Cloud network resource allocation method, device, computer equipment and storage medium

Abstract

The present application relates to a cloud network resource allocation method, apparatus, computer equipment, and storage medium. The method comprises: determining a first service flow and a second service flow from cloud network service flows to which cloud network resources are to be allocated; the priority of the first service flow is higher than the priority of the second service flow; determining first resource allocation information for allocating the cloud network resources to the first service flow, and determining the remaining resources of the cloud network resources based on the first resource allocation information; determining second resource allocation information for allocating the remaining resources to the second service flow; and allocating the cloud network resources to the first service flow and the second service flow based on the first resource allocation information and the second resource allocation information. This method can improve the utilization rate of cloud network resources.

Description

Technical Field

[0001] The present application relates to the field of communication technology, and in particular to a cloud network resource allocation method, apparatus, computer equipment, and storage medium. Background Art

[0002] Traffic engineering refers to the rational scheduling of traffic in the network to optimize network resources, such as minimizing and maximizing link bandwidth utilization, maximizing the number of concurrent flows, and minimizing costs.

[0003] With the development of cloud-network convergence technology, traditional traffic engineering has focused solely on optimizing network resources, failing to rationally utilize cloud resources, resulting in low cloud resource utilization. Furthermore, to address random network failures and traffic uncertainty, a large amount of redundant bandwidth is often required to avoid service congestion, keeping network links underloaded, resulting in low network bandwidth utilization.

[0004] Therefore, the current cloud network resource allocation technology has the problem of low resource utilization. Summary of the Invention

[0005] Based on this, it is necessary to provide a cloud network resource allocation method, device, computer equipment, computer-readable storage medium and computer program product that can improve resource utilization in response to the above technical problems.

[0006] In a first aspect, the present application provides a cloud network resource allocation method. The method comprises:

[0007] Determining a first service flow and a second service flow from the cloud network service flows to which cloud network resources are to be allocated; the priority of the first service flow is higher than the priority of the second service flow;

[0008] Determining first resource allocation information for allocating the cloud network resource to the first service flow, and determining remaining resources of the cloud network resource based on the first resource allocation information;

[0009] Determine second resource allocation information for allocating the remaining resources to the second service flow;

[0010] The cloud network resources are allocated to the first service flow and the second service flow according to the first resource allocation information and the second resource allocation information.

[0011] In one embodiment, determining first resource allocation information for allocating the cloud network resource to the first service flow includes:

[0012] Determining, based on the cloud network resources, a first candidate path that matches a first service requirement of the first service flow;

[0013] A first target path is determined from the first candidate paths, and resource allocation information corresponding to the first target path is determined as the first resource allocation information.

[0014] In one embodiment, determining the remaining resources of the cloud network resources according to the first resource allocation information includes:

[0015] Determining a resource redundancy amount of the first service flow;

[0016] Obtaining a predicted resource occupancy of the first service flow according to the first resource allocation information and the resource redundancy;

[0017] The predicted resource occupancy is removed from the cloud network resources to obtain the remaining resources.

[0018] In one embodiment, determining the resource redundancy of the first service flow includes:

[0019] Determining an excess probability of the first service flow based on the historical resource occupancy of the first service flow; the excess probability is used to represent a probability that the actual resource occupancy of the first service flow exceeds the first resource allocation information;

[0020] The resource redundancy of the first service flow is determined according to the excess probability.

[0021] In one embodiment, determining second resource allocation information for allocating the remaining resources to the second service flow includes:

[0022] determining, based on the remaining resources, a second candidate path that matches a second service requirement of the second service flow;

[0023] A second target path is determined from the second candidate paths, and resource allocation information corresponding to the second target path is determined as the second resource allocation information.

[0024] In one embodiment, allocating the cloud network resources to the first service flow and the second service flow according to the first resource allocation information and the second resource allocation information includes:

[0025] In response to the update operation on the first business requirement, determining a new first business requirement;

[0026] performing a first adjustment operation on the first resource allocation information and a second adjustment operation on the second resource allocation information according to the new first service requirement; wherein the first adjustment operation is in opposite directions to the second adjustment operation;

[0027] The cloud network resources are allocated to the first business flow and the second business flow according to the adjusted first resource allocation information and the adjusted second resource allocation information.

[0028] In one embodiment, determining a first target path from the first candidate paths and determining resource allocation information corresponding to the first target path as the first resource allocation information includes:

[0029] Determining a resource update model for the first service flow; wherein the resource update model takes minimizing the total delay of the first service flow as an update goal;

[0030] The first candidate path is input into the resource update model to obtain the first target path and the resource allocation information corresponding to the first target path, and the resource allocation information is used as the first resource allocation information.

[0031] In a second aspect, the present application further provides a cloud network resource allocation device. The device includes:

[0032] A service classification module, configured to determine a first service flow and a second service flow from cloud network service flows to which cloud network resources are to be allocated; the priority of the first service flow is higher than the priority of the second service flow;

[0033] A first information module is configured to determine first resource allocation information for allocating the cloud network resource to the first service flow, and determine remaining resources of the cloud network resource based on the first resource allocation information;

[0034] A second information module is used to determine second resource allocation information for allocating the remaining resources to the second service flow;

[0035] A resource allocation module is used to allocate the cloud network resources to the first business flow and the second business flow according to the first resource allocation information and the second resource allocation information.

[0036] In a third aspect, the present application further provides a computer device. The computer device includes a memory and a processor, wherein the memory stores a computer program, and when the processor executes the computer program, the following steps are performed:

[0037] Determining a first service flow and a second service flow from the cloud network service flows to which cloud network resources are to be allocated; the priority of the first service flow is higher than the priority of the second service flow;

[0038] Determining first resource allocation information for allocating the cloud network resource to the first service flow, and determining remaining resources of the cloud network resource based on the first resource allocation information;

[0039] Determine second resource allocation information for allocating the remaining resources to the second service flow;

[0040] The cloud network resources are allocated to the first service flow and the second service flow according to the first resource allocation information and the second resource allocation information.

[0041] In a fourth aspect, the present application further provides a computer-readable storage medium having a computer program stored thereon, which, when executed by a processor, implements the following steps:

[0042] Determining a first service flow and a second service flow from the cloud network service flows to which cloud network resources are to be allocated; the priority of the first service flow is higher than the priority of the second service flow;

[0043] Determining first resource allocation information for allocating the cloud network resource to the first service flow, and determining remaining resources of the cloud network resource based on the first resource allocation information;

[0044] Determine second resource allocation information for allocating the remaining resources to the second service flow;

[0045] The cloud network resources are allocated to the first service flow and the second service flow according to the first resource allocation information and the second resource allocation information.

[0046] In a fifth aspect, the present application further provides a computer program product. The computer program product includes a computer program that, when executed by a processor, implements the following steps:

[0047] Determining a first service flow and a second service flow from the cloud network service flows to which cloud network resources are to be allocated; the priority of the first service flow is higher than the priority of the second service flow;

[0048] Determining first resource allocation information for allocating the cloud network resource to the first service flow, and determining remaining resources of the cloud network resource based on the first resource allocation information;

[0049] Determine second resource allocation information for allocating the remaining resources to the second service flow;

[0050] The cloud network resources are allocated to the first service flow and the second service flow according to the first resource allocation information and the second resource allocation information.

