Dynamic scheduling and operation management method and platform for cloud machine resources

Abstract

The invention belongs to the technical field of cloud computing and mobile computing, and discloses a dynamic scheduling and operation management method and platform for cloud machine resources, and the method comprises the steps: maintaining and summarizing the inventory information of each resource pool, and setting a global resource view of a maximum concurrent streaming instance number threshold based on the streaming processing capability for each physical server; in response to a cloud machine creation request, in combination with the threshold, the scheduling strategy and the current load of the target host, deciding and creating a cloud machine instance; during the running period of the cloud machine, monitoring the streaming session state of the cloud machine in real time, and automatically triggering online upgrading or descending of computing resources according to the streaming starting / ending state; and in response to a capacity expansion request for the running cloud machine data disk, expanding the underlying storage volume online and triggering the expansion of a file system in the cloud machine to utilize the newly increased capacity. According to the mode, unified and efficient scheduling of multiple resource pools, refined resource management and control of streaming services and flexible and elastic supply of storage resources are realized.

Description

Technical Field

[0001] This invention relates to the fields of cloud computing and mobile computing technology, and in particular to a method and platform for dynamic scheduling and operation management of cloud machine resources. Background Technology

[0002] Existing technologies, especially those involving dynamic scheduling of cloud platform resources, primarily focus on virtual machine resource monitoring, load balancing, and energy consumption optimization in general computing environments. When applied to mobile cloud computing scenarios with specific business models and complex management needs, such as "cloud phones," they suffer from the following drawbacks: (1) There are shortcomings in the unified management of large-scale resource pools with multiple regions and strong isolation: Existing technical solutions typically target resource scheduling for single or logically centralized data centers, lacking an effective mechanism for unified management, inventory maintenance, and efficient collaborative scheduling of multiple physically dispersed and logically isolated large-scale resource pools (such as those spanning provinces or regions). In scenarios requiring unified management of resource pools across 30 provinces, ensuring strong isolation, and with strict requirements for resource allocation response time (e.g., no more than 5 minutes), existing general scheduling methods face challenges in terms of scheduling efficiency, policy consistency, and isolation assurance.

[0003] (2) Insufficient optimization of resource management for specific business types such as "streaming": Existing technologies primarily aim at load balancing of computing nodes or reduction of overall energy consumption, lacking refined control over the core streaming service characteristics of "cloud phones." For example, they fail to effectively address how to dynamically set and adjust thresholds based on the server's streaming processing capacity to limit the number of concurrent streams per server, thereby ensuring streaming quality and user experience; they also fail to implement differentiated dynamic resource allocation strategies (such as upgrading streaming cloud machines and downgrading non-streaming cloud machines) based on whether the service is streaming or not.

[0004] (3) In terms of storage resource management, there is a lack of flexible and refined control capabilities that are closely linked to business operations: Existing technologies focus more on the dynamic scheduling of computing resources, while the management of storage resources (such as local cloud disks) that are strongly related to "cloud phones" is relatively weak. Specifically, there is a lack of on-demand, GB-level flexible definition and dynamic adjustment mechanisms for cloud phone data disk capacity, as well as the ability to expand capacity without affecting business operations (no downtime), making it difficult to meet the business's demand for agile and elastic supply of storage resources.

[0005] The above content is only used to help understand the technical solution of the present invention and does not represent an admission that the above content is prior art. Summary of the Invention

[0006] The main objective of this invention is to provide a dynamic scheduling and operation management method and platform for cloud machine resources, aiming to solve the technical problems of insufficient unified management of large-scale resource pools with strong isolation in multiple regions, insufficient optimization of resource control for specific business types such as "streaming", and lack of flexible and refined control capabilities for storage resources that are closely linked to business.

[0007] To achieve the above objectives, this invention provides a method for dynamic scheduling and operation management of cloud server resources. This method is applied to a platform that manages multiple physically dispersed and logically isolated resource pools in a unified manner. The method includes: Maintain a global resource view that aggregates the resource inventory information of physical servers in each resource pool, and set a maximum concurrent streaming instance threshold for each physical server based on the streaming processing capacity of the global resource view. In response to the cloud machine creation request, a target physical host that has not exceeded the threshold and has sufficient resources is determined based on at least the maximum number of concurrent streaming instances contained in the global resource view, the preset scheduling policy, and the current number of concurrent streaming instances of the target physical server, and a cloud machine instance is created on the target physical host. During the operation of the cloud machine instance, the streaming session status of each cloud machine instance is monitored in real time. The streaming session status includes the streaming start status and the streaming end status. When the target cloud machine instance is detected to be in the streaming start state, the computing resources of the target cloud machine instance are upgraded online. When the target cloud server instance is detected to be in a streaming termination state and meets the preset conditions, the computing resources of the target cloud server instance are reduced online. In response to a request to expand the data disk bound to a running cloud instance, the corresponding storage volume in the underlying storage system is expanded online, and the operating system inside the cloud instance is triggered to expand the file system online to utilize the increased capacity.

