Resource allocation methods, cloud computing platforms, and computing device clusters
By unifying the management and scheduling of cloud resources in different regions through a cloud management platform, the problem of underutilization of resources in existing technologies has been solved, enabling cross-regional resource sharing and optimized data flow transmission, thereby improving user experience and resource utilization.
Patent Information
- Application Number
- CN202510631798.8
- Authority / Receiving Office
- CN · China
- Patent Type
- Applications(China)
- Current Assignee / Owner
- Filing Date
- 2025-05-15
- Publication Date
- 2026-11-17
AI Technical Summary
In existing technologies, cloud resource allocation is based on regions, which leads to underutilization of resources, failure to achieve user sharing and maximum resource utilization in different regions, non-reusability of network resources, waste of computing resources, poor disaster recovery reliability, inability to switch across regions, and inflexible resource expansion.
The cloud management platform provides unified management of cloud resources distributed across different regions. Based on the tenant's resource subscription information, it allocates cloud resources to multiple target regions for the tenant, supports cross-regional resource scheduling and site access, and utilizes CDN acceleration nodes to optimize data flow transmission, thereby achieving unified allocation and scheduling of resources.
It enables full utilization of cloud resources, improves user experience, reduces network bandwidth consumption, enhances resource flexibility and reliability, and supports cross-regional user sharing and data stream transmission efficiency.
Smart Images

Figure CN123123340A_ABST
Abstract
Description
Technical Field
[0001] This application relates to the field of cloud computing technology, and in particular to a resource allocation method, a cloud computing platform, and a cluster of computing devices. Background Technology
[0002] As the business scenarios and volume of cloud services increase, the number of cloud services that rely on large-scale computing power also increases, and the requirements for computing and network resources to support cloud services also increase accordingly.
[0003] Currently, related technologies allocate and schedule cloud resources based on regions. This means that the user and the cloud resource's site belong to the same region. In terms of network connectivity, users access sites located in the same region via the nearest public network or dedicated line, thereby using cloud services running through cloud resources.
[0004] However, allocating and scheduling resources according to regions cannot achieve full utilization of resources. Summary of the Invention
[0005] This application provides a resource allocation method, a cloud computing platform, and a computing device cluster, which can realize the unified allocation and scheduling of cloud resources in multiple regions, thereby achieving full utilization of resources.
[0006] In a first aspect, this application provides a resource allocation method applied to a cloud computing platform, the cloud computing platform including a cloud management platform for managing cloud resources distributed in different regions, the method comprising:
[0007] The cloud management platform receives resource subscription information from tenants, which is used to indicate multiple regions;
[0008] The cloud management platform allocates cloud resources in multiple target regions to tenants based on resource subscription information. These cloud resources are used to provide cloud services to tenants.
[0009] Among them, multiple target areas are the same as or partially the same as multiple areas indicated by the resource subscription information, or multiple target areas are different from multiple areas indicated by the resource subscription information.
[0010] According to this application, a cloud management platform can manage cloud resources distributed across different regions. Thus, the cloud management platform can uniformly allocate and schedule cloud resources across different regions. When a tenant needs to subscribe to resources, the tenant's resource subscription information can indicate multiple regions. Because the cloud management platform can uniformly allocate and schedule cloud resources across different regions, it can allocate cloud resources in multiple target regions to the tenant based on the tenant's resource subscription information. These cloud resources are then used to provide cloud services to the tenant, thereby achieving full utilization of resources.
[0011] In one possible implementation, the resource subscription information includes the identifiers of multiple regions and the resource specifications for each region.
[0012] In this way, the cloud management platform can allocate resources to tenants based on the region identifier and resource specifications in the resource information, thereby achieving full utilization of resources.
[0013] In one possible implementation, a cloud management platform is used to manage sites distributed across different regions, the sites are used to provide cloud resources, and tenants are used to manage users. The method further includes:
[0014] The cloud management platform obtains the user's access request;
[0015] In response to the access request, the cloud management platform determines the first resource usage information of the first site. The first site is a site within multiple target areas, and the first resource usage information is used to indicate the resource usage status of the site.
[0016] The cloud management platform determines the target site from among the multiple primary sites based on their respective primary resource usage information.
[0017] The cloud management platform sends the gateway address of the target site to the client, enabling the client to access the target site based on the gateway address.
[0018] In this way, the cloud management platform can manage sites in different regions uniformly. When a user needs to access a site, the cloud management platform can obtain the resource usage of each site and, based on that usage, allocate sites that the user can access, thus achieving full utilization of cloud resources.
[0019] In one possible implementation, the resource subscription information includes a first policy for user access to the site, which indicates the policy for allocating the site to the user. The cloud management platform determines the target site from the multiple first sites based on the first resource usage information of each of the multiple first sites, including:
[0020] The cloud management platform determines the target site from among the multiple primary sites based on their respective primary resource usage information and primary policies.
[0021] In this way, tenants can manage users. Therefore, when subscribing to resources, tenants can deploy policies for user access to sites. When a user needs to access a site, the system can allocate sites that the user can access based on the resource usage of sites in multiple target areas and the user's site access policies, thereby improving user experience and making full use of resources.
[0022] In one possible implementation, the resource subscription information includes a second policy for the user's use of the resource, which indicates the restrictions on the user's use of the resource. The access request also includes user information, which indicates the user's business situation. The method further includes:
[0023] The cloud management platform receives business requests from clients, which are used to indicate the target cloud service requested by the user.
[0024] The cloud management platform determines the target server from multiple servers at the target site based on the first site's first resource usage information, user information, and second policy. The target server is used to provide the target cloud service to the user.
[0025] In this way, when subscribing to resources, tenants can deploy restrictions on user resource usage, i.e., a second strategy, such as the maximum number of users using the resources, the maximum amount of memory a user can use, and the maximum number of Central Processing Units (CPUs) a user can use. Based on the site's resource usage information and the second strategy, target servers providing the desired cloud services are allocated to users, satisfying tenant requirements while ensuring full resource utilization.
[0026] In one possible implementation, the resource subscription information further includes a first resource scaling policy for the first site, which indicates a policy for adjusting the resources used by the cloud service. The first resource usage information includes second resource usage information, which indicates the resource usage of different cloud services. The method includes:
[0027] The cloud management platform adjusts the resources of the target cloud service based on the second resource usage information and the first resource scaling strategy;
[0028] Adjusting user resources includes expanding or shrinking user resources.
[0029] In this way, the cloud management platform can adjust the resources used by the target cloud service based on the tenant's resource scaling strategy, thereby achieving full utilization of resources and improving user experience.
[0030] In one possible implementation, the cloud computing platform also includes a business management and control platform distributed across different regions, and the method further includes:
[0031] The target business management platform receives business requests from clients, which also include user identity information. The target business management platform is the business management platform within the region where the target site is located.
[0032] The target business management platform authenticates users based on their identity information;
[0033] The cloud management platform receives business requests from clients, including:
[0034] Once user authentication is successful, the target business management platform sends a business request to the cloud management platform.
[0035] In this way, the business management platform of a certain region can authenticate the identity of users accessing the sites in that region, thereby preventing resources from being misused.
[0036] In one possible implementation, the cloud management platform manages the cloud services running on the second site through a business control platform, and the region to which the second site belongs is the same as the region to which the business control platform belongs.
[0037] In this way, the cloud management platform manages the cloud services running on the site through the business control platform, thereby achieving full utilization of resources.
[0038] In one possible implementation, the method further includes:
[0039] The target business management platform determines the Content Delivery Network (CDN) acceleration node based on the Internet Protocol (IP) address of the business request;
[0040] The target business management platform sends the address information of the CDN acceleration node to the client so that the client can establish a communication connection with the CDN acceleration node.
[0041] In this way, a communication channel can be established between the client, CDN acceleration nodes, and the server, reducing the bandwidth occupied by the communication channel between the client and the server.
[0042] In one possible implementation, the method further includes:
[0043] The target server receives input events from the first client, which are used to indicate events that perform operations on the target cloud service;
[0044] The target server responds to the input event and generates the first data stream;
[0045] The target server sends a first data stream to the first CDN acceleration node, so that the first CDN acceleration node sends a first data stream to the first client.
[0046] In this way, the client transmits input events directly to the server, while the data stream is transmitted to the client through CDN acceleration nodes, which improves the transmission efficiency of the data stream and reduces the bandwidth consumption of the communication channel between the client and the server.
[0047] In one possible implementation, the method further includes:
[0048] The target server sends a first data stream to the second CDN acceleration node so that the second client can obtain the first data stream from the second CDN acceleration node. The second client is a client that accesses the target cloud service, and the second client is a different client from the first client.
[0049] In this way, when the input event of the first client generates a data stream, the server can cache it to different CDN acceleration nodes, so that the second client can quickly obtain the data stream corresponding to the first client from the CDN acceleration node, thereby improving the data stream transmission efficiency in shared and collaborative scenarios.
[0050] In one possible implementation, the regions to which the first client belongs are different from the regions to which the second client belongs.
[0051] In this way, when multiple clients from different regions access the same cloud service, the data stream of one client can be cached in CDN acceleration nodes in different regions, thereby enabling it to be quickly sent to other clients and improving the data stream transmission speed.
[0052] In one possible implementation, the target site is located in a region different from the user's region.
[0053] In this way, the regions where users can access websites can be different from the regions where users are located, thereby enabling cloud resources to be shared across different regions.