[0051] The above-mentioned cloud network resource allocation method, device, computer equipment, storage medium and computer program product determine the first business flow and the second business flow from the cloud network business flow to which the cloud network resources are to be allocated, determine the first resource allocation information for allocating the cloud network resources to the first business flow, and determine the remaining resources of the cloud network resources based on the first resource allocation information, determine the second resource allocation information for allocating the remaining resources to the second business flow, and allocate cloud network resources to the first business flow and the second business flow based on the first resource allocation information and the second resource allocation information; cloud network resources can be preferentially allocated to the first business flow with a higher priority, and redundant resources can be allocated to the first business flow to ensure the reliable operation of the first business flow, while the second business flow with a lower priority does not need to be allocated with redundant resources, thereby improving the overall cloud network resource utilization. BRIEF DESCRIPTION OF THE DRAWINGS

[0052] Figure 1 Schematic diagram of a process for allocating cloud network resources in one embodiment;

[0053] Figure 2 Schematic diagram of a cloud network topology in one embodiment;

[0054] Figure 3 This is a flowchart of a cloud network traffic engineering optimization method;

[0055] Figure 4 A flowchart of a cloud network resource allocation method according to another embodiment;

[0056] Figure 5 This is a structural block diagram of a cloud network resource allocation device in one embodiment;

[0057] Figure 6 FIG. 1 is a diagram showing the internal structure of a computer device in one embodiment. DETAILED DESCRIPTION

[0058] In order to make the purpose, technical solutions and advantages of this application more clear, the following further describes this application in detail with reference to the accompanying drawings and embodiments. It should be understood that the specific embodiments described herein are only used to explain this application and are not intended to limit this application.

[0059] In one embodiment, Figure 1 As shown, a cloud network resource allocation method is provided. This embodiment uses the method applied to a terminal as an example for illustration. It is understandable that the method can also be applied to a server, and can also be applied to a system including a terminal and a server, and implemented through the interaction between the terminal and the server. In this embodiment, the method includes the following steps:

[0060] Step S110: Determine a first service flow and a second service flow from the cloud network service flows to which cloud network resources are to be allocated; the priority of the first service flow is higher than the priority of the second service flow.

[0061] Among them, cloud network resources can include network resources and cloud resources, among which network resources can be but are not limited to network link bandwidth, and cloud resources can be but are not limited to cloud storage space.

[0062] Among them, the cloud network business flow can be a business flow that requires allocation of cloud network resources.

[0063] The first service flow may be a high-priority service flow, for example, a critical flow.

[0064] The second service flow may be a low-priority service flow, such as a normal flow.

[0065] In a specific implementation, the terminal can obtain cloud network resources and classify cloud network service flows that need to be allocated cloud network resources to obtain a first service flow with a relatively high priority and a second service flow with a relatively low priority.

[0066] In actual applications, the terminal can obtain the network topology of cloud network resources, the link bandwidth of each link in the network topology, and the cloud resources of each cloud node (for example, cloud storage resources) from the cloud resource management system and network management system. It can also obtain the business needs of each business flow and classify the business flow based on the traffic characteristics of the business flow to obtain key flows and ordinary flows, where the priority of the key flow is higher than that of the ordinary flow. The key flow and the ordinary flow correspond to different optimization planes, where the optimization plane is a virtual plane used to represent different optimization levels. For example, if the key flow optimization plane has excess traffic, it can occupy the resources of the ordinary flow optimization plane and squeeze the ordinary flow. The ordinary flow can only occupy the resources of the key flow optimization plane when there are redundant resources in the key flow optimization plane.

[0067] Step S120: determine first resource allocation information for allocating cloud network resources to the first service flow, and determine remaining resources of the cloud network resources based on the first resource allocation information.

[0068] The first resource allocation information may be information on allocating cloud network resources to the first business flow, including the network link, network link bandwidth, and cloud storage space allocated to the first business flow.

[0069] Among them, the remaining resources can be the remaining resources obtained by removing the resources allocated to the first business flow from the cloud network resources, including the remaining network link bandwidth and cloud storage space.

[0070] In a specific implementation, the terminal can determine the first resource allocation information for allocating cloud network resources to the first business flow, including the network link allocated to the first business flow, and the network link bandwidth and cloud storage space corresponding to the network link. It can also remove the network link bandwidth and cloud storage space allocated to the first business flow from the cloud network resources to obtain the remaining bandwidth and remaining storage space respectively, and use the remaining bandwidth and remaining storage space as the remaining resources.

[0071] In practical applications, terminals can identify candidate paths for key flows from cloud network resources based on their business needs and establish a multi-commodity flow model based on the candidate paths. The optimization objectives of the multi-commodity flow model can be to minimize the latency and cost of all key flows, maximize the reliability of all key flows, or minimize the maximum link utilization of all key flows. By solving the multi-commodity flow model, the overall traffic distribution of key flows is optimized, and the target paths and resource allocation information for the key flows are obtained. The multi-commodity flow model can be an optimization model based on the network flow problem of multiple commodities flowing from different source nodes to different sink nodes in the network.

[0072] The terminal can also perform a probabilistic assessment of the traffic distribution results of key flows, determine the probability that the key flows may exceed the quota in each link segment and each cloud node, and based on the probabilistic assessment results, set a certain amount of redundant bandwidth for each link segment and configure certain redundant resources for each cloud node to ensure the reliable transmission of key flows. Among them, a larger redundant bandwidth and more redundant resources can be set to improve the utilization of cloud network resources.

[0073] Step S130: Determine second resource allocation information for allocating remaining resources to a second service flow.

[0074] Among them, the second resource allocation information can be information on allocating cloud network resources to the second business flow, including the network link, network link bandwidth and cloud storage space allocated to the second business flow.

[0075] In a specific implementation, the terminal can determine the second resource allocation information for allocating the remaining resources to the second business flow based on the business needs of the second business flow, including the network link allocated to the second business flow, and the network link bandwidth and cloud storage space corresponding to the network link.

[0076] In actual applications, after determining the resource allocation information, redundant bandwidth, and redundant resources of the critical flow, the terminal can remove the cloud network resources, redundant bandwidth, and redundant resources occupied by the critical flow from the original cloud network resources to obtain the remaining resources. Based on the remaining resources, the ordinary flow is optimized to obtain the target path and resource allocation information of the ordinary flow. Similar to the critical flow, a multi-commodity flow model of all ordinary flows can be established, and the overall objective function can be set for solution to achieve optimization of the overall traffic distribution.

[0077] It should be noted that the optimization objective of the common flow multi-commodity flow model may be different from the optimization objective of the critical flow multi-commodity flow model. For example, the optimization objective of the common flow multi-commodity flow model may be set to maximize throughput.

[0078] Step S140: Allocate cloud network resources to the first service flow and the second service flow according to the first resource allocation information and the second resource allocation information.

[0079] In the specific implementation, the terminal can allocate cloud network resources to the first business flow based on information such as the network link, network link bandwidth and cloud storage space allocated to the first business flow, and can also allocate the remaining cloud network resources to the second business flow based on information such as the network link, network link bandwidth and cloud storage space allocated to the second business flow.