[0008] In one embodiment, the preset scheduling strategy includes at least a load balancing strategy for selecting a low-load resource pool when no region is specified, and a strategy for selecting a low-load host that has not exceeded a threshold within the resource pool of a specified region.

[0009] In one embodiment, the online upgrade is to adjust the computing specifications from the basic configuration to the streaming enhancement configuration, and the online downgrade is to adjust the computing specifications from the streaming enhancement configuration to the basic configuration.

[0010] In one embodiment, the maximum number of concurrent streaming instances of the physical server is determined dynamically or statically based on at least one of the following factors: server model, GPU encoding performance, or network bandwidth.

[0011] In one embodiment, the response time from responding to the cloud machine creation request to completing the creation of the cloud machine instance on the target physical host does not exceed a preset time threshold, and the online upgrade or downgrade is completed without restarting the target cloud machine instance.

[0012] In one embodiment, the online capacity expansion of the corresponding storage volume in the underlying storage system includes: Call the interface of the underlying storage system to expand the capacity of the storage volume corresponding to the data disk; notify the physical host where the cloud machine instance is located; The disk capacity change notification is sent to the operating system of the cloud machine instance through the agent on the physical host using the virtualization channel; In response to the notification, the operating system of the cloud instance expands its file system online to cover the newly added storage space.

[0013] In addition, to achieve the above objectives, the present invention also proposes a dynamic scheduling and operation management platform for cloud server resources, which is configured to execute the method described above.

[0014] In one embodiment, the dynamic scheduling and operation management platform for cloud server resources includes: The control plane service cluster shall include at least: The resource pool management service module is used to maintain the global resource view; The scheduling strategy service module is used to execute scheduling decisions during the cloud machine intelligent scheduling creation operation; The threshold control service module is used to manage and provide the maximum concurrent streaming instance count threshold. The cloud machine lifecycle service module is used to perform cloud machine intelligent scheduling creation operations and resource adjustments in the dynamic tuning operations of the streaming status perception. The storage management service module is used to perform online expansion operations of cloud machine storage resources; The monitoring and reporting service module is used to monitor and report the streaming session status to trigger dynamic tuning operations for streaming status awareness. Multiple data plane proxies are deployed on various physical servers to execute instructions from the control plane service cluster, collect local status data, and report events.

[0015] Furthermore, to achieve the above objectives, the present invention also proposes a dynamic scheduling and operation management device for cloud server resources. The dynamic scheduling and operation management device for cloud server resources includes: a memory, a processor, and a dynamic scheduling and operation management program for cloud server resources stored in the memory and executable on the processor. The dynamic scheduling and operation management program for cloud server resources is configured to implement the steps of the dynamic scheduling and operation management method for cloud server resources as described above.

[0016] Furthermore, to achieve the above objectives, the present invention also proposes a storage medium storing a dynamic scheduling and operation management program for cloud server resources. When the dynamic scheduling and operation management program for cloud server resources is executed by a processor, it implements the steps of the dynamic scheduling and operation management method for cloud server resources as described above.

[0017] Furthermore, to achieve the above objectives, the present invention also proposes a computer program product, wherein the computer program product stores a dynamic scheduling and operation management program for cloud server resources, and when the dynamic scheduling and operation management program for cloud server resources is executed by a processor, it implements the steps of the dynamic scheduling and operation management method for cloud server resources as described above.

[0018] This invention maintains and summarizes the inventory information of each resource pool and sets a global resource view for each physical server, defining a threshold for the maximum number of concurrent streaming instances based on its streaming processing capacity. In response to cloud machine creation requests, it decides and creates cloud machine instances based on the thresholds, scheduling policies, and the current load of the target host. During cloud machine operation, it monitors the streaming session status in real time and automatically triggers online upgrades or downgrades of computing resources based on the start / end status of the streaming. In response to requests to expand the data disk of a running cloud machine, it expands the underlying storage volume online and triggers file system expansion within the cloud machine to utilize the increased capacity. This approach achieves unified and efficient scheduling of multiple resource pools, refined resource management of streaming services, and flexible and elastic supply of storage resources. Attached Figure Description

[0019] Figure 1 This is a flowchart illustrating the first embodiment of the dynamic scheduling and operation management method for cloud server resources of the present invention. Figure 2 This is a schematic diagram of the overall architecture of the dynamic scheduling and operation management method for cloud machine resources in this invention; Figure 3 This is a schematic diagram of the cloud machine creation and intelligent scheduling process in the dynamic scheduling and operation management method of cloud machine resources of the present invention; Figure 4 This is a schematic diagram of the dynamic allocation process of streaming status perception in the dynamic scheduling and operation management method of cloud machine resources of the present invention. Figure 5This is a schematic diagram of the online expansion process of the cloud-based mobile cloud disk in the dynamic scheduling and operation management method of cloud machine resources of the present invention.