[0054] Secondly, this application provides a cloud computing platform, which includes a cloud management platform for managing cloud resources distributed in different regions;
[0055] The cloud management platform is also used to receive resource subscription information from tenants. This resource subscription information is used to indicate multiple regions, and based on the resource subscription information, cloud resources are allocated to tenants in multiple target regions. These cloud resources are used to provide cloud services to tenants.
[0056] In one possible implementation, the resource subscription information includes the identifiers of multiple regions and the resource specifications for each region.
[0057] In one possible implementation, a cloud management platform is used to manage sites distributed in different regions, the sites are used to provide cloud resources, the resource subscription information includes a first policy for users to access the sites, the first policy is used to indicate the policy for allocating sites to users, and tenants are used to manage users.
[0058] The cloud management platform is also used to obtain user access requests;
[0059] The cloud management platform is also used to respond to access requests and determine the first resource usage information of the first site. The first site is a site in multiple target areas, and the first resource usage information is used to indicate the resource usage of the site.
[0060] The cloud management platform is also used to determine the target site from multiple first sites based on their respective first resource usage information and first policies, and send the gateway address of the target site to the client so that the client can access the target site based on the gateway address of the target site.
[0061] In one possible implementation, the resource subscription information includes a second strategy for the user to use the resource, which indicates the restrictions on the user's use of the resource, and the access request includes the user's user information, which indicates the user's business situation.
[0062] The cloud management platform is also used to obtain business requests from clients, which indicate the target cloud service requested by the user.
[0063] The cloud management platform is also used to determine the target server from multiple servers of the target site based on the first resource usage information, user information, and second policy of the first site. The target server is used to provide the target cloud service to the user.
[0064] In one possible implementation, the resource subscription information also includes a first resource scaling policy of the first site, which is used to indicate the policy for adjusting the resources used by the cloud service. The first resource usage information includes second resource usage information, which is used to indicate the resource usage of different cloud services.
[0065] The cloud management platform is also used to adjust the resources of the target cloud service based on the second resource usage information and the first resource scaling strategy;
[0066] Adjusting user resources includes expanding or shrinking user resources.
[0067] In one possible implementation, the cloud computing platform also includes a business management and control platform distributed across different regions;
[0068] The target business management platform is used to receive business requests from clients. These requests also include the user's identity information. The target business management platform is the business management platform within the region where the target site is located.
[0069] The target business management platform is also used to authenticate users based on their identity information, and when the user authentication is successful, the target business management platform sends a business request to the cloud management platform.
[0070] In one possible implementation, the cloud management platform manages the cloud services running on the second site through a business control platform, and the region to which the second site belongs is the same as the region to which the business control platform belongs.
[0071] In one possible implementation, the target business management platform is also used to determine the Content Delivery Network (CDN) acceleration node based on the Internet Protocol (IP) address of the business request, and send the CDN acceleration node's address information to the client so that the client can establish a communication connection with the CDN acceleration node.
[0072] In one possible implementation, the target server is also used to receive input events from the first client, which are used to indicate events that perform operations on the target cloud service;
[0073] The target server is also used to respond to the input event, generate a first data stream, and send the first data stream to the first CDN acceleration node, so that the first CDN acceleration node sends the first data stream to the first client.
[0074] In one possible implementation, the target server is also used to send a first data stream to a second CDN acceleration node for a second client to obtain the first data stream from the second CDN acceleration node. The second client is a client that accesses the target cloud service, and the second client is a different client from the first client.
[0075] In one possible implementation, the regions to which the first client belongs are different from the regions to which the second client belongs.
[0076] In one possible implementation, the target site is located in a region different from the user's region.
[0077] Thirdly, this application provides a cloud computing platform, including:
[0078] Business management platforms located in different regions;
[0079] A cloud management platform is used to manage cloud resources distributed in different regions. The cloud resources are communicatively connected to a business management platform located in the region to which the cloud resources belong, so that the cloud computing platform can implement the method as described in any one of claims 1-11.
[0080] Fourthly, this application also provides a computing device cluster. The computing device cluster includes at least one computing device, each computing device including a processor and a memory. The processor of the at least one computing device is used to execute instructions stored in the memory of the at least one computing device, causing the computing device cluster to perform the method provided in the first aspect.
[0081] Fifthly, this application also provides a computer-readable storage medium storing instructions that, when executed on a computer, cause the computer to implement the method provided in the first aspect.
[0082] Sixthly, this application also provides a computer program product containing instructions that, when run on a computer, cause the computer to implement the method provided in the first aspect.
[0083] Any of the devices, computer storage media, or computer program products provided above are used to execute the methods provided above. Therefore, the beneficial effects they can achieve can be referred to the beneficial effects of the corresponding solutions in the corresponding methods provided above, and will not be repeated here. Attached Figure Description
[0084] Figure 1 This is a schematic diagram of the system architecture of a cloud computing platform provided in an embodiment of this application;
[0085] Figure 2 This is a schematic diagram of the structure of a cloud management platform in a cloud computing platform provided in an embodiment of this application;
[0086] Figure 3 This is a flowchart illustrating a site registration method provided in an embodiment of this application;
[0087] Figure 4 This is a flowchart illustrating a resource allocation method provided in an embodiment of this application;
[0088] Figure 5 This is a schematic diagram of a tenant subscription resource interface provided in an embodiment of this application;
[0089] Figure 6 This is a schematic diagram of another interface for tenant subscription of resources provided in an embodiment of this application;
[0090] Figure 7 This is a schematic diagram of another interface for tenant subscription of resources provided in an embodiment of this application;
[0091] Figure 8 This is a flowchart illustrating a site access method provided in an embodiment of this application;
[0092] Figure 9 This is a schematic diagram of an interface for requesting a cloud application provided in an embodiment of this application;
[0093] Figure 10 This is a flowchart illustrating a site resource scaling method provided in an embodiment of this application;
[0094] Figure 11 This is a flowchart illustrating a resource scaling method for cloud services provided in an embodiment of this application;
[0095] Figure 12 This is a schematic diagram of the structure of a cloud computing platform provided in an embodiment of this application;
[0096] Figure 13 This is a schematic diagram of the structure of a computing device provided in an embodiment of this application;
[0097] Figure 14 This is a schematic diagram of the structure of a computing device cluster provided in an embodiment of this application;
[0098] Figure 15 This is a schematic diagram of another computing device cluster structure provided in an embodiment of this application. Detailed Implementation
[0099] The technical solutions of the embodiments of this application will be described below with reference to the accompanying drawings. Obviously, the described embodiments are only some embodiments of this application, and not all embodiments. Based on the embodiments of this application, all other embodiments obtained by those of ordinary skill in the art without creative effort are within the scope of protection of this application.
[0100] Before introducing the embodiments of this application, the terms mentioned in the embodiments of this application will be explained below.
[0101] Cloud desktop, also known as desktop virtualization or cloud PC, is a technology that replaces traditional computers. Specifically, a cloud desktop centralizes computing and storage resources on a remote server, which then projects the interface of a remote virtual desktop onto the terminal device via a remote desktop protocol. A cloud desktop consists of a cloud desktop client and a cloud desktop server. The cloud desktop client can be deployed on the terminal device, while the cloud desktop server is deployed on a remote server, which can be called a cloud desktop server.
[0102] Cloud applications are applications deployed in the cloud. Users can use these applications via the internet or local area network without installing them on their local devices. Cloud applications allow users to access applications anytime, anywhere via their devices, improving work efficiency and flexibility. Cloud applications also include cloud application clients and cloud application servers. Cloud application clients can be deployed on terminal devices, while cloud application servers are deployed on remote servers, which can be called cloud application servers.
[0103] A region represents the location of a cloud data center within a cloud platform, and a region may include at least one cloud data center. In the embodiments of this application, cloud data centers in different regions can share cloud resources. Cloud resources may include, but are not limited to, computing resources, storage resources, load balancer services, elastic internet protocol (EIP) services, cloud backup services, and host security services.
[0104] A tenant is a top-level object used to manage cloud services and / or cloud resources. Tenants register a tenant account and set a tenant password on the cloud management platform through a local client (such as a browser). The local client remotely logs into the cloud management platform through the tenant account and set the tenant password. The cloud management platform provides a configuration interface or API for tenants to configure and use cloud services, including cloud desktops and cloud applications.
[0105] A site refers to a physical geographical location and its associated infrastructure, specifically designed for the centralized deployment of hardware resources (such as servers, storage devices, network equipment, etc.) required for cloud computing. In the embodiments of this application, sites include central sites, edge sites, tenant data centers, and third-party leased sites. A central site refers to a site established by the cloud service provider. An edge site refers to a site located close to the tenant; a third-party leased site refers to a site leased from a third party.
[0106] An agent typically refers to a self-running program or service that can perform specific tasks on behalf of a user, system, or other program. In this embodiment, the agent is deployed on the site to obtain real-time updates on resource changes within the site.
[0107] Resources: Resources include computing resources, storage resources, network resources, etc. Among them, resources can be divided into different dimensions, such as site-level resources and server-level resources.
[0108] As the business scenarios and volume of cloud services increase, the number of cloud services that rely on large-scale computing power also increases, and the requirements for computing and network resources to support cloud services also increase accordingly.
[0109] Currently, cloud resource allocation in related technologies is based on regional allocation and scheduling. This means that the tenant and the cloud resource's hosting cloud platform belong to the same region. In terms of network, tenants access the cloud platform via the nearest public network or dedicated line to use cloud services.