[0080] In practical applications, terminals can configure actual cloud network resources for critical flows and ordinary flows based on their target paths and resource allocation information. Dynamic regulation can also be implemented during the configuration process. As cloud network service traffic dynamically changes, service flow traffic may increase or decrease, or even exceed the allocated redundant bandwidth and redundant resources. In this case, timely traffic guidance is required, and adjustments must be made based on the original optimization to protect critical flows. Specifically, if a critical flow exceeds its original optimization value and the allocated redundant bandwidth or redundant resources are insufficient, the critical flow is guided to the ordinary flow optimization plane, reducing the number of ordinary flows passing through the same links and nodes, prioritizing the critical flow. Conversely, if a critical flow does not exceed its original optimization value, an ordinary flow that exceeds its original optimization value can occupy the allocated redundant bandwidth and redundant resources. This allows for rapid dynamic regulation based on the original traffic optimization to meet the needs of various service flows while simultaneously improving the link and resource utilization of cloud network resources.

[0081] The above-mentioned cloud network resource allocation method determines the first business flow and the second business flow from the cloud network business flow to which the cloud network resources are to be allocated, determines the first resource allocation information for allocating the cloud network resources to the first business flow, and determines the remaining resources of the cloud network resources based on the first resource allocation information, determines the second resource allocation information for allocating the remaining resources to the second business flow, and allocates cloud network resources to the first business flow and the second business flow based on the first resource allocation information and the second resource allocation information; cloud network resources can be preferentially allocated to the first business flow with a higher priority, and redundant resources can be allocated to the first business flow to ensure the reliable operation of the first business flow, while the second business flow with a lower priority does not need to be allocated with redundant resources, thereby improving the overall utilization of cloud network resources.

[0082] In one embodiment, the above-mentioned step S120 may specifically include: determining a first candidate path that matches the first business demand of the first business flow based on cloud network resources; determining a first target path from the first candidate path, and determining resource allocation information corresponding to the first target path as the first resource allocation information.

[0083] The first business requirement may be a key flow business requirement, the first candidate path may be a key flow candidate path, and the first target path may be a key flow target path.

[0084] In a specific implementation, a first candidate path can be determined in cloud network resources according to the first business demand of the first business flow, a multi-commodity flow model can be established for the first candidate path, and all first candidate paths and their corresponding cloud network resources can be optimized through the multi-commodity flow model to obtain the first target path in the first candidate path, as well as the network link bandwidth and cloud storage space corresponding to the first target path, and the first target path and its corresponding network link bandwidth and cloud storage space are determined as the first resource allocation information.

[0085] Figure 2 Provides a schematic diagram of the cloud network topology. Figure 2 Assuming that the link bandwidth is 100Gbps, node B and node D have cloud resources, both of which are 1000G storage resources. Through traffic classification technology, based on traffic characteristics, etc., the key flows and common flows in the business flow are determined. The key flows include: flow AB, flow AD, flow AE, and the common flows include: flow FB, flow FC, flow FD. The requirements of key flows and common flows for cloud network resources are shown in Table 1.

[0086] Table 1

[0087] Flow (start-end) Bandwidth requirements Cloud resource requirements Flow AB (key flow) 50Gbps 25G Flow AD (key flow) 50Gbps 25G Flow AE (Key Flow) 50Gbps 0 Stream FB (normal stream) 30Gbps 20G Flow FC (normal flow) 30Gbps 0 Stream FD (normal stream) 30Gbps 20G

[0088] Calculate candidate paths based on the requirements of key flows. Assume that the candidate paths for flow AB are path 1: AB, path 2: AEB, and path 3: AFEB; the candidate paths for flow AD are path 1: AED, path 2: ABCD, and path 3: AFD; and the candidate paths for flow AE are path 1: AE, path 2: ABE, and path 3: AFE. Assuming that the goal is to improve the quality of service for key flows, the overall optimization objective for key flows is to minimize the latency of all key flows. A multi-commodity flow linear scaling model can be established, with the variables being the bandwidth of each candidate path for the key flows and the constraints being the link capacity constraint: the traffic carrying the key flows on each link segment must not exceed the link capacity. There are also cloud node resource capacity constraints: the sum of the resources required to carry the key flows on each cloud node must not exceed the cloud resource capacity of that node. Based on this linear programming model, the bandwidth and resource allocation of key flows are obtained: the final path of flow AB is path: AB, with an allocated bandwidth of 50 Gbps and 25 GB of storage resources allocated to node B; the final path of flow AD is path: AED, with an allocated bandwidth of 50 Gbps and 50 GB of storage resources allocated to node D; the candidate path of flow AE is path: AFE, with an allocated bandwidth of 50 Gbps.

[0089] In this embodiment, a first candidate path that matches the first business demand of the first business flow is determined based on cloud network resources; a first target path is determined from the first candidate path, and resource allocation information corresponding to the first target path is determined as the first resource allocation information. Resource allocation can be performed for key flows based on all cloud network resources to ensure reliable transmission of key flows and improve the utilization rate of cloud network resources.

[0090] In one embodiment, the above-mentioned step S120 may further specifically include: determining the resource redundancy of the first business flow; obtaining the predicted resource occupancy of the first business flow based on the first resource allocation information and the resource redundancy; and removing the predicted resource occupancy from the cloud network resources to obtain the remaining resources.

[0091] The resource redundancy may be redundant bandwidth and redundant storage space configured to ensure reliable transmission of the first service flow.

[0092] The predicted resource occupancy may be the estimated resource occupancy of the first service flow.

[0093] In a specific implementation, after obtaining the network link bandwidth and cloud storage space allocated to the first business flow, the redundant bandwidth and redundant storage space of the first business flow can be determined, the network link bandwidth allocated to the first business flow and the redundant bandwidth are added to obtain the predicted bandwidth of the first business flow, the cloud storage space allocated to the first business flow and the redundant storage space are added to obtain the predicted storage space of the first business flow, the predicted bandwidth and predicted storage space are used as the predicted resource occupancy of the first business flow, the predicted resource occupancy is subtracted from the cloud network resources, and the remaining resources of the cloud network resources are obtained.

[0094] For example, according to Figure 2 , the bandwidth and resource allocation for key flows are obtained: the final path for flow AB is path: AB, with an allocated bandwidth of 50 Gbps and 25 GB of storage resources allocated to node B; the final path for flow AD is path: AED, with an allocated bandwidth of 50 Gbps and 50 GB of storage resources allocated to node D; the candidate path for flow AE is path: AFE, with an allocated bandwidth of 50 Gbps. Assume that the redundant bandwidth and redundant storage resources for flow AB are 40 Gbps and 20 GB, respectively; the redundant bandwidth and redundant storage resources for flow AD are 5 Gbps and 5 GB, respectively; and the redundant bandwidth for flow AE is 20 Gbps. Adding the allocated bandwidth to the redundant bandwidth, and the allocated storage resources to the redundant storage resources, we obtain the predicted bandwidth and predicted storage resources for flow AB of 90 Gbps and 45 GB, respectively; the predicted bandwidth and predicted storage resources for flow AD of 55 Gbps and 55 GB, respectively; and the predicted bandwidth for flow AE of 70 Gbps. The calculated predicted bandwidth and predicted storage resources can then be subtracted from the cloud network resources to obtain the remaining bandwidth and remaining storage resources of the cloud network resources, respectively.