[0020] The realization of the objective, functional features and advantages of the present invention will be further explained in conjunction with the embodiments and with reference to the accompanying drawings. Detailed Implementation

[0021] It should be understood that the specific embodiments described herein are for illustrative purposes only and are not intended to limit the scope of the invention.

[0022] This invention provides a method for dynamic scheduling and operation management of cloud server resources, referring to... Figure 1 , Figure 1 This is a flowchart illustrating the first embodiment of a dynamic scheduling and operation management method for cloud server resources according to the present invention.

[0023] In this embodiment, the method for dynamic scheduling and operation management of cloud server resources includes the following steps: Step S10: Maintain a global resource view.

[0024] In this embodiment, the execution entity is a dynamic scheduling and operation management device for cloud server resources. This device has functions such as data processing, data communication, and program execution. The dynamic scheduling and operation management device for cloud server resources can be a computer terminal device or other network device, or other devices with similar functions. This embodiment does not limit the scope of the application.

[0025] It should be noted that existing technologies, especially those involving dynamic scheduling of cloud platform resources, primarily focus on virtual machine resource monitoring, load balancing, and energy consumption optimization in general computing environments. When applied to mobile cloud computing scenarios with specific business models and complex management needs, such as "cloud phones," the following drawbacks exist: Insufficient unified management of large-scale resource pools with strong isolation across multiple regions: Existing technologies typically schedule resources for a single or logically centralized data center, lacking an effective mechanism for unified management, inventory maintenance, and efficient collaborative scheduling of multiple physically dispersed and logically isolated large-scale resource pools (e.g., across provinces or regions). In scenarios requiring unified management of resource pools across 30 provinces, ensuring strong isolation, and with strict requirements for resource allocation response time (e.g., no more than 5 minutes), existing general scheduling methods face challenges in terms of scheduling efficiency, policy uniformity, and isolation assurance. Insufficient optimization of resource control for specific business types such as "streaming": Existing technologies primarily aim at load balancing of computing nodes or overall energy consumption reduction, lacking refined control over the core streaming business characteristics of "cloud phones." For example, there is a lack of effective solutions on how to dynamically set and adjust thresholds based on the server's streaming processing capabilities to limit the number of concurrent streams per server, thereby ensuring streaming quality and user experience. Furthermore, there is a lack of differentiated dynamic resource allocation strategies based on whether the business is streaming or not (e.g., upgrading streaming cloud servers and downgrading non-streaming cloud servers). In storage resource management, there is a lack of flexible and refined control capabilities closely linked to business operations: existing technologies focus more on the dynamic scheduling of computing resources, with relatively weak management of storage resources strongly related to "cloud phones" (such as local cloud disks). Specifically, there is a lack of flexible, on-demand, GB-level definition and dynamic adjustment mechanisms for cloud phone data disk capacity, as well as the ability to expand capacity without affecting business operations (no downtime), making it difficult to meet the business's demand for agile and elastic storage resource supply.

[0026] To address the aforementioned technical issues, this embodiment maintains and summarizes the inventory information of each resource pool and sets a global resource view for each physical server, defining a threshold for the maximum number of concurrent streaming instances based on its streaming processing capacity. In response to cloud machine creation requests, it decides and creates cloud machine instances based on the threshold, scheduling strategy, and the current load of the target host. During cloud machine operation, its streaming session status is monitored in real time, and online upgrades or downgrades of computing resources are automatically triggered based on the start / end status of the streaming. In response to requests to expand the data disk of a running cloud machine, the underlying storage volume is expanded online, and the file system within the cloud machine is expanded to utilize the increased capacity. This approach achieves unified and efficient scheduling of multiple resource pools, refined resource management of streaming services, and flexible and elastic supply of storage resources. Specifically, it can be implemented as follows.