[0110] However, due to the limitation that tenants and cloud computing platforms are located within the same region, cloud services can only conduct one-to-one data interactions. For shared business scenarios (such as shared online editing applications, online meetings, live streaming, etc.), cloud services cannot be shared with users in different regions, requiring one-to-one data stream distribution, resulting in the inability to reuse network resources. Moreover, users in different regions cannot share the same cloud service or achieve optimal scheduling, and computing power cannot be maximized, leading to a waste of computing resources.
[0111] Moreover, when users access the area without any services, and the current service command events and media streams are not separated and share the same network link, signaling and media streams need to go through the public network to remote ends or cross international lines to connect to services and obtain data. The long network lines and network instability result in obvious latency and stuttering, and the user experience cannot be optimal.
[0112] Furthermore, cloud desktop and cloud application services cannot switch across regions for disaster recovery reliability, which results in the inability to fully utilize resources in each region.
[0113] In addition, in related technologies, computing resources are concentrated in the cloud, which cannot be flexibly expanded. It is impossible to utilize resources such as edge computing or self-built data centers within the enterprise, and it is also impossible to support cloud service business through externally leased servers when business suddenly grows.
[0114] Based on this, embodiments of this application provide a resource allocation method, a cloud computing platform, and a computing device cluster. The resource allocation method is applied to a cloud computing platform, which includes a cloud management platform capable of managing cloud resources distributed across different regions. When the cloud management platform receives resource subscription information from a tenant, it allocates cloud resources to the tenant based on this information. The resource subscription information indicates multiple regions, meaning the tenant needs to subscribe to cloud resources located in different regions. Since the cloud management platform can manage cloud resources distributed across different regions, it can allocate cloud resources in multiple target regions to the tenant based on the different regions indicated by the resource subscription information, thereby providing cloud services to the tenant through the cloud resources in the target regions. Here, the target regions can be the regions indicated in the resource subscription information, or the target regions may differ from the regions indicated in the resource subscription information, or multiple target regions may partially overlap with multiple regions indicated in the resource subscription information, etc.
[0115] The technical solution of this application is described below. First, the system architecture of the cloud computing platform in the embodiment of this application is introduced.
[0116] Figure 1 This is a schematic diagram of the system architecture of a cloud computing platform provided in an embodiment of this application. Figure 1As shown, the cloud computing platform provided in this application embodiment includes a cloud management platform 110, a business control platform 120 (distributed access control layer) distributed in different regions, sites 130 (distributed resource layer) distributed in different regions, content delivery network (CDN) acceleration nodes 140 distributed in different regions, and clients 150. Sites access the cloud management platform 110 through the business control platform 120, and the cloud management platform 110 can manage the sites 130 through the business control platform 120.
[0117] The cloud management platform 110 includes a site management unit 111, a resource management unit 112, a global scheduling unit 113, an operation and maintenance unit 114, a business management unit 115, and storage 116. Among these, the site management unit 111 can manage sites in all regions, such as... Figure 2 As shown, the site management unit 111 manages and monitors sites 130 distributed in different regions and the servers within those sites. For example, the site management unit needs to manage the site's regional affiliation, operator network, site type (e.g., central site, edge site, tenant's data center, third-party leased site, etc.), and the quantity of resources within the site. The site management unit 111 also supports site registration and reporting of site status. During site registration, the site needs to report its region and basic information to the site management unit. This basic information includes the site's identity, operator network, site status (including online, paused, and offline), site server information (e.g., number of servers, server model, and server specifications), and access gateway information (gateway domain name or IP address, gateway port address). Furthermore, the site management unit also collects real-time site operational status information through an in-site agent. This operational status information includes the number of running or non-running servers and network connection status.
[0118] like Figure 2 As shown, the resource management unit 112 includes functions such as computing power resource management, network resource management, and resource assessment and management. For example, the resource management unit 112 senses the resource load of cloud services in real time, performs resource estimation and planning, and receives computing power resource requests and valid computing power resource allocations from the business management unit. The resource management unit 112 also senses the operating status of resource sites, locks invalid computing power resources and network resources, and notifies the downstream global scheduling unit and business management unit. Furthermore, the resource management unit 112 performs a comprehensive assessment of site resources within a region in real time based on the usage of site resources, the number of services carried by the sites, and cost consumption, obtains the assessment results, and provides the assessment results to subsequent resource scheduling.
[0119] like Figure 2 As shown, the global scheduling unit 113 includes scheduling data management, scheduling policy management, policy and decision distribution, and CDN scheduling functions. Specifically, the global scheduling unit 113 is used to perceive the overall resource consumption within a site in real time (hereinafter referred to as the first resource usage information), and allocate cloud resources to tenants based on this information. The global scheduling unit is also used to determine the optimal region and carrier line based on the user's Internet Protocol (IP) when the user accesses the site, thereby allocating a site to the user and enabling them to access the site. Furthermore, the global scheduling unit is used to perform resource scheduling based on policies selected by the tenant, and to send the resource scheduling results to the business management platform 120, ensuring that users receive optimal resource allocation after accessing the site.
[0120] The Business Management Unit 115 is a unit that manages global business operations. Specifically, it is used for configuring cloud services and managing business data globally. The Business Management Unit 115 can perform cross-regional business configuration and global management of business data, thus facilitating disaster recovery and failover for the Business Management Platform 120. Simultaneously, the Business Management Unit 115 is also used to distribute cloud service configurations to the Business Management Platform 120, thereby enabling business access and data retrieval.
[0121] The business management unit 115 can deploy various management services, thereby enabling the management of cloud services provided to users. For example... Figure 2 As shown, the services provided by the business management unit 115 include resource scaling and scheduling services, access scheduling services, cloud service management, tenant management, user management, business policy management, application management, image management, operation and maintenance management, and other functions.
[0122] The operation and maintenance unit 114 is used to operate and maintain the site and cloud services based on user resource usage logs and site operation logs. Specifically, the operation and maintenance unit 114 can obtain the overall resource consumption in real time, i.e., the primary resource usage information, and synchronize this information to the business management unit, global scheduling unit, resource management unit, etc., so that each platform can manage the global site from different dimensions based on the primary resource information.
[0123] Storage unit 116 is used to store information about the site during site registration, subscription information of tenants during tenant access, site operation status information and user resource usage information during user cloud service usage, etc. In this embodiment, storage unit 116 may include a relational database service (rds), a distributed cache service (dcs), and an object storage service (obs), thereby providing storage space for the cloud management platform to manage sites in different regions.
[0124] In this embodiment, the cloud management platform 110 can be deployed on a computing device cluster, which includes at least one computing device. The computing device can be a host, server, etc. For example, the cloud management platform 110 can be deployed at sites in one or more regions.
[0125] As one possible implementation, the cloud management platform 110 adopts a dual-active design or a primary-backup design. For example, such as... Figure 1 As shown, the cloud management platform 110 can be divided into a primary cloud management platform and a backup cloud management platform. Data synchronization is possible between the primary and backup cloud management platforms. Furthermore, when the primary cloud management platform fails, unified management can be achieved through the backup cloud management platform. As another example, such as... Figure 1 As shown, the cloud management platform 110 adopts a dual-active design, in which the two cloud management platforms can negotiate which cloud management platform is used for management. The two can be flexibly switched. The communication channel between the cloud management platform 110 used for management and the business control platform 120 in different regions and the site 130 in different regions is opened, while the communication channel between the other cloud management platform 110 and the business control platform 120 in different regions and the site 130 in different regions is closed.
[0126] The business management platform 120 is used for login authentication and pre-allocation of computing resources for cloud service users accessing the site. Furthermore, the business management platform 120 is used to provide users with the means to establish a connection between the client and the site. Figure 1 As shown, the business management platform 120 includes a business management virtual private cloud (VPC) and a load distribution gateway. The business management VPC includes access authentication services and access control services, thereby authenticating and controlling the access of users to the access sites.
[0127] Sites 130, distributed across different regions, are used to provide resources for cloud services, such as computing resources, storage resources, and network resources. For example, such as... Figure 1As shown, site 130 can be a central site and tenant data center distributed in region A, or a central site distributed in region B. The distributed resource layer 130 also includes intelligent edge cloud and third-party leased data center, etc.
[0128] Based on the aforementioned cloud computing platform, the process of resource allocation by the cloud computing platform will be explained in detail below.
[0129] First, before describing the resource allocation method provided in the embodiments of this application, the site creation process will be described in detail.
[0130] Figure 3 This is a flowchart illustrating the site registration method provided in an embodiment of this application. Figure 3 As shown, the site registration method provided in this application embodiment includes the following steps S301 to S303.
[0131] S301, the site sends a registration request to the site management unit, which includes the site's initial information. This initial information refers to the site's basic information at startup. Examples include the site's identity, status, region, type, network information, resource information, access gateway information, and basic server information within the site.
[0132] The status of a site includes normal operation, temporary suspension, and offline. When a site is operating normally, it provides computing, storage, and network resources to the cloud service. When a site needs upgrades or modifications, it will be temporarily suspended and will resume operation after the upgrades or modifications are completed. A site being offline means that it no longer provides computing, storage, and network resources to the cloud service. During the site registration process, the site's status is "normal operation." Site types include central sites, edge sites, tenant data centers, and third-party leased sites. The site's network information includes network bandwidth, the network provider, and network quality. Specifically, network bandwidth information includes backbone bandwidth, international bandwidth, non-international bandwidth, backbone bandwidth latency, international bandwidth latency, and non-international bandwidth latency. The site's resource information includes the number of physical cores of the Central Processing Unit (CPU) within the site, the total memory size of all CPU physical cores, the total number of physical cores of the Graphics Processing Unit (GPU) within the site, the total memory size of the GPU physical cores, the total video memory size of the GPU physical cores, and the site's storage resource size. The site's access gateway information includes the domain name of the access gateway, or the site's Internet Protocol (IP) address and port address. Basic server information includes the number of servers within the site, server identifiers, server models, and server access IP addresses.