[0095] In this embodiment, by determining the resource redundancy of the first business flow; obtaining the predicted resource occupancy of the first business flow based on the first resource allocation information and the resource redundancy; removing the predicted resource occupancy from the cloud network resources to obtain the remaining resources, redundant resources can be allocated to the key flow to ensure the reliable transmission of the key flow and improve the utilization rate of the cloud network resources.

[0096] In one embodiment, the above-mentioned step of determining the resource redundancy of the first business flow may specifically include: determining the excess probability of the first business flow based on the historical resource occupancy of the first business flow; the excess probability is used to characterize the probability that the actual resource occupancy of the first business flow exceeds the first resource allocation information; and determining the resource redundancy of the first business flow based on the excess probability.

[0097] Among them, the historical resource occupancy can be the cloud network resources occupied by the first business flow at a historical moment.

[0098] Among them, the actual resource occupancy can be the cloud network resources actually occupied by the first business flow.

[0099] In the specific implementation, the cloud network resources occupied by the first business flow at a historical moment can be obtained, and based on this, the probability that the cloud network resources actually occupied by the first business flow exceeds the first resource allocation information determined in step S120 can be determined to obtain the excess probability, and based on the size of the excess probability, the resource redundancy corresponding to the first resource allocation information can be determined.

[0100] For example, according to Figure 2 After determining the bandwidth and resource allocation for key flows, combined with cloud network resources, the overage probability of key flows can be analyzed based on historical traffic, bandwidth, and resource allocation data, or queuing theory. Assume that for flow AB in path 1: AB has a 90% overage probability, so it is configured with larger redundant bandwidth and redundant storage resources, 40 Gbps and 20 GB, respectively. Flow AD has a lower overage probability of 10%, so it is configured with smaller redundant bandwidth and redundant storage resources, 5 Gbps and 5 GB, respectively. Flow AE has a moderate overage probability of 50%, so it is configured with a moderate redundant bandwidth of 20 Gbps.

[0101] In this embodiment, the excess probability of the first business flow is determined based on the historical resource occupancy of the first business flow; the excess probability is used to characterize the probability that the actual resource occupancy of the first business flow exceeds the first resource allocation information; based on the excess probability, the resource redundancy of the first business flow is determined, and the resource redundancy can be reasonably determined based on the probability that the actual resources of the key flow exceed the allocated resources to ensure reliable transmission of the key flow.

[0102] In one embodiment, the above-mentioned step S130 may specifically include: determining a second candidate path that matches the second business demand of the second business flow based on the remaining resources; determining a second target path from the second candidate paths, and determining resource allocation information corresponding to the second target path as the second resource allocation information.

[0103] The second service requirement may be a normal flow service requirement, the second candidate path may be a normal flow candidate path, and the second target path may be a normal flow target path.

[0104] In a specific implementation, a second candidate path can be determined from the remaining resources based on the second business demand of the second business flow, a multi-commodity flow model can be established for the second candidate path, and all second candidate paths and their corresponding cloud network resources can be optimized through the multi-commodity flow model to obtain the second target path in the second candidate path, as well as the network link bandwidth and cloud storage space corresponding to the second target path, and the second target path and its corresponding network link bandwidth and cloud storage space are determined as the second resource allocation information.

[0105] For example, according to Figure 2, we can select routes for common flows, and the candidate paths for flow FB are FEB, FAB, and FDEB; for flow FC, FDC, FABC, and FEBC; and for flow FD, FD, FED, and FED. Based on the remaining cloud network resources, assuming that the overall optimization goal for common flows is to minimize cost, we can establish a multi-commodity flow linear scaling model. The variables are the bandwidths of the candidate paths for common flows. Constraints include link bandwidth constraints: the traffic carrying common flows on each link segment must not exceed the link capacity; and cloud node resource constraints: the sum of the resources required to carry common flows on each cloud node must not exceed the cloud resource capacity of that node. Based on this linear programming model, the final path, bandwidth, and resource allocation of common flows are obtained: the final path of flow FB is FEB, with a bandwidth allocation of 30 Gbps and 20 GB of cloud resources allocated to cloud node B; the final path of flow FC is FDC, with a bandwidth allocation of 30 Gbps; the final path of flow FD is FD, with a bandwidth allocation of 30 Gbps and 20 GB of cloud resources allocated to cloud node D.

[0106] In this embodiment, a second candidate path that matches the second business demand of the second business flow is determined based on the remaining resources; a second target path is determined from the second candidate path, and resource allocation information corresponding to the second target path is determined as the second resource allocation information. Resource allocation can be performed for ordinary flows based on the remaining resources of the cloud network resources, thereby improving the utilization rate of the cloud network resources.

[0107] In one embodiment, the above-mentioned step S140 may specifically include: determining a new first business demand in response to an update operation for the first business demand; performing a first adjustment operation on the first resource allocation information and a second adjustment operation on the second resource allocation information according to the new first business demand; the first adjustment operation is in opposite directions to the second adjustment operation; and allocating cloud network resources to the first business flow and the second business flow according to the adjusted first resource allocation information and the adjusted second resource allocation information.

[0108] The first adjustment operation may be an operation to increase or decrease the network link bandwidth and / or cloud storage space, and correspondingly, the second adjustment operation may be an operation to decrease or increase the network link bandwidth and / or cloud storage space.

[0109] In a specific implementation, when the first business demand changes, a new first business demand is obtained, and the first resource allocation information is increased or decreased according to the new first business demand to obtain the adjusted first resource allocation information. Correspondingly, the second resource allocation information is reduced or increased to obtain the adjusted second resource allocation information. Thereafter, cloud network resources can be allocated to the first business flow according to the adjusted first resource allocation information, and cloud network resources can be allocated to the second business flow according to the adjusted second resource allocation information.

[0110] In actual applications, based on the dynamic changes of actual business traffic and the operating status of the cloud network, the dual planes can dynamically adjust the traffic distribution of key flows and common flows on the basis of the optimized allocation in step S140 to ensure the reliable transmission of key flows and improve the utilization of cloud network resources. Figure 2 The demand for the critical flow AE suddenly changes, and the path AFE changes from the original 50Gbps to 80Gbps. Since it passes through the same link FE as the normal flow FB (the path is FEB), and the link bandwidth is only 100Gbps, it is necessary to squeeze 10Gbps of the bandwidth of the normal flow FB, and the normal flow FB is reduced from the original 30Gbps to 20Gbps.

[0111] In this embodiment, a new first business demand is determined by responding to an update operation for the first business demand; based on the new first business demand, a first adjustment operation is performed on the first resource allocation information, and a second adjustment operation is performed on the second resource allocation information; the first adjustment operation is in opposite directions to the second adjustment operation; based on the adjusted first resource allocation information and the adjusted second resource allocation information, cloud network resources are allocated to the first business flow and the second business flow, and the cloud network resource allocation can be dynamically adjusted in real time to adapt it to the actual business demand, thereby ensuring the reliable implementation of the actual business demand.

