layer being communicatively connected to the middle platform layer, the frontend layer obtaining the description model from the middle platform layer, and the frontend layer performing operation and maintenance on the target instance based on the description model.

[0016] Further, the middle platform layer is used to: obtain at least one network component corresponding to the target instance, wherein the network component is hardware and / or software that supports the target instance to run on the network; collect time-series data from at least one network component, wherein the time-series data is data collected for metrics on that network component; determine the attribute information of the target instance based on the collected time-series data and / or the geographical location of the network component from which the time-series data originates, wherein the attribute information includes: time attributes corresponding to the metrics and / or geographical location attributes corresponding to the metrics; obtain configuration information related to the target instance, wherein the related configuration information includes at least one of the following: configuration information of the target instance itself, configuration information of the network components, connection relationship information between at least one network component, and connection relationship information between the target instance and at least one network component; and establish a description model of the target instance based on the attribute information and configuration information.

[0017] Furthermore, determining the attribute information of the target instance based on the collected time-series data and / or the geographical location of the network component from which the time-series data originates includes: filtering out the time-series data belonging to the target instance from the time-series data collected from at least one network component; aggregating the filtered time-series data according to the required dimensions; and determining the attribute information of the target instance based on the aggregated time-series data.

[0018] Furthermore, obtaining configuration information related to the target instance includes: obtaining the dimension table corresponding to the target instance, which is obtained from the network's management layer; and obtaining configuration information related to the target instance based on the dimension table.

[0019] Furthermore, the configuration information of the network component includes at least one of the following: the geographical location of the network component, the specification information of the network component, and the ID information of the network component.

[0020] Furthermore, the front-end layer is also used to: determine the target network information according to the requirements of the task, the target network information being information related to the task; obtain a description model of at least one target instance based on the target network information; and establish a task model based on at least one description model and additional data, the task model being used to display the network information corresponding to the task.

[0021] According to a sixth aspect of the present invention, a description model for a network instance is also provided, wherein the description model is used to display network information of the network instance, and the description model includes: attribute information of the network instance, the attribute information of the network instance including: time attribute corresponding to an indicator and / or geographical location attribute corresponding to the indicator, the time attribute corresponding to the indicator being determined based on time-series data collected from at least one network component, the network component being hardware and / or software that supports the network instance to run on the network, the time-series data being data collected for indicators on the network component, and the geographical location attribute corresponding to the indicator being determined based on the geographical location of the network component from which the time-series data originates; configuration information related to the network instance, wherein the related configuration information includes at least one of the following: configuration information of the network instance itself, configuration information of the network component, connection relationship information between at least one network component, and connection relationship information between the network instance and at least one network component.

[0022] Furthermore, the configuration information of the network component includes at least one of the following: the geographical location of the network component, the specification information of the network component, and the ID information of the network component.

[0023] In implementation, the network data processing method of this invention first acquires at least one network component corresponding to the target instance. The network component is hardware and / or software that supports the target instance's operation on the network. Then, time-series data is collected for the network component's metrics. Based on the time-series data and / or the geographical location of the network component from which the time-series data originates, the target instance's attribute information is determined. This attribute information includes the time attribute corresponding to the metric and / or the geographical location attribute corresponding to the metric. During implementation, the network data processing method also acquires configuration information related to the target instance. This configuration information includes at least one of the following: the target instance's own configuration information, the network component's configuration information, the connection relationship information between at least one network component, and the connection relationship information between the target instance and at least one network component. After acquiring the target instance's attribute information and configuration information, a description model of the target instance is established based on the attribute information and configuration information. This description model is used to display the target instance's network information. Thus, the model established by combining the target instance's attribute information and its related configuration information can better describe the target instance's attributes, achieving a relatively complete description of the target instance within a standardized basic model. This approach integrates attribute information, including time and / or geographic location attributes (regional attributes), with configuration information related to the target instance, resulting in a network architecture model that better matches the target instance—a descriptive model of the target instance. This provides a more accurate and complete data model for upper-layer tasks. When analyzing the target instance, only this descriptive model is needed, eliminating the need to extract large amounts of scattered data from various network components. This effectively reduces the difficulty of data acquisition and understanding, solving the technical problem of high difficulty in acquiring and understanding cloud network data in existing technologies. Attached Figure Description

[0024] The accompanying drawings, which are included to provide a further understanding of the invention and form part of this application, illustrate exemplary embodiments of the invention and, together with their description, serve to explain the invention and do not constitute an undue limitation thereof. In the drawings:

[0025] Figure 1 A flowchart illustrating a network data processing method provided in an embodiment of the present invention;

[0026] Figure 2 A schematic diagram illustrating the execution process of a network data processing method provided in an embodiment of the present invention;

[0027] Figure 3 A schematic diagram illustrating the generation process of an index with spatiotemporal attributes in a network data processing method provided in an embodiment of the present invention;

[0028] Figure 4 This is a schematic diagram of the dimension table data processing process of a network data processing method provided in an embodiment of the present invention. Detailed Implementation

[0029] To enable those skilled in the art to better understand the present invention, the technical solutions of the present invention will be clearly and completely described below with reference to the accompanying drawings of the embodiments of the present invention. Obviously, the described embodiments are only some embodiments of the present invention, and not all embodiments. Based on the embodiments of the present invention, all other embodiments obtained by those skilled in the art without creative effort should fall within the scope of protection of the present invention.