[0133] In some embodiments, such as Figure 1 As shown, a proxy can be deployed on the server in the site. When the server in the site starts running, the proxy can also run, thereby obtaining the site's initial information.
[0134] S302, The site management unit determines the site's initial information based on the registration request.
[0135] S303, the site management unit sends the first information to the resource management unit so that the resource management unit can manage the resources within the site.
[0136] In some embodiments, to improve the data flow transmission efficiency of cloud services, the site management unit can also group the site's resources based on the basic information of the servers within the site in the first information to obtain resource group information. This resource group information includes the name of the resource group, the identifier of the resource group, the type of cloud service run by the resources in the resource group, the resource information of the resource group, and network operator information. Resources in different resource groups run different cloud services; for example, resource group Z1 is used to run cloud applications, and resource group Z2 is used to run cloud desktops. The resource information of a resource group refers to information such as the quantity of computing resources and storage resources in the resource group, such as the number of CPUs, the identifier of the server where the CPU is located, the model of the server where the CPU is located, and the access IP address of the server where the CPU is located, and the number of GPUs, the identifier of the server where the GPU is located, and the access IP address of the server where the GPU is located. Different resource groups correspond to different network operators, and the network operator information includes the network operator providing the network for the resource group and the Uniform Resource Locator (URL) of the access gateway for the resource group. For example, the network operators providing the network for the site include network operator Y1, network operator Y2, and network operator Y3. Therefore, the site's resources need to be divided into three resource groups, named Resource Group Z1, Resource Group Z2, and Resource Group Z3. Resource Group Z1 is identified as Z1, Resource Group Z2 as Z2, and Resource Group Z3 as Z3. Network operator Y1 provides network access to Resource Group Z1, network operator Y2 provides network access to Resource Group Z2, and network operator Y3 provides network access to Resource Group Z3. The specific resource allocation strategy can be set according to actual needs; this embodiment does not impose specific limitations on it.
[0137] After the site management unit divides the resources into groups, the resource group information of each group is synchronized to the resource management unit, enabling the resource management unit to manage the site's resources. Furthermore, different resource groups correspond to different network operators. When providing cloud services, resources from different resource groups can be allocated to users based on their region and network operator. In this way, the data flow for cloud services provided to users is transmitted to the client through the same network lines provided by the same network operator, improving data transmission efficiency.
[0138] Through the above process, each site in the distributed resource layer can complete registration, enabling the cloud management platform to uniformly manage sites in all regions and uniformly schedule the resources provided by sites in all regions, thereby improving resource utilization.
[0139] Once all regional sites are connected to the cloud management platform, the sites become usable, providing cloud services to users. During site usage, sites can be expanded or new sites can be added. Site expansion refers to adding resources within the site, such as increasing computing power, storage resources, and network resources. After site resource expansion, the agent within the site can send first information to the site management unit in the cloud management platform 110. This first information, generated after the site resource addition, updates the site's first information in the resource management unit 112. When a new site is added to the distributed resource layer, the newly added site can complete site registration through steps S301 and S303, allowing the resource management unit 112 to add the new site's first information and manage the newly added site.
[0140] In some embodiments, the global scheduling unit comprehensively scores a site based on its initial information and operational status, thereby comprehensively evaluating the site's resource specifications and quality to adjust and optimize its resources. Different indicators correspond to different weights. For example, as shown in Table 1, the global scheduling unit can evaluate a site's performance, cost, and reliability based on indicators such as memory, computing power, and network.
[0141] Table 1
[0142]
[0143] After successful site registration, the site can be put into use. The following section will explain in detail the implementation process of tenant access and user access to the site.
[0144] Based on the above Figures 1 to 3 The corresponding embodiments will provide a detailed description of the resource allocation method provided in the embodiments of this application.
[0145] Figure 4 This is a flowchart illustrating a resource allocation method provided in an embodiment of this application. The resource allocation method provided in this embodiment is applied to... Figure 1 and Figure 2 The cloud computing platform shown. (As shown) Figure 4 The resource allocation method provided in the illustrated embodiment includes the following steps S401 and S402.
[0146] S401, the cloud management platform receives resource subscription information from the tenant, the resource subscription information being used to indicate multiple regions.
[0147] Tenants can select resources to subscribe to in different regions through the tenant management platform. These resources include computing resources, storage resources, and network resources of the servers within the site. Tenants can choose the resources they need to subscribe to based on their own requirements. In some instances, such as... Figure 5 As shown, tenant administrators can select the regions where the resources they wish to subscribe to are located through the tenant management platform. For example, the tenant administrator can select regions Q1, Q2, Q3, and Q4. After selecting the regions, the tenant administrator also needs to select the specifications of the resources to be subscribed to for each region. Specifically, the tenant administrator can select different server models and quantities for each region through the tenant management platform. Figure 5 As shown, in region Q1, the tenant administrator subscribed to 2 servers S1 and 3 servers S2. In region Q2, the tenant administrator subscribed to 4 servers S2 and 1 server S3. In region Q3, the tenant administrator subscribed to 3 servers S3 and 2 servers S1. In region Q4, the tenant administrator subscribed to 5 servers S4 and 1 server S1.
[0148] The resource subscription information includes identifiers for multiple regions and resource specifications for each region. Resource specifications can include, for example, different server models, the required number of CPUs, the number of CPU cores in the server, etc. This is just an example; the specific resource specifications should be selected according to actual needs.
[0149] The tenant management platform generates resource subscription information based on the input from the tenant administrator and sends the resource subscription information to the cloud management platform. For example, the tenant management platform can send resource subscription information to the cloud service management module within the business management unit.
[0150] S402, the cloud management platform allocates cloud resources in multiple target areas to the tenant based on the resource subscription information, and the cloud resources are used to provide cloud services to the tenant.
[0151] The cloud management platform allocates cloud resources within multiple target regions to tenants based on resource subscription information. These cloud resources are used to provide cloud services to tenants. In this embodiment, the multiple target regions may be the regions indicated by the resource subscription information, or the multiple target regions may be partially the same as or different from the regions indicated by the resource subscription information.
[0152] In some embodiments, after a site completes registration, the operation and maintenance unit in the cloud management platform can obtain the site's primary resource usage information in real time through an agent. This primary resource usage information includes resource load information across various dimensions, such as resource load at the business type level, load information for various resource metrics at the server group level, resource load information at the tenant level, and load information for resource metrics at the server level.
[0153] The operations and maintenance unit can synchronize the site's primary resource usage information to the global scheduling unit, site management unit, resource management unit, and business management unit for use by these platforms. For example, the global scheduling unit can allocate cloud resources to tenants and users based on the site's primary resource usage information, such as allocating sites to tenants, sites to users, and servers to users.
[0154] To ensure full resource utilization and improve the user experience of cloud services, the cloud management platform can allocate cloud resources to tenants in different regions based on the initial resource usage information of cloud resources within each region. For example, the cloud management platform can obtain the initial resource usage information of sites within each region in real time, and then allocate sites to tenants based on this information. In this way, because the cloud management platform can manage cloud resources in different regions, it can allocate cloud resources from multiple target regions to tenants, enabling resource sharing across different regions and achieving full resource utilization.
[0155] As one possible implementation, when allocating cloud resources to tenants, the cloud management platform can also allocate cloud resources in multiple target regions to tenants based on information such as the first resource usage information of cloud resources in each region, the region indicated in the resource subscription information, and the distance between regions.
[0156] After cloud resources are allocated to tenants, the users managed by those tenants can access the site where the cloud resources are located, thereby accessing the site and using cloud services. For example, using cloud desktops and cloud applications. The specific implementation process for user access to the site is explained in detail below.
[0157] The cloud management platform can receive access requests sent by clients. These requests indicate that users need to access a site to access cloud resources. Based on the site's primary resource usage information, the cloud management platform determines the target site that the user can access from a pool of primary sites. The primary sites refer to sites within the target area. After determining the target site, the cloud management platform sends the target site's gateway address to the client, thereby enabling the client to access the target site.
[0158] In some embodiments, the resource subscription information further includes a first policy, which indicates the policy for allocating sites to users. There are multiple strategies for resource allocation, and tenants can choose different resource allocation strategies based on the needs of different regions. The first policy includes strategies such as experience priority, cost priority, resource load priority, and security priority. When users use cloud services, the cloud management platform provides site allocation to users in different regions based on the first policy selected by the tenant for different regions. Experience priority means allocating sites to users with the best user experience of the cloud service as the first priority. Cost priority means allocating sites to users with the best cost priority. Resource load priority means allocating sites to users according to the amount of resources required by the cloud service. Security priority means allocating sites to users according to the security level required by the tenant or user for the cloud service.