[0112] In one embodiment, the above-mentioned steps of determining the first target path from the first candidate path and determining the resource allocation information corresponding to the first target path as the first resource allocation information may specifically include: determining a resource update model for the first business flow; the resource update model takes minimizing the total delay of the first business flow as the update target; inputting the first candidate path into the resource update model to obtain the first target path and the resource allocation information corresponding to the first target path, and using the resource allocation information as the first resource allocation information.

[0113] Among them, the resource update model can be a multi-commodity flow model.

[0114] In a specific implementation, a first candidate path can be determined based on the first business demand of the first business flow, and a resource update model can be established for the first candidate path. The optimization goal of the resource update model can be to minimize the total delay of the first business flow. By solving the resource update model, a first target path can be determined from the first candidate path, as well as the network link bandwidth and cloud storage space corresponding to the first target path. The network link bandwidth and cloud storage space corresponding to the first target path are used as the first resource allocation information.

[0115] For example, according to Figure 2 The candidate paths for the key flows are calculated as follows: Path 1: AB, Path 2: AEB, Path 3: AFEB for flow AB; Path 1: AED, Path 2: ABCD, Path 3: AFD for flow AD; and Path 1: AE, Path 2: ABE, Path 3: AFE for flow AE. Assuming the goal is to improve the quality of service for key flows, the overall optimization objective for key flows is to minimize the latency of all key flows. A multi-commodity flow linear scaling model can be established, with the bandwidth of each candidate path for the key flows as the variable and the constraints being the link capacity constraint: the traffic carrying the key flows on each link segment must not exceed the link capacity. There is also a cloud node resource capacity constraint: the sum of the resources required to carry the key flows on each cloud node must not exceed the cloud resource capacity of that node. Based on this linear programming model, the bandwidth and resource allocation of key flows are obtained: the final path of flow AB is path: AB, with an allocated bandwidth of 50 Gbps and 25 GB of storage resources allocated to node B; the final path of flow AD is path: AED, with an allocated bandwidth of 50 Gbps and 50 GB of storage resources allocated to node D; the candidate path of flow AE is path: AFE, with an allocated bandwidth of 50 Gbps, which is the first resource allocation information.

[0116] In this embodiment, the resource update model of the first business flow is determined; the resource update model takes the minimum total delay of the first business flow as the update target; the first candidate path is input into the resource update model to obtain the first target path and the resource allocation information corresponding to the first target path, and the resource allocation information is used as the first resource allocation information. The first resource allocation information can be directly determined through the resource update model to improve the efficiency of cloud network resource allocation.

[0117] In order to facilitate those skilled in the art to have a deeper understanding of the embodiments of the present application, a specific example will be used for illustration below.

[0118] This application proposes a cloud network traffic engineering optimization method and system based on differentiated traffic optimization, probabilistic assessment of redundant bandwidth and resources, and dual-plane dynamic regulation. This method and system comprehensively consider the overall status of cloud network resources in cloud network traffic engineering optimization, perform differentiated optimization for different types of traffic, and rationally design redundant bandwidth and other resources to ensure business flow SLA (Service Level Agreement) requirements and improve overall cloud network resource utilization.

[0119] The cloud network traffic engineering optimization system can be composed of the following parts: traffic differentiation module, multi-factor routing module, probability evaluation module, key flow optimization module, common flow optimization module, output module and dual-plane control module.

[0120] Among them, the input module is used to provide the abstraction of cloud network resources (network topology, link bandwidth, cloud resource distribution, etc.), business flow requirements and inter-cloud network traffic matrix; it can be obtained through interaction with related cloud resource management systems and network management systems.

[0121] The traffic differentiation module is used to differentiate flows in the cloud network based on traffic characteristics through traffic classification technology, determine the priority of their traffic, and the optimization plane on which they are located. Different optimization planes may have different overall optimization goals. Among them, the optimization plane can be a virtual plane corresponding to different optimization levels. If the critical flow optimization plane has excess traffic, it can occupy the resources of the ordinary flow plane and squeeze out the ordinary flow. Ordinary flow can only be occupied when there are redundant resources on the critical flow plane.

[0122] The key flow optimization module uses the multi-factor routing module to comprehensively consider cloud network resources and calculate Pareto-optimal paths. It can also further filter out optimal paths that meet SLA constraints from these Pareto-optimal paths. Based on the candidate paths, a multi-commodity flow model is established for all key flows, and an overall objective function is set and solved to optimize the overall traffic distribution. The overall objective function to be optimized can be minimizing the latency or cost of all key flows, maximizing overall reliability, or minimizing maximum link utilization.

[0123] The probability assessment module is used to assess the traffic distribution results and needs of key flows. Machine learning can be used to train a probability model using historical data to determine the probability of key flows exceeding the capacity limit for each link segment and each node. Queuing theory can also be used for probabilistic analysis. Based on the probabilistic assessment results, a certain amount of redundant bandwidth is set for each link segment and a certain amount of redundant resources is configured for each node to ensure critical flows. Compared to the light-load approach commonly used in traditional network capacity configuration, this capacity configuration can be more aggressive and less conservative, thereby improving cloud network resource utilization.

[0124] The Common Flow Optimization module optimizes common flows based on the remaining cloud network resources after redundant bandwidth and resources are determined and configured for critical flows. Similarly, the Multi-Factor Routing Module can be used for routing. A multi-commodity flow model is then established for all common flows, and an overall objective function is set and solved to optimize the overall traffic distribution. It should be noted that the overall objective function here can be different from that in the Critical Flow Optimization module; for example, it can be set to maximize throughput.

[0125] The output module is used to output the paths and resource allocation results of key flows and common flows, and configure them into the actual cloud network.

[0126] The dual-plane control module optimizes traffic engineering for both critical and common flows based on pre-determined requirements (given traffic and required resources). However, due to the dynamic changes in cloud network business traffic, the traffic of business flows may increase or decrease compared to the original optimized initial value, exceeding the allocated redundant bandwidth and resources. A dynamic module is needed to guide traffic in a timely manner and make adjustments based on the original optimization to protect critical flows. If a critical flow exceeds the allocated value and the configured redundant bandwidth or resources cannot meet the needs, it will be guided to the common flow plane, reducing the number of common flows passing through the same links and nodes as it, and giving priority to protecting the critical flow. Conversely, if the critical flow does not exceed the original optimized value, the common flow may occupy the allocated redundant bandwidth and resources if it exceeds the allocated value. In this way, based on the original traffic optimization, through rapid dynamic control, we strive to meet the needs of various flows while improving link and resource utilization.

[0127] Figure 3 A flow chart of the cloud network traffic engineering optimization method is provided. Figure 3 ,The cloud network traffic engineering optimization method may include the following steps:

[0128] Step S201: Obtain input from relevant systems, including relevant resource and demand data, cloud network topology, cloud network traffic matrix, etc.

[0129] Step S202: Divide the traffic into key flows and common flows.

[0130] Step S203: Invoke the multi-factor routing module to calculate candidate paths based on the key flow requirements. Simultaneously, based on the key flow requirements and the overall optimization goal, establish a multi-commodity network flow model to solve the optimal key flow flow and resource allocation.