[0030] It should be noted that the user information (including but not limited to user device information, user personal information, etc.) and data (including but not limited to data used for analysis, data stored, data displayed, etc.) involved in this application are all information and data authorized by the user or fully authorized by all parties. Furthermore, the collection, use and processing of the relevant data must comply with the relevant laws, regulations and standards of the relevant countries and regions, and corresponding operation portals are provided for users to choose to authorize or refuse.

[0031] Figure 1 This is a network data processing method according to an embodiment of the present invention, such as... Figure 1 As shown, the method includes the following steps:

[0032] Step S102: Obtain at least one network component corresponding to the target instance, wherein the network component is hardware and / or software that supports the target instance to run on the network;

[0033] Step S104: Collect time-series data from at least one network component, wherein the time-series data is data collected for metrics on the network component;

[0034] Step S106: Determine the attribute information of the target instance based on the collected time-series data and / or the geographical location of the network component from which the time-series data originates. The attribute information includes: the time attribute corresponding to the indicator.

[0035] Step S108: Obtain configuration information related to the target instance, wherein the configuration information includes at least one of the following: configuration information of the target instance itself, configuration information of network components, connection relationship information between at least one network component, and connection relationship information between the target instance and at least one network component;

[0036] Step S110: Establish a description model of the target instance based on the attribute information and configuration information, wherein the description model is used to display the network information of the target instance.

[0037] The network data processing method employing the above scheme, during implementation, first acquires at least one network component corresponding to the target instance. This network component is the hardware and / or software that supports the target instance's operation on the network. Then, time-series data is collected for the network component's metrics. Based on the time-series data and / or the geographical location of the network component from which the time-series data originates, the target instance's attribute information is determined. This attribute information includes the time attribute corresponding to the metric and / or the geographical location attribute corresponding to the metric. During implementation, the network data processing method also acquires configuration information related to the target instance. This configuration information includes at least one of the following: the target instance's own configuration information, the network component's configuration information, the connection relationship information between at least one network component, and the connection relationship information between the target instance and at least one network component. After acquiring the target instance's attribute and configuration information, a descriptive model of the target instance is established based on these information. This descriptive model is used to display the target instance's network information. Thus, the model established by combining the target instance's attribute information and its related configuration information can effectively describe the target instance's attributes, providing a relatively complete description of the target instance within a standardized basic model. This approach integrates attribute information, including time and / or geographic location attributes (regional attributes), with configuration information related to the target instance, resulting in a network architecture model that better matches the target instance—a descriptive model of the target instance. This provides a more accurate and complete data model for upper-layer tasks. When analyzing the target instance, only this descriptive model is needed, eliminating the need to extract large amounts of scattered data from various network components. This effectively reduces the difficulty of data acquisition and understanding, solving the technical problem of high difficulty in acquiring and understanding cloud network data in existing technologies.

[0038] A target instance, also known as a network instance, is a product or network object abstracted from basic network infrastructure and configuration. It is an independent computing unit obtained through cloud network resource virtualization, such as an SLB (Server Load Balancing) instance, an EIP (Elastic IP Address) instance, a NAT (Network Address Translation) instance, and so on. It should be noted that there is no strict order between the steps of obtaining the target instance's configuration information and obtaining its attribute information; either one can be executed first, the other later, or both can be executed simultaneously.

[0039] It's important to note that any hardware and / or software that supports the operation of a target instance on the network can be called a network component. For example, a network component can be a virtual machine, a physical machine, a network forwarding device, etc. Time-series data refers to data collected from metrics on network components. These metrics can take many forms, such as inbound traffic, outbound traffic, packet loss rate, etc. There can be one or multiple metrics. Time-series data, or time series data, is a data column recorded in chronological order. Data within the same column are of the same standard and are comparable. In cloud networks, time-series metrics primarily represent information collected from instances and devices at the minute or second level. This type of metric information has a time attribute, and its value changes at different times, such as an instance's inbound traffic, outbound traffic, packet loss rate, etc. A descriptive model is used to display the network information of a target instance. As the name suggests, a descriptive model can represent the network-related information of the target instance, characterize the network attributes of the target instance, and provide a relatively complete description of the target instance. For example, it can be a dataset stored in various formats.

[0040] In a preferred embodiment, determining the attribute information of the target instance based on the collected time-series data includes: filtering time-series data belonging to the target instance from time-series data collected from at least one network component; aggregating the filtered time-series data according to the required dimensions; and determining the attribute information of the target instance based on the aggregated time-series data. That is, in this embodiment, during the process of determining the attribute information of the target instance based on the time-series data, the time-series data collected from at least one network component is filtered to select the time-series data belonging to the target instance and filter out the time-series data that does not belong to the target instance. For example, the collected time-series data is cleaned through data cleaning operations to remove time-series data that is irrelevant to the target instance (e.g., filtering out classic IP addresses and leaving only EIP addresses). Since the time-series data extracted through the network component is only the data of that network component itself, its granularity is relatively dispersed (e.g., traffic data of a certain port of a certain underlying physical device, with excessively fine granularity). In order to better obtain the attribute information of the target instance, after filtering the time series data, the time series data will also be aggregated to the required dimension (for example, the time series data will be aggregated to the dimension of the instance, so as to better represent the relevant attributes of the target instance), and comprehensive data of the granularity corresponding to the target instance will be obtained, which will help to build a more accurate and effective descriptive model of the target instance in the future.