[0159] For example, a tenant administrator subscribes to resources located in regions Q1 through Q4. For different regions, the tenant administrator can choose different primary strategies based on their needs. For example, ... Figure 6 As shown, for region Q1, the tenant's first strategy is resource load priority; for region Q2, the tenant's first strategy is cost priority; for region Q3, the tenant's first strategy is security priority; and for region Q4, the tenant's first strategy is cost priority. Based on the tenant's first strategy selected for different regions, the tenant management platform sends resource subscription information to the business management unit in the cloud management platform. The resource subscription information received by the business management unit includes the information shown in Table 2 below.
[0160] Table 2
[0161] area First Strategy Region Q1 Prioritize based on resource load Region Q2 Cost priority Area Q3 Safety first Region Q4 Cost priority
[0162] As another example, a tenant administrator can select different primary policies based on the user's group. For instance, a tenant can group the users they manage based on the tenant's organizational structure and select different primary policies for users in different groups. Figure 7 As shown, the users managed by the tenant are divided into three groups: Group 1, Group 2, and Group 3. For Group 1, the tenant selects resource load as the primary strategy. For Group 2, the tenant selects cost as the primary strategy. For Group 3, the tenant selects security as the primary strategy. Based on the primary strategy selected by the tenant for each group, the tenant management platform sends resource subscription information to the business management unit in the cloud management platform.
[0163] As another example, tenant administrators can select different primary policies based on user identity levels. For instance, tenants determine different user identity levels based on their organizational structure. Different identity levels correspond to different primary policies. For example, for high-level users, user experience should be prioritized; therefore, the primary policy for high-level users is experience-first. For medium-level users, resource load should be prioritized; therefore, the primary policy for these users is load-first. For low-level users, the primary policy is cost-first. If a user's identity level is a dedicated line, the primary policy for this user is security-first.
[0164] The above examples are only for the purpose of clearly describing the solution of this application. There are many possible first strategies. Specific cloud service providers can set them according to actual needs, and tenant administrators can choose according to their own needs. Therefore, the embodiments of this application do not make specific limitations.
[0165] In this embodiment, the cloud management platform can determine the target site from multiple first sites based on the first resource usage information and the first policy of the first site.
[0166] For example, such as Figure 8 As shown, based on Figure 2 The cloud management platform shown in this application embodiment includes the following steps S801 to S805 for site access.
[0167] S801, the service management unit receives the client's access request.
[0168] The business management unit in the cloud management platform can manage user services. When a client needs to request cloud services, it can send an access request to the business management unit in the cloud management platform. This access request includes user information.
[0169] S802, the global scheduling unit issues periodic site scheduling policies to the business management unit.
[0170] A periodic site scheduling strategy refers to a scheduling strategy that allocates sites to users within a certain period. For example, a site scheduling strategy with a 5-minute interval, or a site scheduling strategy with a 10-minute interval. The period is set according to specific needs; the above is only an example, and this application does not specifically limit the setting of the period. The periodic site scheduling strategy is generated based on the site's primary resource usage information. For example, the global scheduling unit can generate a site's load score based on the site's primary resource usage information and perform site scheduling based on the site's load score.
[0171] The site's primary resource usage information indicates the consumption of resources within the site and reflects its load status. This load status is based on load data from different dimensions, such as cloud service type, service group, tenant, and individual server. In this way, the global scheduling unit and business management unit can coordinate and schedule the site's load from various perspectives, thereby maximizing resource utilization.
[0172] For example, the sites allocated to the tenant to which the user belongs include sites S1, S2, and S3. The user's region is Shanghai, and the dedicated line access weight ratio is (Shanghai:Beijing = 2:1). The periodic site scheduling strategy for each of sites S1, S2, S3, and S4 is shown in Table 3.
[0173] Table 3
[0174]
[0175] Therefore, based on Table 3, the site scheduling strategy for site S1 includes: site S1's region is Shanghai, it has dedicated line access, the proportion of low-cost resources is 10%, and the load calculation score is 0.56. The site scheduling strategy for site S2 includes: site S2's region is Beijing, it has dedicated line access, the proportion of low-cost resources is 1%, and the load calculation score is 0.435. The site scheduling strategy for site S3 includes: site S3's region is Guiyang, it has public network access (no dedicated line access available), the proportion of low-cost resources is 100%, and the load calculation score is 0.49. The site scheduling strategy for site S4 includes: site S4's region is Guangzhou, it has public network access (no dedicated line access available), the proportion of low-cost resources is 80%, and the load calculation score is 0.66.
[0176] S803, the business management unit determines the target site from multiple first sites based on the first strategy and the periodic site scheduling strategy.
[0177] For example, based on Table 3, users are assigned access sites according to the tenant's first strategy. If the first strategy is experience priority, then site S1 is preferred for the user; if the first strategy is resource load priority, then site S2 is preferred; if the first strategy is cost priority, then site S3 is preferred; if the first strategy is security priority, then according to the dedicated line access weight ratio, Shanghai:Beijing = 2:1, that is, Shanghai is 0.56*2 = 1.12 and Beijing is 0.435*1 = 0.435, so site S1 is preferred.
[0178] S804, the business management unit determines the gateway address of the target site.
[0179] S805: The business management platform sends the gateway address of the target site to the client so that the client can access the target site.
[0180] In some embodiments, the access request includes the user's user information. This user information indicates the user's business details, such as the type of cloud service requested, the user's identity level, the user's network operator information, and the region the user belongs to. The cloud management platform can also determine the target site from multiple first sites based on the first policy, the resource usage information of the first site, and the user information.
[0181] For example, based on the initial resource usage information of each site, the periodic site scheduling strategies for sites S1 to S4 are shown in Table 3. The user's region is Shanghai, and the user's status level is high. Since the initial strategy for users in the Shanghai region is cost-first, and the user's status level is high, user experience needs to be considered. Therefore, assigning the user to site S1 satisfies both user experience and allocates low-cost resources.
[0182] In other embodiments, the periodic site scheduling strategy may also include load thresholds for sites across different dimensions, such as load thresholds at the site level, load thresholds at the service type level, and load thresholds at the tenant level. The cloud management platform can also combine these load thresholds to allocate target sites to users, thereby improving the fit between site allocation and user experience.
[0183] In practical applications, there are various ways to allocate sites to users based on site resource usage information. The specific method depends on the actual situation. The above implementation method is only an example and is not specifically limited in this application.
[0184] In some embodiments, the target site is located in a different region than the user's region. Thus, because the cloud management platform can manage sites in different regions, users can access sites in different regions, enabling resource sharing across different regions and ensuring full utilization of resources.
[0185] After the cloud management platform assigns a site to a user, the user can use cloud services. The following is a detailed explanation of the specific process by which users access cloud services.
[0186] Before a user uses a cloud service, the cloud management platform needs to allocate a server, i.e., the target server, to run the cloud service. This target server is used to provide the cloud service to the user. In this process, the cloud management platform can obtain business requests from clients, which indicate the cloud service requested by the user, i.e., the target cloud service. For example, a user can select the cloud service they wish to request through a client. Figure 9As shown, the cloud service requested by the user is application A within the cloud application. The client sends the identifier of application A (such as name, ID, etc.) in the business request to the cloud management platform. In this way, after receiving the business request, the cloud management platform can determine that the cloud service requested by the user is application A.
[0187] In response to a business request, the cloud management platform determines the target server from multiple servers at the target site based on the resource usage information of the first site and the user information in the access request. The target server then provides the target cloud service to the user.
[0188] In some embodiments, the resource subscription information further includes a second policy, which indicates restrictions on user resource usage. For example, the second policy may include limits on the maximum number of users per server within different sites, maximum CPU usage thresholds, maximum memory usage thresholds, and maximum GPU usage thresholds. When a tenant subscribes to resources, they can select the second policy through the tenant management platform, thereby including the second policy in the resource subscription information and sending it to the cloud management platform. Cloud resource providers may also set different restrictions for tenants to choose from based on actual requests; tenants can choose according to their own needs, and this embodiment does not specifically limit this.
[0189] Based on the resource usage information of the first site, the cloud management platform can determine the resource usage of each server within the first site. Based on the resource usage of each server, user information, and a second policy, it determines the target server for the user's access from multiple servers. In this way, user needs are met while resources are fully utilized.
[0190] As one possible implementation, each unit in the cloud management platform caches the initial resource usage information of a site. The global scheduling unit can generate periodic resource scheduling information based on this initial resource usage information. This periodic resource scheduling information refers to the resource scheduling strategy within a certain period, such as the maximum resource usage strategy for servers or server groups, and the service level of tenants. The maximum resource usage strategy includes a threshold for the number of users a single server can support; if this threshold is reached, the server cannot be allocated to users. The maximum resource usage strategy can also include server resource usage thresholds, such as CPU usage thresholds, memory usage thresholds, and GPU usage thresholds; if the usage of any of these resources reaches a certain threshold, the server cannot be allocated to users. The maximum resource usage strategy can also include server resource usage inflection point values, such as CPU usage inflection point values, GPU usage inflection point values, and memory usage inflection point values; if the usage of any of these resources reaches a certain inflection point value, the server cannot be allocated to users.
[0191] In some embodiments, the cloud management platform can also define application resource consumption models for different cloud services under different operating scenarios based on historical resource usage information for different types of cloud services, thereby predicting the type and size of resource consumption by different cloud services.
[0192] The cloud management platform combines application resource consumption models, primary resource usage information from the primary site, user information, and secondary strategies to determine the target server from the target sites. In this way, by combining application resource consumption models to predict future user resource usage, the user experience can be improved.