[0131] Step S204 , based on the traffic and resource allocation results of the key flows, a probability evaluation is performed on the results of carrying the key flows on each link and each node, and redundant bandwidth and resources (key flow optimization plane) are determined based on the probability of excess.

[0132] In step S205, the resources reserved for key flow optimization are subtracted from the initial cloud network resources to obtain the remaining resources (normal flow optimization plane). Similarly, the multi-factor routing module is called to calculate candidate paths based on the normal flow demand. At the same time, based on the normal flow demand and the overall goal to be optimized, a multi-commodity network flow model is established to solve the optimal key flow traffic and resource allocation.

[0133] Step S206: Output the paths and resource allocation results of the key flows and common flows, and configure them in the cloud network.

[0134] In step S207, based on the dynamic changes of actual business traffic and the operating status of the cloud network, the dual-plane control module dynamically adjusts the traffic distribution of key flows and common flows on the basis of the original optimized allocation to ensure key flows and improve resource utilization.

[0135] according to Figure 2 The cloud network topology diagram provided assumes a 100 Gbps link bandwidth, Nodes B and D have cloud resources, and both have 1000 GB of storage resources. There are three key flows: AB, AD, and AE, and three common flows: FB, FC, and FD. The cloud network resource requirements for these key and common flows are shown in Table 1.

[0136] Cloud network traffic engineering optimization based on this cloud network topology can include the following steps:

[0137] Step 1: The traffic differentiation module uses traffic classification technology to determine key flows and common flows based on traffic characteristics. Key flows are flows AB, AD, and AE, and common flows are flows FB, FC, and FD.

[0138] Step 2: Call the multi-factor routing module to calculate candidate paths based on the requirements of the key flows. Assume that the candidate paths for flow AB are path 1: AB, path 2: AEB, and path 3: AFEB; the candidate paths for flow AD are path 1: AED, path 2: ABCD, and path 3: AFD; and the candidate paths for flow AE are path 1: AE, path 2: ABE, and path 3: AFE. Assuming that the goal here is to improve the service quality of key flows, the overall optimization goal for key flows is to minimize the latency of all key flows. A linear scale model for multi-commodity flows can be established, with the variables being the bandwidth of each candidate path for the key flows and the constraints being the link capacity constraint: the traffic carrying the key flows on each link segment must not exceed the link capacity. There are also cloud node resource capacity constraints: the sum of the resources required to carry the key flows on each cloud node must not exceed the cloud resource capacity of that node. Based on this linear programming model, the bandwidth and resource allocation of key flows are obtained: the final path of flow AB is path: AB, with an allocated bandwidth of 50 Gbps and 25 GB of cloud resources allocated to node B; the final path of flow AD is path: AED, with an allocated bandwidth of 50 Gbps and 50 GB of cloud resources allocated to node D; the candidate path of flow AE is path: AFE, with an allocated bandwidth of 50 Gbps.

[0139] In step 3, based on the bandwidth and resource allocation results from step 2 and combined with cloud network resources, the overage probability of key flows can be analyzed based on historical traffic, bandwidth, and resource allocation data, or queuing theory. Assuming that the probability of overage on path 1 for flow AB is 90%, a larger amount of redundant bandwidth and resources, 40 Gbps and 20 GB, respectively, is allocated. Flow AD has a lower probability of overage, 10%, so a smaller amount of redundant bandwidth, 5 Gbps, and resources, 5 GB, is allocated. Flow AE has a moderate probability of overage, 50%, so a moderate amount of redundant bandwidth, 20 Gbps, is allocated. This results in the total bandwidth and resources required for key flow optimization.

[0140] In step 4, the bandwidth and resources required for the key flow optimization determined in step 3 are subtracted from the original cloud network resource topology. Similarly, the multi-factor routing module is called to select routes for ordinary flows: candidate path 1 of flow FB is FEB, candidate path 2 is FAB, and candidate path 3 is FDEB; candidate path 1 of flow FC is FDC, candidate path 2 is FABC, and candidate path 3 is FEBC; candidate path 1 of flow FD is FD, candidate path 2 is FED, and candidate path 3 is FED. Based on the remaining cloud network resources, assuming that the overall optimization goal of ordinary flows is to minimize costs, a multi-commodity flow linear scale model can be established. The variables are the bandwidths of each candidate path of ordinary flows, and the constraints are link bandwidth constraints: the traffic carrying ordinary flows on each link segment does not exceed the link capacity. There are also cloud node resource constraints: the sum of the resources required to carry ordinary flows on each cloud node does not exceed the cloud resource capacity of the node. Based on this linear programming model, the final path, bandwidth, and resource allocation of common flows are obtained: the final path of flow FB is FEB, with a bandwidth allocation of 30 Gbps and 20 GB of cloud resources allocated to cloud node B; the final path of flow FC is FDC, with a bandwidth allocation of 30 Gbps; the final path of flow FD is FD, with a bandwidth allocation of 30 Gbps and 20 GB of cloud resources allocated to cloud node D.

[0141] In step 5, the paths of the key flows and common flows determined in steps 2 and 4, as well as their bandwidth allocation and resource allocation, are configured in the cloud network.

[0142] In step 6, based on the optimized allocation in step 5 and the dynamic changes in actual service traffic and the cloud network's operational status, the dual-plane control module dynamically adjusts the traffic distribution of critical flows and common flows to protect critical flows and improve resource utilization. For example, if the demand for critical flow AE suddenly changes, the path AFE will increase from 50 Gbps to 80 Gbps. Since it and common flow FB (path FEB) pass through the same link FE, which only has 100 Gbps bandwidth, the bandwidth of common flow FB needs to be squeezed by 10 Gbps, reducing the common flow's bandwidth from 30 Gbps to 20 Gbps.

[0143] Step 7: Through continuous dynamic design optimization and dynamic traffic adjustment in steps 1 to 6, the protection of key flows is improved while improving the utilization of cloud network resources.

[0144] The above-mentioned cloud network traffic engineering optimization method comprehensively considers the business flow type, performs differentiated traffic optimization for different types of business flows, introduces a probabilistic assessment method, rationally designs redundant bandwidth and capacity, and considers a dynamic adjustment mechanism to dynamically adjust traffic, thereby improving the protection of key businesses and the overall cloud network resource utilization.

[0145] In one embodiment, Figure 4 As shown, a cloud network resource allocation method is provided, which is described by taking the application of the method to a terminal as an example, and includes the following steps:

[0146] Step S301: Determine a first service flow and a second service flow from the cloud network service flows to which cloud network resources are to be allocated; the priority of the first service flow is higher than the priority of the second service flow;

[0147] Step S302: determining a first candidate path that matches a first service requirement of a first service flow based on cloud network resources;

[0148] Step S303: determining a first target path from the first candidate paths, and determining resource allocation information corresponding to the first target path as first resource allocation information;

[0149] Step S304, determining the resource redundancy of the first service flow;

[0150] Step S305: obtaining a predicted resource occupancy of the first service flow according to the first resource allocation information and the resource redundancy;

[0151] Step S306: Subtract the predicted resource occupancy from the cloud network resources to obtain the remaining resources;

[0152] Step S307: determining a second candidate path that matches the second service requirement of the second service flow based on the remaining resources;

[0153] Step S308: determining a second target path from the second candidate paths, and determining resource allocation information corresponding to the second target path as second resource allocation information;

[0154] Step S309: Allocate cloud network resources to the first service flow and the second service flow according to the first resource allocation information and the second resource allocation information.