[0041] After incorporating the time information and corresponding geographic location information from the time-series data into the model, a model that can roughly describe some characteristics of the target instance (such as traffic, bandwidth, packet loss, etc.) is obtained. However, these metrics cannot fully describe the attributes of the target instance. Besides metrics with spatiotemporal attributes, the target instance may have relationships with other objects (connection information between the target instance and at least one network component). In one embodiment, obtaining configuration information related to the target instance includes: obtaining the dimension table corresponding to the target instance, which is obtained from the network's management layer; and obtaining configuration information related to the target instance based on the dimension table. Dimension table data, also known as dimension table data, mainly stores configuration-related information. This type of information mainly stores relationships between entities, entity specifications, etc., and usually includes geographic information, or more granular availability zone or cluster information. By obtaining the dimension table from the network's management layer, and through the content configured in the dimension table, configuration information related to the target instance can be obtained relatively easily and accurately, including the target instance's own configuration information, network component configuration information, connection information between at least one network component, and connection information between the target instance and at least one network component. For example, by performing ETL processing on dimension table data, a network topology can be obtained. Based on the network topology, the association information between the target instance and network components can be determined, such as the association information between the target instance and other instances, the association information between the target instance and a specific user, or the association information between the target instance and a physical machine. This facilitates the accurate creation of a description model of the target instance based on the configuration information. The configuration information of the network component refers to the configuration information related to the network component. In specific implementations, it can include a variety of configuration information, as long as it is related to the network component. For example, in a preferred embodiment, the configuration information of the network component includes at least one of the following: the geographical location information of the network component, the specification information of the network component, and the ID information of the network component.

[0042] The ETL process described above, or Extract-Transform-Load, is used to extract, transform, and load data from the source to the destination. After processing the dimension table data in this way, the network topology can be obtained. Based on the network topology built upon the dimension table data, the relationship information between the target instance and network components can be determined, thus enabling the established descriptive model to more comprehensively describe the target instance.

[0043] Based on the aforementioned descriptive model of the target instance, this model can comprehensively describe the attributes of the target instance, thereby effectively reducing the difficulty of acquiring and understanding data related to the target instance. However, in reality, different tasks have different specific data requirements, which may not be limited to the data of the target instance. In this embodiment, the network data processing method further includes: determining target network information according to the task requirements, whereby the target network information is information related to the task requirements; obtaining at least one descriptive model of the target instance based on the target network information; and establishing a task model based on at least one descriptive model and additional data. The task model is used to display the network information corresponding to the task (i.e., related to the task). In this embodiment, the target network information is determined according to the specific task requirements. Since the target instance model is used to display the network information of the target instance, based on the specific requirements of the target network information, at least one target instance model can be determined. After obtaining the corresponding target instance model, a task model is established based on at least one model and additional data. This task model is used to display the network information corresponding to the task. The additional data can be any data, and its specific content can be selected according to the actual data requirements. In other words, based on one or more basic models (models of the target instance) and any type of supplementary data, higher-dimensional domain models (task models) can be built. These domain models are used by operations within the target domain to meet specific task requirements. They provide the information needed for operations in that target domain. This hierarchical modeling approach better meets the different data needs of different domains, eliminating the need for extensive data collection and analysis. This makes the completed domain model closer to the actual task, providing better convenience for task operations in the corresponding domain. For example, in the operations and maintenance (O&M) domain, an O&M model can be built based on at least one relevant basic model and supplementary data. This O&M model includes the data required for O&M operations, effectively facilitating O&M operations or the development of O&M tools. Similarly, in the operations domain, an operations model can be built based on relevant basic models and supplementary data, more specifically meeting the needs of the operational process.

[0044] In this embodiment, the network data processing method further includes: receiving model request information from a target terminal; and in response to the model request information, sending the model of the target instance or the task model to the target terminal. After the model or task model of the target instance is built, the model and task model of the target instance can be used by different objects. For example, when the target terminal needs to perform operation and maintenance on the target instance, it can respond to the model request information sent by the target terminal and send the model of the target instance to the target terminal. After the target terminal obtains the model of the target instance, it can easily obtain the network information (attribute description) related to the target instance, thus eliminating the need to extract a large amount of scattered data from various network components, effectively reducing the difficulty of data acquisition and understanding, and facilitating the operation and maintenance of the target terminal. For example, in the process of developing network operation and maintenance tools, the target terminal needs to rely on a large amount of task data. After the operation and maintenance-related task model is established, the target terminal can directly obtain the task model, thereby achieving efficient data acquisition without extracting a large amount of scattered data from various network components, effectively reducing the difficulty of data acquisition and understanding.