[0193] In some embodiments, the target service management platform can receive service requests from clients. The region to which the target service management platform belongs is the same as the region to which the target site belongs. The target service management platform authenticates the user based on the user's identity information in the service request, and if authentication is successful, sends the service request to the cloud management platform. In this way, the cloud management platform manages the cloud services running on sites within the region where the service management platform is located through the service management platform, thereby enabling the cloud management platform to manage sites in different regions through service management platforms in different regions.
[0194] In some embodiments, there may be multiple target servers. The cloud management platform can send the protocol network addresses and exchange codes (such as server tokens) of multiple target servers to the client. For example, the cloud management platform can sort the multiple target servers. This could be done by sorting them according to their load scores, or by combining factors such as load scores, low-cost resource coefficients, and the availability of dedicated line access. The user selects a target server that matches their needs through the client. The client then accesses the target server and uses the target cloud service based on its protocol network address and exchange code.
[0195] It should be noted that in the above embodiments, the cloud management platform allocates resources to users at the server level. In practical applications, the cloud management platform can also allocate resources to users at the virtual instance level, such as virtual machines and containers. This application embodiment does not specifically limit this.
[0196] In some embodiments, after a user accesses the target server, the site reports the user's business session and first resource usage information when using cloud services in real time, thereby adjusting the periodic site scheduling strategy and the periodic resource allocation strategy.
[0197] In some embodiments, to improve the data transmission speed during cloud service usage, the cloud management platform can establish a media channel after the user determines the target server to access. The media channel is used to send data streams to the client, while signaling channels are used to transmit input events generated by the client, such as mouse and keyboard events. This decouples the channels, reduces the bandwidth usage of the signaling channel, and improves the transmission efficiency of the data stream.
[0198] Upon receiving a business request, the cloud management platform determines the CDN acceleration node based on the request's IP address. One possible implementation is that the cloud management platform can determine the client's region using its IP address, thereby identifying the address information of the CDN acceleration nodes within that region. The cloud management platform then sends the CDN acceleration node address information to both the client and the target server, establishing communication connections between the client and the CDN acceleration node, and vice versa. This forms a media channel connecting the client, the CDN acceleration node, and the target server.
[0199] For example, the global scheduling unit has CDN scheduling functionality. After receiving a service request, the global scheduling unit allocates a CDN acceleration node to the client. This CDN acceleration node is located in a region close to the client's region, or the region where the CDN acceleration node is located is within the same regional range as the client's region.
[0200] In this embodiment of the application, the service request also includes the client's session identifier. The target server can manage the association between the media channel and the signaling channel based on the session identifier, thereby determining the association between the input event and the data stream.
[0201] After the client connects to the target server, the client can send input events to the target server. The target server generates a data stream corresponding to the input event based on the input event and sends the input event to the CDN acceleration node, thereby sending the data stream to the client through the CDN acceleration node.
[0202] In some embodiments, in scenarios such as cloud service collaboration and sharing, when multiple clients access the same cloud service, CDN acceleration nodes can also improve the transmission efficiency of data streams. Taking a first client and a second client accessing a target cloud service as an example, the first client sends an input event to the target server, and the target server generates a first data stream. The target server sends the first data stream to a first CDN acceleration node, which is the CDN acceleration node connected to the first client. Furthermore, the target server sends the first data stream to a second CDN acceleration node. The second client can obtain the first data stream from the second CDN acceleration node without inputting any input event, thereby achieving rapid data stream transmission in scenarios such as cloud service collaboration and sharing.
[0203] In some embodiments, to improve user experience and fully utilize resources, the cloud management platform can also elastically scale resources within the site. For example, Figure 10 As shown, the site's resource elastic scaling method includes the following steps S1001 to S1003.
[0204] S1001, the resource management unit determines the first resource adjustment result of the site based on the first resource usage information and the site resource elastic scaling strategy.
[0205] S1002, the resource management unit synchronizes the first resource adjustment result with the site management unit.
[0206] S1003, the site management unit adjusts the site's resources based on the first resource adjustment result.
[0207] Different sites are configured with different resource elastic scaling strategies. The resource management unit can synchronize resource scaling strategies from the site management unit. The resource scaling strategy includes the following information: site identifier, the number of resources scaled up at one time, the overall scaling strategy for the site's resources, and the overall scaling strategy for the site's resources.
[0208] The overall site resource expansion strategies include scaling up based on total CPU utilization (upgrading after reaching a threshold), total memory utilization (upgrading after reaching a threshold), total video memory utilization (upgrading after reaching a threshold), and total storage utilization (upgrading after reaching a threshold). The overall site resource contraction strategies include scaling down based on total CPU utilization (downgrading after reaching a threshold), total memory utilization (downgrading after reaching a threshold), and total storage utilization (downgrading after reaching a threshold).
[0209] The resource management unit can periodically synchronize the first resource usage information of the site from the operation and maintenance monitoring platform. Combining the first resource usage information of the site with the site resource elastic scaling strategy, it calculates the first resource adjustment result of the site. The first resource adjustment result includes the number of units that need to be expanded or reduced.
[0210] The resource management unit can synchronize the first adjustment result with the site management unit, so that the site management unit can adjust the site's resources according to the first resource adjustment result, such as expanding or shrinking the site's resources.
[0211] In other embodiments, the resources provided to tenants can be elastically scaled.
[0212] Tenant administrators can send the tenant's first resource scaling policy to the resource scaling and scheduling service of the business management unit through the tenant management platform. This first resource scaling policy includes policies targeting different dimensions, such as access user sessions, CPU / GPU / memory / network card bandwidth, etc., supporting maximum total number of servers to scale and the number of servers scaled at a time. For example, taking server groups as a unit, the first resource scaling policy targeting access user sessions includes the following information: whether the policy is enabled, maximum scaling quantity, number of servers to scale at a time, total session utilization of the server group (scaling after reaching a threshold), and releasing instances without session connections within a given time. Resource-based elastic scaling policies include the following information: total CPU utilization of the server group (scaling after reaching a threshold), total memory utilization of the server group (scaling after reaching a threshold), and total GPU memory utilization of the server group (scaling after reaching a threshold).
[0213] Cloud service providers can configure various first-resource scaling policies for different tenants and display them to tenant administrators through the tenant management platform. When subscribing to cloud services, tenant administrators can select the first-resource scaling policy to be enabled.
[0214] After a tenant configures its first resource scaling policy, the cloud management platform can adjust the tenant's resources based on the first resource scaling policy and second resource usage information during the user's use of resources to access cloud services. The second resource usage information refers to the cloud service's resource usage, and the first resource usage information includes the second resource usage information. For example, as shown... Figure 11 As shown, the tenant resource elastic scaling method provided in this application embodiment includes the following steps S1101 to S1105.
[0215] S1101, the operations and maintenance unit determines the second resource usage information during the user's use of cloud services.
[0216] The second resource usage information refers to the amount of resources used by the cloud service during its operation, such as CPU utilization, GPU utilization, memory utilization, network card bandwidth utilization, and other indicators.
[0217] S1102, the operations and maintenance unit synchronizes the second resource usage information with the business management unit.
[0218] S1103, the business management unit determines the second resource adjustment result of the cloud service based on the second resource usage information, the first resource usage information of the site, and the first resource scaling strategy.
[0219] S1104, the business management unit sends the second resource adjustment result to the resource management unit.
[0220] S1105, the resource management unit adjusts the cloud service resources based on the second resource adjustment result.
[0221] The business management unit notifies the cloud desktop server management service and cloud application server management service to add or delete resources, thereby achieving maximum resource availability and global resource cost control.
[0222] According to an embodiment of this application, the resource allocation method is applied to a cloud computing platform, which includes a cloud management platform capable of managing cloud resources distributed across different regions. When the cloud management platform receives resource subscription information from a tenant, it can allocate cloud resources to the tenant based on this information. The resource subscription information indicates multiple regions, meaning the tenant needs to subscribe to cloud resources located in different regions. Since the cloud management platform can manage cloud resources distributed across different regions, it can allocate cloud resources in multiple target regions to the tenant based on the different regions indicated by the resource subscription information, thereby providing cloud services to the tenant through the cloud resources in the target regions. Here, the target regions can be the regions indicated in the resource subscription information, or the target regions may be different from the regions indicated in the resource subscription information, or multiple target regions may partially overlap with multiple regions indicated in the resource subscription information, etc.
[0223] Based on the same concept as the method embodiments of this application, this application also provides a cloud computing platform. The cloud computing platform includes a cloud management platform, a business control platform distributed across different regions, and cloud resources distributed across different regions. Each component of the cloud computing platform is used to execute the various steps in the resource allocation method provided in this application. The specific working process of the above-described cloud computing platform can be referred to the corresponding process in the foregoing method embodiments, and will not be repeated here.
[0224] Figure 12This application provides a cloud computing platform, which includes a cloud management platform 1201 for managing cloud resources 1202 distributed in different regions.
[0225] The cloud management platform 1201 is also used to receive resource subscription information from tenants, which is used to indicate multiple regions, and to allocate cloud resources 1202 in multiple target regions to tenants based on the resource subscription information. The cloud resources 1202 are used to provide cloud services to tenants.
[0226] In one possible implementation, the resource subscription information includes the identifiers of multiple regions and the resource specifications for each region.
[0227] In one possible implementation, the cloud management platform 1201 is used to manage sites distributed in different regions. The sites are used to provide cloud resources. The resource subscription information includes a first policy for users to access the sites. The first policy is used to indicate the policy for allocating sites to users. Tenants are used to manage users.