[0155] In the specific implementation, the cloud network business flow can be distinguished to obtain key flows and ordinary flows. The cloud network resources are first allocated to the key flow to obtain the first candidate path of the key flow, and the first target path and the first resource allocation information corresponding to the first target path are determined from the first candidate path through the resource update model, and the resource redundancy of the key flow is determined. The predicted resource occupancy of the key flow is obtained according to the sum of the first resource allocation information and the resource redundancy, and the predicted resource occupancy is removed from the cloud network resources to obtain the remaining resources. The remaining resources are then allocated to the ordinary flow to obtain the second candidate path of the ordinary flow. The second target path and the second resource allocation information corresponding to the second target path are determined from the second candidate path through the resource update model. Finally, the cloud network resources can be allocated to the key flow according to the first resource allocation system information, and the remaining resources of the cloud network resources can be allocated to the ordinary flow according to the second resource allocation information.

[0156] The above-mentioned cloud network resource allocation method determines the first business flow and the second business flow from the cloud network business flow to which the cloud network resources are to be allocated, determines a first candidate path that matches the first business demand of the first business flow based on the cloud network resources, determines a first target path from the first candidate path, and determines the resource allocation information corresponding to the first target path as the first resource allocation information, determines the resource redundancy of the first business flow, obtains the predicted resource occupancy of the first business flow based on the first resource allocation information and the resource redundancy, removes the predicted resource occupancy from the cloud network resources to obtain the remaining resources, determines a second candidate path that matches the second business demand of the second business flow based on the remaining resources, determines a second target path from the second candidate path, and determines the resource allocation information corresponding to the second target path as the second resource allocation information, allocates cloud network resources to the first business flow and the second business flow based on the first resource allocation information and the second resource allocation information, thereby improving the utilization of cloud network resources while ensuring the reliable transmission of key flows.

[0157] It should be understood that, although the various steps in the flowcharts involved in the various embodiments described above are displayed in sequence according to the instructions of the arrows, these steps are not necessarily executed in sequence in the order indicated by the arrows. Unless otherwise specified herein, there is no strict order restriction on the execution of these steps, and these steps can be executed in other orders. Moreover, at least a portion of the steps in the flowcharts involved in the various embodiments described above can include multiple steps or multiple stages, and these steps or stages are not necessarily executed and completed at the same time, but can be executed at different times, and the execution order of these steps or stages is not necessarily to be carried out in sequence, but can be executed in turn or alternately with other steps or at least a portion of steps or stages in other steps.

[0158] Based on the same inventive concept, embodiments of the present application also provide a cloud network resource allocation device for implementing the cloud network resource allocation method involved above. The implementation solution provided by this device is similar to the implementation solution described in the above method. Therefore, the specific limitations in one or more cloud network resource allocation device embodiments provided below can be found in the above-mentioned limitations on the cloud network resource allocation method and will not be repeated here.

[0159] In one embodiment, Figure 5 As shown, a cloud network resource allocation device is provided, including: a service classification module 410, a first information module 420, a second information module 430 and a resource allocation module 440, wherein:

[0160] The service classification module 410 is configured to determine a first service flow and a second service flow from the cloud network service flows to which cloud network resources are to be allocated; the priority of the first service flow is higher than the priority of the second service flow;

[0161] A first information module 420 is configured to determine first resource allocation information for allocating the cloud network resource to the first service flow, and determine remaining resources of the cloud network resource based on the first resource allocation information;

[0162] A second information module 430 is configured to determine second resource allocation information for allocating the remaining resources to the second service flow;

[0163] The resource allocation module 440 is used to allocate the cloud network resources to the first business flow and the second business flow according to the first resource allocation information and the second resource allocation information.

[0164] In one embodiment, the above-mentioned first information module 420 is also used to determine a first candidate path that matches the first business demand of the first business flow based on the cloud network resources; determine a first target path from the first candidate path, and determine the resource allocation information corresponding to the first target path as the first resource allocation information.

[0165] In one embodiment, the above-mentioned first information module 420 is also used to determine the resource redundancy of the first business flow; obtain the predicted resource occupancy of the first business flow based on the first resource allocation information and the resource redundancy; remove the predicted resource occupancy from the cloud network resources to obtain the remaining resources.

[0166] In one embodiment, the above-mentioned first information module 420 is also used to determine the excess probability of the first business flow based on the historical resource occupancy of the first business flow; the excess probability is used to characterize the probability that the actual resource occupancy of the first business flow exceeds the first resource allocation information; based on the excess probability, the resource redundancy of the first business flow is determined.

[0167] In one embodiment, the above-mentioned second information module 430 is also used to determine a second candidate path that matches the second business demand of the second business flow based on the remaining resources; determine a second target path from the second candidate path, and determine the resource allocation information corresponding to the second target path as the second resource allocation information.

[0168] In one embodiment, the above-mentioned resource allocation module 440 is also used to determine a new first business demand in response to an update operation for the first business demand; perform a first adjustment operation on the first resource allocation information and a second adjustment operation on the second resource allocation information according to the new first business demand; the first adjustment operation is opposite to the second adjustment operation; and allocate the cloud network resources to the first business flow and the second business flow according to the adjusted first resource allocation information and the adjusted second resource allocation information.

[0169] In one embodiment, the above-mentioned first information module 420 is also used to determine the resource update model of the first business flow; the resource update model takes the minimum total delay of the first business flow as the update target; the first candidate path is input into the resource update model to obtain the first target path, and the resource allocation information corresponding to the first target path, and the resource allocation information is used as the first resource allocation information.

[0170] Each module in the aforementioned cloud network resource allocation device may be implemented in whole or in part through software, hardware, or a combination thereof. Each module may be embedded in or independent of a processor in a computer device in hardware form, or may be stored in a computer device memory in software form, so that the processor can call and execute the corresponding operations of each module.

[0171] In one embodiment, a computer device is provided. The computer device may be a terminal, and its internal structure diagram may be as follows: Figure 6As shown. The computer device includes a processor, a memory, an input / output interface, a communication interface, a display unit, and an input device. The processor, memory, and input / output interface are connected via a system bus, and the communication interface, display unit, and input device are connected to the system bus via the input / output interface. The processor of the computer device is used to provide computing and control capabilities. The memory of the computer device includes a non-volatile storage medium and an internal memory. The non-volatile storage medium stores an operating system and a computer program. The internal memory provides an environment for the operation of the operating system and computer program in the non-volatile storage medium. The input / output interface of the computer device is used to exchange information between the processor and an external device. The communication interface of the computer device is used to communicate with an external terminal in a wired or wireless manner. The wireless manner can be achieved through WIFI, a mobile cellular network, NFC (near field communication) or other technologies. When the computer program is executed by the processor, a cloud network resource allocation method is implemented. The display unit of the computer device is used to form a visually visible image, and can be a display screen, a projection device, or a virtual reality imaging device. The display screen can be a liquid crystal display screen or an electronic ink display screen, and the input device of the computer device can be a touch layer covering the display screen, or a button, trackball or touchpad set on the computer device casing, or an external keyboard, touchpad or mouse.