[0045] By establishing an operation and maintenance (O&M) model based on a foundational model, which includes network information required for O&M operations, the model effectively adapts to O&M data needs. Developing cloud network O&M tools based on this model significantly simplifies data acquisition during the tool development process, improving efficiency. Since target instances in the network involve multiple underlying physical devices, manually locating faults requires checking each device individually, which is inefficient and impractical. However, using an O&M model built on the aforementioned network data processing methods allows for fault location from a big data perspective, effectively facilitating the O&M of instances within the network.

[0046] The following will use Alibaba Cloud's network operation and maintenance process as an example to explain in detail: Currently, Alibaba Cloud's cloud network is mainly implemented through SDN, separating the network control plane and forwarding plane. The control plane is responsible for distributing network configurations, while the forwarding plane carries network data traffic. This allows the cloud network to abstract various network infrastructures and basic configurations, extracting various product forms, making the network more flexible and reliable. At the same time, the objects and instances that need to be operated and maintained by the network also become more complex, requiring operation and maintenance personnel to accurately select useful information from a large amount of data across various dimensions.

[0047] Although cloud network operations and maintenance (O&M) are moving towards automation and intelligence, and a large number of cloud network O&M tools and systems have emerged, greatly improving O&M efficiency, this merely delegates some repetitive and predictable operations to these tools. The developers of these tools still need to invest significant resources in understanding and selecting data. Currently, various O&M systems and tools suffer from difficulties in data acquisition, high costs in data understanding, scattered data distribution, numerous data dimensions, and diverse data storage. The data acquisition process involves a large amount of customization work, redundant development, and low efficiency.

[0048] For the operation and maintenance of large-scale cloud networks, data is the foundation. Therefore, the problems caused by scale include: the huge amount of data collected; numerous data dimensions, with different operation and maintenance scenarios requiring different data support; high data understanding costs; cloud network operation and maintenance involves many network components, and cloud network product developers are often only familiar with their own team's products and not very familiar with the implementation methods of other teams' products. This leads to the need for a platform to carry the data of the entire cloud network and to be able to reasonably model the data in the global operation and maintenance process across components and products.

[0049] The network data processing method of this invention collects two types of data: time-series indicator data and dimension table data. Time-series indicator data mainly refers to indicator information collected on instances and devices at the minute or second level. This type of indicator information has a time attribute, and the value of the indicator changes at different times. Dimension table data mainly contains configuration-related information. This type of information primarily stores relationships between entities, entity specifications, etc. This information usually includes geographic information, or more granular availability zone or cluster information. A cluster is a smaller scope below an availability zone, and an availability zone is a smaller scope below that. For example, a region includes multiple availability zones, and an availability zone includes multiple clusters.

[0050] The specific implementation method is as follows:

[0051] First, such as Figure 3As shown, time-series data is collected geographically (e.g., time-series indicators from Hangzhou, Shanghai, and Shenzhen are collected separately), and the data is cleaned and aggregated according to the required dimensions, while also being labeled with regional tags to obtain indicators with spatiotemporal attributes. At this stage, the data can only describe some attributes of network objects (target instances). The complete attributes of network objects may be scattered across multiple time-series indicators in multiple regions and dimensions, and the descriptions of objects at different times will differ. Therefore, it is necessary for the object description to include time attributes. Time-series indicators naturally have time attributes, and with the regional tags added during data collection, the cloud network object model built based on this data simultaneously possesses both time and spatial attributes.

[0052] like Figure 2 As shown, time-series data from VMs (virtual machines), NCs (physical machines), and forwarding devices are cleaned and aggregated before being stored in a data warehouse. While this time-series data is collected from VMs, NCs, and forwarding devices, since VMs also run on physical devices, this time-series data is essentially collected from the underlying physical devices of the cloud network. When a portion of these time-series metrics is used as part of the cloud network object model (basic model), a model that roughly describes some characteristics of the object (instance) is obtained. However, these metrics cannot meet all the task's requirements for the cloud network object model. Besides time-series metrics, it is also necessary to obtain the relationships between various objects or instances (part of the configuration information) and some non-time-series data, such as specifications, the physical machine where the VM resides, and the user to which the instance belongs. This information cannot be obtained from time-series metrics, but can be effectively obtained by relying on dimension table data.

[0053] like Figure 4 As shown, taking the search for the association information of EIP instances (the connection relationships between EIP instances and other network components) as an example, dimension table data, after ETL processing, can be used to solidify the network topology. Based on the network topology built upon the dimension table data, association information can be added to the cloud network object model, such as... Figure 4Through ETL processing of the dimension table data, the association information between User, EIP (Elastic IP Address), VPC (Virtual Private Cloud), Ecs (Elastic Compute Service), and nc (physical machine) was obtained. This association was then linked to other objects related to the EIP object, ultimately forming a complete description of the cloud network object, i.e., the target instance model (basic model). By combining the target instance's attribute information and related configuration information to build the target instance model, this model can effectively describe the target instance's attributes. It provides a relatively complete description of the target instance within a standardized basic model, thus offering a more accurate and complete data model for upper-layer tasks. When analyzing the target instance, only this basic model needs to be used, eliminating the need to extract large amounts of scattered data from various network components. This effectively reduces the difficulty of data acquisition and understanding, solving the technical problem of high difficulty in acquiring and understanding cloud network data in existing technologies.