[0228] The cloud management platform 1201 is also used to obtain user access requests;
[0229] The cloud management platform 1201 is also used to respond to access requests and determine the first resource usage information of the first site. The first site is a site in multiple target areas, and the first resource usage information is used to indicate the resource usage status of the site.
[0230] The cloud management platform 1201 is also used to determine the target site from multiple first sites based on their respective first resource usage information and first policies, and send the gateway address of the target site to the client so that the client can access the target site based on the gateway address of the target site.
[0231] In one possible implementation, the resource subscription information includes a second strategy for the user to use the resource, which indicates the restrictions on the user's use of the resource, and the access request includes the user's user information, which indicates the user's business situation.
[0232] The cloud management platform 1201 is also used to obtain business requests from clients, which are used to indicate the target cloud service requested by the user.
[0233] The cloud management platform 1201 is also used to determine the target server from multiple servers of the target site based on the first resource usage information, user information and second policy of the first site. The target server is used to provide the target cloud service to the user.
[0234] In one possible implementation, the resource subscription information also includes a first resource scaling policy of the first site, which is used to indicate the policy for adjusting the resources used by the cloud service. The first resource usage information includes second resource usage information, which is used to indicate the resource usage of different cloud services.
[0235] The cloud management platform 1201 is also used to adjust the resources of the target cloud service based on the second resource usage information and the first resource scaling strategy;
[0236] Adjusting user resources includes expanding or shrinking user resources.
[0237] In one possible implementation, the cloud computing platform also includes a business management and control platform 1203 distributed across different regions;
[0238] The target business management platform is used to receive business requests from clients. These requests also include the user's identity information. The target business management platform is the business management platform within the region where the target site is located.
[0239] The target business management platform is also used to authenticate users based on their identity information, and when the user authentication is successful, the target business management platform sends a business request to the cloud management platform.
[0240] In one possible implementation, the cloud management platform manages the cloud services running on the second site through a business control platform, and the region to which the second site belongs is the same as the region to which the business control platform belongs.
[0241] In one possible implementation, the target business management platform is also used to determine the Content Delivery Network (CDN) acceleration node based on the Internet Protocol (IP) address of the business request, and send the CDN acceleration node's address information to the client so that the client can establish a communication connection with the CDN acceleration node.
[0242] In one possible implementation, the target server is also used to receive input events from the first client, which are used to indicate events that perform operations on the target cloud service;
[0243] The target server is also used to respond to the input event, generate a first data stream, and send the first data stream to the first CDN acceleration node, so that the first CDN acceleration node sends the first data stream to the first client.
[0244] In one possible implementation, the target server is also used to send a first data stream to a second CDN acceleration node for a second client to obtain the first data stream from the second CDN acceleration node. The second client is a client that accesses the target cloud service, and the second client is a different client from the first client.
[0245] In one possible implementation, the regions to which the first client belongs are different from the regions to which the second client belongs.
[0246] In one possible implementation, the target site is located in a region different from the user's region.
[0247] Based on the same concept as the embodiments of the method in this application, embodiments of this application also provide a computing device. This computing device may be a cloud management platform, a business control platform, or a server for providing cloud resources as described in the above embodiments. The following uses a cloud management platform as an example to introduce the computing device provided in the embodiments of this application. Similarly, examples of business control platforms and servers can be found in the examples of cloud management platforms.
[0248] This application also provides a computing device 1300. For example... Figure 13 As shown, the computing device 1300 includes a bus 1302, a processor 1304, a memory 1306, and a communication interface 1308. The processor 1304, the memory 1306, and the communication interface 1308 communicate with each other via the bus 1302. The computing device 1300 can be a server or a terminal device. It should be understood that this application does not limit the number of processors and memories in the computing device 1300.
[0249] Bus 1302 can be a Peripheral Component Interconnect (PCI) bus or an Extended Industry Standard Architecture (EISA) bus, etc. Buses can be divided into address buses, data buses, control buses, etc. For ease of representation, Figure 13 The bus 802 may be represented by a single line, but this does not mean that there is only one bus or one type of bus. The bus 802 may include a path for transmitting information between various components of the computing device 1300 (e.g., memory 1306, processor 1304, communication interface 1308).
[0250] The processor 1304 may include any one or more processors such as a central processing unit (CPU), a graphics processing unit (GPU), a microprocessor (MP), or a digital signal processor (DSP).
[0251] The memory 1306 may include volatile memory, such as random access memory (RAM). The memory 1306 may also include non-volatile memory, such as read-only memory (ROM), flash memory, hard disk drive (HDD), or solid state drive (SSD).
[0252] The memory 1306 stores executable program code, and the processor 1304 executes this executable program code to implement the functions of the site management unit, resource management unit, business management unit, operation and maintenance unit, etc., in the aforementioned cloud management platform, thereby achieving... Figure 4 The method shown. That is, the memory 1306 stores the method for execution. Figure 4 The instructions for the method shown.
[0253] The communication interface 1308 uses transceiver modules such as, but not limited to, network interface cards and transceivers to enable communication between the computing device 1300 and other devices or communication networks.
[0254] This application also provides a computing device cluster.
[0255] Figure 14 This is a computing device cluster 1400 provided in an embodiment of this application. For example... Figure 14 As shown, the computing device cluster includes at least one computing device 1300. The memory 1306 of one or more computing devices 1300 in the computing device cluster may store the same memory for executing... Figure 4 The instructions for the method shown. The cloud management platform or cloud resources can be deployed on a single computing device 1300 or distributed across multiple computing devices 1300.
[0256] The computing device can be a server, such as a central server, an edge server, or a local server in a local data center. In some embodiments, the computing device can also be a terminal device such as a desktop computer, a laptop computer, or a smartphone.
[0257] In some possible implementations, the memory 1306 of one or more computing devices 1300 in the computing device cluster may also store memory for execution. Figure 4 The instructions of the method shown are partial. In other words, a combination of one or more computing devices 1300 can coexist for executing... Figure 4 The instructions for the method shown.
[0258] It should be noted that the memory 1306 in different computing devices 1300 within the computing device cluster can store different instructions, each for execution. Figure 13 The diagram illustrates the functions of the site management unit, resource management unit, business management unit, and operation and maintenance unit. In other words, the instructions stored in the memory 1306 of different computing devices 1300 can implement the functions of one or more modules in the cloud management platform or cloud resources.
[0259] In some possible implementations, one or more computing devices in a computing device cluster can be connected via a network. This network can be a wide area network (WAN) or a local area network (LAN), etc. Figure 15 One possible implementation is shown. For example... Figure 15 As shown, the two computing devices 1300A and 1300B are connected via a network. Specifically, they are connected to the network through communication interfaces in each computing device. In this possible implementation, the memory 1306 in computing device 1300A stores instructions for implementing some functions of the cloud management platform, such as resource management unit and site management unit. Meanwhile, the memory 1306 in computing device 1300B stores instructions for implementing other functions of the cloud management platform, such as business management unit and operation and maintenance unit.
[0260] It should be understood that Figure 15 The functions of computing device 1300A shown can also be performed by multiple computing devices 1300. Similarly, the functions of computing device 1300B can also be performed by multiple computing devices 1300.
[0261] In addition to the methods, apparatus, and computing devices described above, embodiments of this application may also provide a computer program product, comprising computer program instructions. When executed by a processor, these computer program instructions cause the processor to perform the steps of the methods in the various embodiments of this application described in the "Method" section of this specification. The computer program product may be written in any combination of one or more programming languages to perform the operations of the embodiments of this application. The programming languages include object-oriented programming languages such as Java and C++, as well as conventional procedural programming languages such as C or similar languages. The computer program code may be in source code form, object code form, executable file, or some intermediate form. The computer program code may be executed entirely on a user's computing device, partially on a user's computing device, as a standalone software package, partially on a user's computing device and partially on a remote computing device, or entirely on a remote computing device or server.
[0262] Furthermore, embodiments of this application may also provide a computer-readable storage medium storing computer program instructions thereon, which, when executed by a processor, cause the processor to perform the steps of the display control method according to various embodiments of this disclosure as described in the "Method" section above. The computer-readable storage medium may be any combination of one or more readable media. A readable medium may be a readable signal medium or a readable storage medium. A readable storage medium may include, for example, an electrical, magnetic, optical, electromagnetic, infrared, or semiconductor system, apparatus, or device, or any combination thereof. More specific examples of readable storage media (a non-exhaustive list) include: an electrical connection having one or more wires, a portable disk, a hard disk, random access memory (RAM), read-only memory (ROM), erasable programmable read-only memory (EPROM or flash memory), optical fiber, portable compact disk read-only memory (CD-ROM), optical storage device, magnetic storage device, or any suitable combination thereof. It should be noted that the content contained in the computer-readable medium may be appropriately added to or subtracted according to the requirements of legislation and patent practice in a jurisdiction. For example, in some jurisdictions, according to legislation and patent practice, a computer-readable medium may not include electrical carrier signals and telecommunication signals.
[0263] The method steps in the embodiments of this application can be implemented in hardware or by a processor executing software instructions. The software instructions can consist of corresponding software modules, which can be stored in random access memory (RAM), flash memory, read-only memory (ROM), programmable read-only memory (PROM), erasable programmable read-only memory (EPROM), electrically erasable programmable read-only memory (EEPROM), registers, hard disks, portable hard disks, CD-ROMs, or any other form of storage medium known in the art. An exemplary storage medium is coupled to the processor, enabling the processor to read information from and write information to the storage medium. Of course, the storage medium can also be a component of the processor. The processor and the storage medium can reside in an ASIC.