[0172] Those skilled in the art will understand that Figure 6 The structure shown in the figure is only a block diagram of a part of the structure related to the solution of the present application, and does not constitute a limitation on the computer device to which the solution of the present application is applied. The specific computer device may include more or fewer components than shown in the figure, or combine certain components, or have a different component arrangement.

[0173] In one embodiment, a computer device is further provided, including a memory and a processor. The memory stores a computer program, and the processor implements the steps in the above method embodiments when executing the computer program.

[0174] In one embodiment, a computer-readable storage medium is provided, on which a computer program is stored. When the computer program is executed by a processor, the steps in the above-mentioned method embodiments are implemented.

[0175] In one embodiment, a computer program product is provided, including a computer program, which implements the steps in the above method embodiments when executed by a processor.

[0176] 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, stored data, displayed data, etc.) involved in this application are all information and data authorized by the user or fully authorized by all parties, and the collection, use and processing of relevant data must comply with the relevant laws, regulations and standards of relevant countries and regions.

[0177] Those skilled in the art will appreciate that all or part of the processes in the above-mentioned embodiment methods can be implemented by instructing the relevant hardware through a computer program, and the computer program can be stored in a non-volatile computer-readable storage medium. When the computer program is executed, it can include the processes of the embodiments of the above-mentioned methods. Among them, any reference to memory, database or other media used in the embodiments provided in this application may include at least one of non-volatile and volatile memory. Non-volatile memory may include read-only memory (ROM), magnetic tape, floppy disk, flash memory, optical memory, high-density embedded non-volatile memory, resistive random access memory (ReRAM), magnetic random access memory (MRAM), ferroelectric random access memory (FRAM), phase change memory (PCM), graphene memory, etc. Volatile memory may include random access memory (RAM) or external cache memory, etc. By way of illustration and not limitation, RAM can be in various forms, such as static random access memory (SRAM) or dynamic random access memory (DRAM). The database involved in the various embodiments provided herein may include at least one of a relational database and a non-relational database. Non-relational databases may include, but are not limited to, distributed databases based on blockchains. The processor involved in the various embodiments provided herein may be, but are not limited to, a general-purpose processor, a central processing unit, a graphics processing unit, a digital signal processor, a programmable logic unit, a data processing logic unit based on quantum computing, and the like.

[0178] The technical features of the above embodiments can be combined arbitrarily. To make the description concise, not all possible combinations of the technical features in the above embodiments are described. However, as long as there is no contradiction in the combination of these technical features, they should be considered to be within the scope of this specification.

[0179] The above-described embodiments merely represent several implementation methods of the present application. While the descriptions are relatively specific and detailed, they should not be construed as limiting the scope of the present application. It should be noted that a person of ordinary skill in the art may make various modifications and improvements without departing from the spirit of the present application, and these modifications and improvements fall within the scope of protection of the present application. Therefore, the scope of protection of the present application shall be determined by the appended claims.

Claims

1. A cloud network resource allocation method, characterized in that: The method comprises: Determining a first service flow and a second service flow from the cloud network service flows to which cloud network resources are to be allocated; the priority of the first service flow is higher than the priority of the second service flow; Determining first resource allocation information for allocating the cloud network resources to the first service flow, determining an excess probability of the first service flow, and determining remaining resources of the cloud network resources based on the first resource allocation information and the excess probability; the excess probability is used to represent a probability that actual resource usage of the first service flow exceeds the first resource allocation information; Determine second resource allocation information for allocating the remaining resources to the second service flow; The cloud network resources are allocated to the first service flow and the second service flow according to the first resource allocation information and the second resource allocation information.

2. The method according to claim 1, characterized in that The determining first resource allocation information for allocating the cloud network resource to the first service flow includes: Determining, based on the cloud network resources, a first candidate path that matches a first service requirement of the first service flow; A first target path is determined from the first candidate paths, and resource allocation information corresponding to the first target path is determined as the first resource allocation information.

3. The method according to claim 2, characterized in that The determining the remaining resources of the cloud network resources according to the first resource allocation information and the excess probability includes: determining a resource redundancy amount of the first service flow according to the excess probability; Obtaining a predicted resource occupancy of the first service flow according to the first resource allocation information and the resource redundancy; The predicted resource occupancy is removed from the cloud network resources to obtain the remaining resources.

4. The method according to claim 1, wherein The determining the excess probability of the first service flow includes: The excess probability of the first service flow is determined according to the historical resource occupancy of the first service flow.

5. The method according to claim 2, characterized in that The determining second resource allocation information for allocating the remaining resources to the second service flow includes: determining, based on the remaining resources, a second candidate path that matches a second service requirement of the second service flow; A second target path is determined from the second candidate paths, and resource allocation information corresponding to the second target path is determined as the second resource allocation information.

6. The method according to claim 5, characterized in that The allocating the cloud network resources to the first service flow and the second service flow according to the first resource allocation information and the second resource allocation information includes: In response to the update operation on the first business requirement, determining a new first business requirement; performing a first adjustment operation on the first resource allocation information and a second adjustment operation on the second resource allocation information according to the new first service requirement; wherein the first adjustment operation is in opposite directions to the second adjustment operation; The cloud network resources are allocated to the first business flow and the second business flow according to the adjusted first resource allocation information and the adjusted second resource allocation information.

7. The method according to claim 2, characterized in that The determining of a first target path from the first candidate paths and determining resource allocation information corresponding to the first target path as the first resource allocation information includes: Determining a resource update model for the first service flow; wherein the resource update model takes minimizing the total delay of the first service flow as an update goal; The first candidate path is input into the resource update model to obtain the first target path and the resource allocation information corresponding to the first target path, and the resource allocation information is used as the first resource allocation information.

8. A cloud network resource allocation device, characterized in that: The device comprises: A service classification module, configured to determine a first service flow and a second service flow from cloud network service flows to which cloud network resources are to be allocated; the priority of the first service flow is higher than the priority of the second service flow; a first information module, configured to determine first resource allocation information for allocating the cloud network resources to the first service flow, determine an excess probability of the first service flow, and determine remaining resources of the cloud network resources based on the first resource allocation information and the excess probability; the excess probability is used to represent a probability that actual resource usage of the first service flow exceeds the first resource allocation information; A second information module is used to determine second resource allocation information for allocating the remaining resources to the second service flow; A resource allocation module is used to allocate the cloud network resources to the first business flow and the second business flow according to the first resource allocation information and the second resource allocation information.

9. A computer device comprising a memory and a processor, wherein the memory stores a computer program, wherein: When the processor executes the computer program, the steps of the method according to any one of claims 1 to 7 are implemented.

10. A computer-readable storage medium having a computer program stored thereon, characterized in that: When the computer program is executed by a processor, the steps of the method according to any one of claims 1 to 7 are implemented.

11. A computer program product comprising a computer program, characterized in that When the computer program is executed by a processor, the steps of the method according to any one of claims 1 to 7 are implemented.

Images