[0054] The construction of a basic cloud network model based on spatiotemporal attribute indicators and network topology models establishes the foundational model of the cloud network. This foundational model models the objects within the cloud network. However, in practical applications, different domains have different data requirements. Service providers or data users can then build higher-dimensional models—domain models (task models)—on top of the foundational model based on their specific needs. For example, in the cloud network operations and maintenance (O&M) domain, the basic EIP model and device models, along with other information from spatiotemporal indicators or network topology, can be packaged to create the required O&M model. This allows for fulfilling cloud network O&M-related task requirements. This layered modeling approach better meets the needs of various domains. In other words, the foundational model aims to unify the description of objects and reduce the difficulty of data understanding and acquisition. Domain models, on the other hand, are more task-oriented and can be built according to the needs of different domains. By modeling the massive amounts of cloud network data, a network product object (instance) can be accurately described at the data level. Combined with multi-dimensional spatiotemporal data and network topology information, development, O&M, and operations personnel can easily understand and acquire the necessary data, facilitating task implementation.

[0055] Secondly, embodiments of the present invention also provide a non-volatile storage medium, wherein the non-volatile storage medium includes a stored program, wherein the program controls the device where the non-volatile storage medium is located to execute the above-described network data processing method during runtime.

[0056] Secondly, embodiments of the present invention also provide a processor, wherein the processor is used to run a program, wherein the program executes the above-described network data processing method during runtime.

[0057] In addition, embodiments of the present invention also provide a network data processing terminal, including a memory, a processor, and a computer program stored in the memory and executable on the processor, wherein the processor executes the computer program to implement the network data processing method described above.

[0058] This data processing terminal can build a model of a target instance by combining its attribute information and related configuration information. This model can effectively describe the attributes of the target instance and provide a relatively complete description of the target instance in a standardized basic model. This allows for a more accurate and complete data model to be provided to upper-layer tasks. When analyzing the target instance, only this basic model needs to be used, without the need to extract large amounts of scattered data from various network components. This effectively reduces the difficulty of data acquisition and understanding, and solves the technical problem of high difficulty in acquiring and understanding cloud network data in existing technologies.

[0059] Furthermore, embodiments of the present invention also provide a cloud network operation and maintenance system, comprising: a backend layer, which includes multiple network devices on which the cloud network operation depends; a middle platform layer, which is communicatively connected to the backend layer, wherein the middle platform layer obtains data related to a target instance from the backend layer, and generates a description model of the target instance based on the data, the description model being used to display the network information of the target instance; and a frontend layer, which is communicatively connected to the middle platform layer, wherein the frontend layer obtains the description model from the middle platform layer, and performs operation and maintenance on the target instance based on the description model.

[0060] As mentioned above, in the operation or maintenance of cloud network instances in related technologies, since instances involve a series of underlying physical devices, operators need to accurately select and understand useful information from the large amount of data generated by these underlying physical devices across various dimensions. In this implementation, the cloud network operation and maintenance system adopts a two-layer architecture design, namely, it includes a backend layer and a frontend layer. The backend layer includes multiple underlying devices. During operation and maintenance, the frontend layer directly extracts data from the multiple underlying devices in the backend layer and processes the data accordingly to meet the operation and maintenance needs of the corresponding instance. In this process, if there are multiple frontend layers, when multiple frontend layers need to use data related to the same instance, they need to extract data from the backend layer separately, process and store it, resulting in high data processing and storage costs. The cloud network operation and maintenance system of this invention, through the design of a middle platform layer, can obtain data from the back-end layer, process the data, and generate a description model of the target instance. The description model is used to display the network information of the target instance. When any front-end layer needs to perform operation and maintenance operations on the target instance, it only needs to obtain the description model of the target instance directly from the middle platform layer and perform operation and maintenance based on the description model. In this process, there is no need for each front-end layer to extract and store data, saving data processing and storage costs and solving the problem of high data processing and storage costs in cloud network operation and maintenance processes in related technologies.

[0061] Since the middle platform layer generates a description model of the target instance based on the acquired data, the front-end layer can directly perform operation and maintenance operations related to the target instance based on this description model. Therefore, the accuracy of the description model directly affects the operation and maintenance effect of the target instance. To improve the accuracy of the operation and maintenance of the target instance, in this embodiment, the middle platform layer is used to: acquire at least one network component corresponding to the target instance, wherein the network component is hardware and / or software that supports the target instance to run on the network; collect time-series data from at least one network component, wherein the time-series data is data collected for the indicators on the network component; determine the attribute information of the target instance based on the collected time-series data and / or the geographical location of the network component from which the time-series data originates, wherein the attribute information includes: the time attribute corresponding to the indicator and / or the geographical location attribute corresponding to the indicator; acquire the configuration information related to the target instance, wherein the related configuration information includes at least one of the following: the configuration information of the target instance itself, the configuration information of the network component, the connection relationship information between at least one network component, and the connection relationship information between the target instance and at least one network component; and establish a description model of the target instance based on the attribute information and the configuration information.