[0264] It is understood that the various numerical designations used in the embodiments of this application are merely for descriptive convenience and are not intended to limit the scope of the embodiments of this application. It should be understood that in the embodiments of this application, the order of the process numbers does not imply the order of execution; the execution order of each process should be determined by its function and internal logic, and should not constitute any limitation on the implementation process of the embodiments of this application.
[0265] Finally, it should be noted that the above embodiments are only used to illustrate the technical solutions of this application, and are not intended to limit them. Although this application has been described in detail with reference to the foregoing embodiments, those skilled in the art should understand that modifications can still be made to the technical solutions described in the foregoing embodiments, or equivalent substitutions can be made to some of the technical features. Such modifications or substitutions do not cause the essence of the corresponding technical solutions to deviate from the protection scope of the technical solutions of the embodiments of this application.
Claims
1. A resource allocation method, characterized in that, Applied to a cloud computing platform, the cloud computing platform including a cloud management platform, the cloud management platform being used to manage cloud resources distributed in different regions, the method includes: The cloud management platform receives resource subscription information from tenants, and the resource subscription information is used to indicate multiple regions; The cloud management platform allocates cloud resources in multiple target regions to the tenant based on the resource subscription information, and the cloud resources are used to provide cloud services to the tenant.
2. The method according to claim 1, characterized in that, The resource subscription information includes the identifiers of multiple regions and the resource specifications for each region.
3. The method according to claim 1 or 2, characterized in that, The cloud management platform is used to manage sites distributed in different regions. Each site provides cloud resources. The resource subscription information includes a first policy for user access to the site, which indicates a policy for allocating sites to the user. The tenant manages the user. The method further includes: The cloud management platform obtains the user's access request; In response to the access request, the cloud management platform determines the first resource usage information of the first site, where the first site is a site within the plurality of target areas, and the first resource usage information is used to indicate the resource usage status of the site. The cloud management platform determines the target site from the multiple first sites based on the first resource usage information and the first policy of each of the multiple first sites; The cloud management platform sends the gateway address of the target site to the client, so that the client can access the target site based on the gateway address of the target site.
4. The method according to claim 3, characterized in that, The resource subscription information includes a second policy for the user's resource usage, the second policy indicating restrictions on the user's resource usage; the access request includes the user's user information, the user information indicating the user's business status; the method further includes: The cloud management platform receives a business request from the client, and the business request is used to indicate the target cloud service requested by the user; The cloud management platform determines a target server from multiple servers of the target site based on the first resource usage information of the first site, the user information, and the second strategy. The target server is used to provide the target cloud service to the user.
5. The method according to claim 4, characterized in that, The resource subscription information further includes a first resource scaling policy for the first site, the first resource scaling policy being used to indicate a strategy for adjusting the resources used by the cloud service. The first resource usage information includes second resource usage information, the second resource usage information being used to indicate the resource usage of different cloud services. The method includes: The cloud management platform adjusts the resources of the target cloud service based on the second resource usage information and the first resource scaling-out strategy; Adjusting the user's resources includes expanding or shrinking the user's resources.
6. The method according to claim 4 or 5, characterized in that, The cloud computing platform also includes a business management and control platform distributed across the different regions, and the method further includes: The target business management platform receives the business request from the client, wherein the business request also includes the user's identity information, and the target business management platform is the business management platform in the region where the target site is located; The target business management platform authenticates the user based on the user's identity information; The cloud management platform receives business requests from the client, including: Once the user authentication is successful, the target business management platform sends the business request to the cloud management platform.
7. The method according to claim 6, characterized in that, The cloud management platform manages the cloud services running on the second site through the business control platform. The region to which the second site belongs is the same as the region to which the business control platform belongs.
8. The method according to any one of claims 4-7, characterized in that, The method further includes: The cloud management platform determines the Content Delivery Network (CDN) acceleration node based on the Internet Protocol (IP) address of the service request. The cloud management platform sends the address information of the CDN acceleration node to the client so that the client can establish a communication connection with the CDN acceleration node.
9. The method according to claim 8, characterized in that, The method further includes: The target server receives an input event from the first client, the input event being used to indicate an event for operating the target cloud service; The target server responds to the input event by generating a first data stream; The target server sends the first data stream to the first CDN acceleration node, so that the first CDN acceleration node sends the first data stream to the first client.
10. The method according to claim 9, characterized in that, The method further includes: The target server sends the first data stream to the second CDN acceleration node so that the second client can obtain the first data stream from the second CDN acceleration node. The second client is a client that accesses the target cloud service, and the second client is a different client from the first client.
11. The method according to claim 10, characterized in that, The regions to which the first client belongs are different from the regions to which the second client belongs.
12. The method according to any one of claims 3-11, characterized in that, The target site is located in a region different from the user's region.
13. A cloud computing platform, characterized in that, The cloud computing platform includes a cloud management platform, which is used to manage cloud resources distributed in different regions; The cloud management platform is also used to receive resource subscription information from tenants, the resource subscription information being used to indicate multiple regions, and to allocate cloud resources in multiple target regions to the tenants based on the resource subscription information, the cloud resources being used to provide cloud services to the tenants.
14. The cloud computing platform according to claim 13, characterized in that, The resource subscription information includes the identifiers of multiple regions and the resource specifications for each region.
15. The cloud computing platform according to claim 13 or 14, characterized in that, The cloud management platform is used to manage sites distributed in different regions. The sites are used to provide cloud resources. The resource subscription information includes a first policy for users to access sites. The first policy is used to indicate the policy for allocating sites to users. The tenant is used to manage users. The cloud management platform is also used to obtain the user's access request; The cloud management platform is also used to respond to an access request and determine the first resource usage information of the first site, wherein the first site is a site within the plurality of target areas, and the first resource usage information is used to indicate the resource usage status of the site. The cloud management platform is also used to determine a target site from the multiple first sites based on the first resource usage information and the first policy of each of the multiple first sites, and send the gateway address of the target site to the client so that the client can access the target site based on the gateway address of the target site.
16. The cloud computing platform according to claim 15, characterized in that, The resource subscription information includes a second strategy for the user to use resources, the second strategy being used to indicate the restrictions on the user's use of resources; the access request includes the user's user information, the user information being used to indicate the user's business status. The cloud management platform is also used to obtain business requests from the client, and the business requests are used to indicate the target cloud service requested by the user; The cloud management platform is also used to determine a target server from multiple servers of the target site based on the first resource usage information of the first site, the user information, and the second strategy. The target server is used to provide the target cloud service to the user.
17. The cloud computing platform according to claim 16, characterized in that, The resource subscription information also includes a first resource scaling strategy of the first site, which is used to indicate a strategy for adjusting the resources used by the cloud service. The first resource usage information includes second resource usage information, which is used to indicate the resource usage of different cloud services. The cloud management platform is also used to adjust the resources of the target cloud service based on the second resource usage information and the first resource scaling strategy; Adjusting the user's resources includes expanding or shrinking the user's resources.
18. The cloud computing platform according to claim 16 or 17, characterized in that, The cloud computing platform also includes business management and control platforms distributed across the different regions; The target business management platform is used to receive the business request from the client, wherein the business request also includes the user's identity information, and the target business management platform is the business management platform in the area where the target site is located; The target business management platform is also used to authenticate the user based on the user's identity information, and if the user authentication is successful, the target business management platform sends the business request to the cloud management platform.
19. The cloud computing platform according to claim 18, characterized in that, The cloud management platform manages the cloud services running on the second site through the business control platform. The region to which the second site belongs is the same as the region to which the business control platform belongs.
20. The cloud computing platform according to claim 19, characterized in that, The target service management platform is also used to determine the Content Delivery Network (CDN) acceleration node based on the Internet Protocol (IP) address of the service request, and send the address information of the CDN acceleration node to the client so that the client can establish a communication connection with the CDN acceleration node.
21. The cloud computing platform according to claim 20, characterized in that, The target server is also configured to receive input events from the first client, the input events being used to indicate events that perform operations on the target cloud service; The target server is also configured to respond to the input event, generate a first data stream, and send the first data stream to a first CDN acceleration node, so that the first CDN acceleration node sends the first data stream to the first client.
22. The cloud computing platform according to claim 21, characterized in that, The target server is also used to send the first data stream to the second CDN acceleration node so that the second client can obtain the first data stream from the second CDN acceleration node. The second client is a client that accesses the target cloud service, and the second client is a different client from the first client.
23. The cloud computing platform according to claim 22, characterized in that, The regions to which the first client belongs are different from the regions to which the second client belongs.
24. The cloud computing platform according to any one of claims 15-23, characterized in that, The target site is located in a region different from the user's region.
25. A computing device cluster, characterized in that, It includes at least one computing device, each computing device including a processor and memory; The processor of the at least one computing device is configured to execute instructions stored in the memory of the at least one computing device to cause the cluster of computing devices to perform the method as described in any one of claims 1-12.
26. A computer program product containing instructions, characterized in that, When the instruction is executed by the computing device cluster, the computing device cluster causes the computing device cluster to perform the method as described in any one of claims 1-12.
27. A computer-readable storage medium, characterized in that, Includes computer program instructions, which, when executed by a cluster of computing devices, perform the method as described in any one of claims 1-12.