[0062] In the process of generating a descriptive model of the target instance based on the acquired data, the middle platform layer first acquires at least one network component corresponding to the target instance. The network component is the hardware and / or software that supports the target instance's operation on the network. Then, time-series data is collected for the network component's metrics. Based on this time-series data, the attribute information of the target instance is determined. This attribute information includes the time attribute corresponding to the metric. During implementation, the network data processing method also acquires configuration information related to the target instance. This configuration information includes at least one of the following: the target instance's own configuration information, the network component's configuration information, the connection relationship information between at least one network component, and the connection relationship information between the target instance and at least one network component. After acquiring the target instance's attribute and configuration information, a descriptive model of the target instance is built based on these information. This descriptive model is used to display the target instance's network information. In this way, the model built by combining the target instance's attribute information and its related configuration information can effectively describe the target instance's attributes, achieving a relatively complete description of the target instance within a standardized basic model. Specifically, by integrating attribute information, including time attributes and / or geographic location attributes (regional attributes), with configuration information related to the target instance, a network architecture model that better matches the target instance is obtained, i.e., a descriptive model of the target instance. This provides a more accurate and complete data model for front-end layer operation and maintenance tasks, which is beneficial to improving the accuracy of target instance operation and maintenance.

[0063] In a preferred embodiment, determining the attribute information of the target instance based on the collected time-series data and / or the geographical location of the network component from which the time-series data originates includes: filtering time-series data belonging to the target instance from time-series data collected from at least one network component; aggregating the filtered time-series data according to the required dimensions; and determining the attribute information of the target instance based on the aggregated time-series data. Since the time-series data extracted from the network component is only the data of that network component itself, its granularity is relatively dispersed (e.g., traffic data from a single port of a certain underlying physical device, with excessively fine granularity). To better obtain the attribute information of the target instance, after filtering the time-series data, an aggregation operation is performed to aggregate the time-series data to the required dimensions (e.g., aggregating the time-series data to the instance dimension to better characterize the attributes related to the target instance), obtaining comprehensive data of the target instance's granularity, which is beneficial for subsequently establishing a more accurate and effective descriptive model of the target instance.

[0064] Obtaining configuration information related to the target instance includes: retrieving the dimension table corresponding to the target instance, which is obtained from the network management layer; and retrieving the configuration information related to the target instance based on the dimension table. Dimension table data primarily stores configuration-related information, such as relationships between entities, entity specifications, etc., and often includes geographic information, or more granular availability zone or cluster information. By retrieving the dimension table from the network management layer, the configuration information related to the target instance can be obtained relatively easily and accurately through the content configured in the dimension table. This includes the target instance's own configuration information, network component configuration information, connection relationships between at least one network component, and connection relationships between the target instance and at least one network component. For example, by performing ETL processing on the dimension table data, the network topology can be obtained. Based on the network topology, the relationship information between the target instance and network components can be determined, such as the relationship between the target instance and other instances, the relationship between the target instance and a specific user, or the relationship between the target instance and physical machines. This facilitates the subsequent accurate creation of a descriptive model of the target instance based on this configuration information. The configuration information of a network component refers to the configuration information related to the network component. In specific implementations, it can include a variety of configuration information, as long as it is related to the network component. For example, in a preferred embodiment, the configuration information of the network component includes at least one of the following: the geographical location information of the network component, the specification information of the network component, and the ID information of the network component.

[0065] In this embodiment, the front-end layer is further configured to: determine target network information based on task requirements, wherein the target network information is information related to the task; obtain a description model of at least one target instance based on the target network information; and establish a task model based on at least one description model and additional data, wherein the task model is used to display the network information corresponding to the task. Based on the established description model of the target instance, this model can describe the attributes of the target instance relatively completely, thereby effectively reducing the difficulty of obtaining and understanding data related to the target instance. However, in reality, different tasks have different specific data requirements, which may not be limited to the data of the target instance. In this embodiment, the network data processing method further includes: determining target network information based on task requirements, wherein the target network information is information related to task requirements; obtaining a description model of at least one target instance based on the target network information; and establishing a task model based on at least one description model and additional data, wherein the task model is used to display the network information corresponding to the task (i.e., task-related). In this embodiment, target network information is determined based on specific task requirements. Since the target instance model is used to display the network information of the target instance, at least one target instance model can be determined based on the specific requirements of the target network information. After obtaining the corresponding target instance model, a task model is built based on at least one model and additional data. This task model is used to display the network information corresponding to the task. The additional data can be any data, and its specific content can be selected according to actual data requirements. That is to say, based on one or more basic models (target instance models) and any type of additional data, a higher-dimensional domain model (task model) can be built. The domain model is used for target domain operations to meet specific task requirements. It can provide the information required for the target domain operations. This hierarchical model building method can better meet the different data requirements of different domains, without the need for large-scale data collection and analysis processing, thus making the built domain model closer to the actual task and providing better convenience for task operations in the corresponding domain. For example, when used in the operations and maintenance (O&M) field, an O&M model can be built based on at least one relevant basic model and additional data. This O&M model includes the data required for O&M operations, thereby effectively facilitating O&M operations or the development of O&M tools. As another example, when used in the operations field, an operations model can be built based on relevant basic models and additional data, thereby more specifically meeting the needs of the operational process.

[0066] Finally, embodiments of the present invention also provide a description model of a network instance, wherein the description model is used to display network information of the network instance, and the description model includes: attribute information of the network instance, which includes: time attribute corresponding to an indicator and / or geographical location attribute corresponding to the indicator. The time attribute corresponding to the indicator is determined based on time-series data collected from at least one network component. The network component is hardware and / or software that supports the operation of the network instance on the network. The time-series data is data collected for indicators on the network component. The geographical location attribute corresponding to the indicator is determined based on the geographical location of the network component from which the time-series data originates. Configuration information related to the network instance, wherein the related configuration information includes at least one of the following: configuration information of the network instance itself, configuration information of the network component, connection relationship information between at least one network component, and connection information between the network instance and at least one network component. In the process of creating the descriptive model for the Wana Luo instance, the connection relationship information of network components is first obtained. This involves acquiring at least one network component corresponding to the target instance. The network component is the hardware and / or software that supports the target instance's operation on the network. Then, time-series data is collected for the network component's metrics. Based on this time-series data, the target instance's attribute information is determined. This attribute information includes the time attribute corresponding to the metric. During implementation, the network data processing method also acquires the target instance's configuration information. This configuration information includes at least one of the following: the target instance's own configuration information, the network component's configuration information, the connection relationship information between at least one network component, and the connection relationship information between the target instance and at least one network component. After acquiring the target instance's attribute and configuration information, a descriptive model of the target instance is built based on these information. This descriptive model is used to display the target instance's network information. In this way, the model built by combining the target instance's attribute information and its related configuration information can effectively describe the target instance's attributes, achieving a relatively complete description of the target instance within a standardized basic model. Specifically, by fusing attribute information, including time attributes and / or geographic location attributes (regional attributes), with configuration information related to the target instance, a network architecture model that better matches the target instance is obtained, i.e., a descriptive model of the target instance. This provides a more accurate and complete data model for upper-layer tasks. When analyzing the target instance, only this descriptive model needs to be used, without needing to extract large amounts of scattered data from various network components. This effectively reduces the difficulty of data acquisition and understanding, and solves the technical problem of high difficulty in acquiring and understanding cloud network data in existing technologies.

[0067] The configuration information of a network component can be in various forms. In one specific embodiment, the configuration information of a network component includes at least one of the following: the geographical location information of the network component, the specification information of the network component, and the ID information of the network component.

[0068] The sequence numbers of the above embodiments of the present invention are merely for descriptive purposes and do not represent the superiority or inferiority of the embodiments. Moreover, the steps shown in the flowcharts of the accompanying drawings can be executed in a computer system such as a set of computer-executable instructions, and although a logical order is shown in the flowcharts, in some cases the steps shown or described may be performed in a different order than that shown here.

[0069] In the above embodiments of the present invention, the descriptions of each embodiment have different focuses. For parts not described in detail in a certain embodiment, please refer to the relevant descriptions of other embodiments.

[0070] In the several embodiments provided in this application, it should be understood that the disclosed technical content can be implemented in other ways. The device embodiments described above are merely illustrative; for example, the division of units can be a logical functional division, and in actual implementation, there may be other division methods. For instance, multiple units or components may be combined or integrated into another system, or some features may be ignored or not executed. Furthermore, the displayed or discussed mutual coupling, direct coupling, or communication connection may be through some interfaces; the indirect coupling or communication connection between units or modules may be electrical or other forms.

[0071] The units described as separate components may or may not be physically separate. The components shown as units may or may not be physical units; that is, they may be located in one place or distributed across multiple units. Some or all of the units can be selected to achieve the purpose of this embodiment according to actual needs.

[0072] Furthermore, the functional units in the various embodiments of the present invention can be integrated into one processing unit, or each unit can exist physically separately, or two or more units can be integrated into one unit. The integrated unit can be implemented in hardware or as a software functional unit.

[0073] If the integrated unit is implemented as a software functional unit and sold or used as an independent product, it can be stored in a computer-readable storage medium. Based on this understanding, the technical solution of the present invention, in essence, or the part that contributes to the prior art, or all or part of the technical solution, can be embodied in the form of a software product. This computer software product is stored in a storage medium and includes several instructions to cause a computer device (which may be a personal computer, server, or network device, etc.) to execute all or part of the steps of the methods described in the various embodiments of the present invention. The aforementioned storage medium includes various media capable of storing program code, such as USB flash drives, read-only memory (ROM), random access memory (RAM), portable hard drives, magnetic disks, or optical disks.

[0074] The above description is only a preferred embodiment of the present invention. It should be noted that for those skilled in the art, several improvements and modifications can be made without departing from the principle of the present invention, and these improvements and modifications should also be considered within the scope of protection of the present invention.

Claims

1. A network data processing method, comprising: obtaining at least one network component corresponding to a target instance, wherein the network component is hardware and / or software providing support for running the target instance on a network, and the target instance is a product or a network object obtained by abstracting underlying network infrastructure and configuration; collecting time series data from at least one network component, wherein the time series data is data collected for an index on the network component; determining attribute information of the target instance according to the collected time series data and a geographical location where the network component from which the time series data is derived is located, wherein the attribute information comprises a time attribute corresponding to the index and a geographical location attribute corresponding to the index; obtaining configuration information related to the target instance, wherein the related configuration information comprises at least one of the following: configuration information of the target instance itself, configuration information of the network component, connection relationship information between at least one network component, and connection relationship information between the target instance and at least one network component; and establishing a description model of the target instance according to the attribute information and the configuration information, wherein the description model is used to show network information of the target instance; wherein the determining of the attribute information of the target instance according to the collected time series data and the geographical location where the network component from which the time series data is derived is located comprises: filtering time series data belonging to the target instance from the time series data collected from at least one network component; aggregating the filtered time series data according to a required dimension; and determining the attribute information of the target instance according to the aggregated time series data; wherein the obtaining of the configuration information related to the target instance comprises: obtaining a dimension table corresponding to the target instance, wherein the dimension table is obtained from a management layer of the network; and obtaining the configuration information related to the target instance according to the dimension table; wherein the configuration information of the network component comprises at least one of the following: regional information of the network component, specification information of the network component, and ID information of the network component; wherein the network data processing method further comprises: determining target network information according to a requirement of a task, wherein the target network information is information related to the task; obtaining a description model of at least one target instance according to the target network information; and establishing a task model according to at least one description model and additional data, wherein the task model is used to show network information corresponding to the task; wherein the network data processing method further comprises: receiving model request information from a target terminal; and sending the description model of the target instance or the task model to the target terminal in response to the model request information; wherein the non-volatile storage medium comprises a stored program, wherein the non-volatile storage medium controls a device in which the non-volatile storage medium is located to perform the network data processing method of any one of claims 1 to 5 when the program is running; and wherein the processor is configured to run a program, wherein the processor performs the network data processing method of any one of claims 1 to 5 when the program is running. ​ ​ ​ ​ ​ ​ ​ ​ ​ 2. The network data processing method of claim 1, wherein, ​ ​ ​ 3. The network data processing method according to claim 1 or 2, wherein, ​ 4. The network data processing method according to claim 1 or 2, wherein, ​ ​ ​ ​ 5. The network data processing method of claim 4, wherein, ​ ​ ​ 6. A non-volatile storage medium, wherein, ​ 7. A processor, wherein, ​ 8. A network data processing terminal comprising a memory, a processor, and a computer program stored in said memory and executable on said processor, wherein, The processor implements the network data processing method of any one of claims 1-5 when executing the computer program.

9. A cloud network operation system, comprising: a background layer including a plurality of network devices relied on by a cloud network; a middle layer in communication connection with the background layer, the middle layer obtaining data related to a target instance from the background layer, the middle layer generating a description model of the target instance according to the data, the description model being used to show network information of the target instance; a front layer in communication connection with the middle layer, the front layer obtaining the description model from the middle layer, the front layer performing operation and maintenance on the target instance based on the description model; wherein the middle layer is configured to: obtain at least one network component corresponding to the target instance, wherein the network component is hardware and / or software providing support for running of the target instance on a network, and the target instance is a product or a network object obtained by abstracting a basic network infrastructure and configuration; collect time series data from at least one of the network components, wherein the time series data is data collected for an index on the network component; determine attribute information of the target instance according to the collected time series data and / or a geographical location where the network component from which the time series data is sourced is located, wherein the attribute information includes a time attribute corresponding to the index and / or a geographical location attribute corresponding to the index; obtain configuration information related to the target instance, wherein the related configuration information includes at least one of the following: configuration information of the target instance itself, configuration information of the network component, connection relationship information between at least one of the network components, and connection relationship information between the target instance and at least one of the network components; establish a description model of the target instance according to the attribute information and the configuration information; wherein determining attribute information of the target instance according to the collected time series data and / or a geographical location where the network component from which the time series data is sourced is located includes: filtering time series data belonging to the target instance from the time series data collected from at least one of the network components; aggregating the filtered time series data according to required dimensions; and determining attribute information of the target instance according to the aggregated time series data.

10. The cloud network operation system of claim 9, wherein, Obtaining configuration information related to the target instance includes: obtaining a dimension table corresponding to the target instance, the dimension table being obtained from a management layer of the network; and obtaining the configuration information related to the target instance according to the dimension table. 11.The cloud network operation and maintenance system of claim 9, wherein, The configuration information of the network component includes at least one of the following: regional information of the network component, specification information of the network component, and ID information of the network component.

12. The cloud network operations system of any of claims 9 to 11, wherein, The front layer is further configured to: determine target network information according to a requirement of a task, the target network information being information related to the task; obtain a description model of at least one of the target instances according to the target network information; establish a task model according to at least one of the description models and additional data, the task model being used to show network information corresponding to the task.

Images