[0027] In its specific implementation, this embodiment provides a dynamic scheduling and operation management platform and method for cloud machine resources, aiming to solve the problems mentioned in the background art, such as unified management of large-scale multi-resource pools, refined control of streaming services, and flexible supply of storage resources. The technical solution of the present invention will be described in detail below with reference to the accompanying drawings and embodiments. First, the overall system architecture of the present invention will be described. (Refer to...) Figure 2 , Figure 2 This is a system architecture diagram of the cloud server resource dynamic scheduling and operation management platform provided in this embodiment. (For example...) Figure 2As shown, the platform adopts a layered distributed architecture, mainly including: a management console, a control plane service cluster, a data plane proxy, and an underlying infrastructure resource pool. 1. Management Console: Provides a graphical user interface for platform administrators and operators, used for resource pool information maintenance, image specification management, scheduling policy configuration, threshold rule setting, resource monitoring, and report viewing. All configuration commands are issued to the control plane service cluster via a RESTful API. 2. Control Plane Service Cluster: This is the core of the platform, composed of a series of microservice-based, horizontally scalable components, specifically including: Resource Pool Management Service: Responsible for maintaining metadata information, resource inventory (such as the number of physical servers, total CPU / memory / storage, number of available streaming ports, etc.) for all regional resource pools (e.g., 30 provinces), and logical isolation policies between resource pools. It maintains a global resource view. Image Specification Management Service: Manages information related to AOSP (Android OpenSource Project) cloud machine images, including image version, associated virtual machine specifications (vCPU, memory, base system disk size, etc.), and is responsible for synchronizing images to the image repositories of each resource pool. **Scheduling Strategy Service:** A built-in strategy engine, based on various administrator-preset strategies (such as load balancing, region affinity, and cost optimization), combined with the global view provided by the resource pool management service, determines the optimal target resource pool and target physical host for cloud machine creation and migration requests. **Threshold Control Service:** Manages streaming threshold rules for physical servers. Rules can be set based on server model, network bandwidth, GPU performance, and other dimensions to determine the maximum number of concurrent streaming instances. This service ensures that resource allocation does not exceed the threshold, guaranteeing streaming quality. **Cloud Machine Lifecycle Service:** Responsible for the full lifecycle management of cloud machine instances, including initiating and coordinating commands for creation, startup, stop, deletion, and dynamic scaling (upgrade / downgrade). **Storage Management Service:** Dedicated to managing the local cloud disk of the "Cloud Smart Phone." Provides capabilities such as cloud disk creation, binding, unbinding, and online expansion (adjusted by GB), and interacts with the underlying storage system to complete volume management operations. Monitoring and Reporting Service: Receives cloud machine status, resource utilization, and key streaming status events (such as streaming start and end) from the data plane agent. This service is the core event source for triggering dynamic scheduling and allocation. 3. Data Plane Agent: Deployed on each physical server (compute node), it is a lightweight daemon process. Its main responsibilities include: Command Execution: Receiving and executing cloud machine management commands issued from the control plane (such as creating / deleting virtual machines via libvirt or container tools). Status Collection and Reporting: Periodically collects resource usage (CPU, memory, IO) of cloud machines on the local machine and host machine, and monitors the streaming session status of cloud machines in real time. Once a streaming connection is established or disconnected, it immediately reports the event to the monitoring and reporting service.Local policy enforcement: Receives and caches the local streaming threshold values, applying hard limits when allocating resources locally. 4. Infrastructure resource pool: Consists of physical server clusters, network devices, and storage systems distributed across various regions (provinces). Each resource pool is logically isolated, enjoying independent network address ranges and storage resources. Virtualization layers (such as KVM) run on the physical servers to host cloud machine instances.

[0028] It should be noted that in this embodiment, the global resource view summarizes the resource inventory information of physical servers in each resource pool, and sets a maximum concurrent streaming instance threshold for each physical server based on the streaming processing capacity of the global resource view. The maximum concurrent streaming instance threshold for a physical server is determined dynamically or statically by at least one of the following factors: server model, GPU encoding performance, or network bandwidth.

[0029] Step S20: In response to the cloud machine creation request, determine the target physical host that has not exceeded the threshold and has sufficient resources based on at least the maximum number of concurrent streaming instances contained in the global resource view, the preset scheduling policy, and the current number of concurrent streaming instances of the target physical server, and create a cloud machine instance on the target physical host.

[0030] In this implementation, it is necessary to determine a target physical host that has not exceeded a threshold and has sufficient resources. The determination of the target physical host is based on a threshold for the maximum number of concurrent streaming instances included in the global resource view, a preset scheduling policy, and the current number of concurrent streaming instances on the target physical server. The preset scheduling policy includes at least a load balancing policy for selecting a low-load resource pool when no region is specified, and a policy for selecting a low-load host that has not exceeded a threshold within the resource pool of a specified region. After determining the target physical host, a cloud server instance is created on the target physical host.

[0031] Furthermore, for ease of understanding, this embodiment combines... Figure 3 The cloud server creation and intelligent scheduling process in this embodiment is illustrated below. When a user requests the creation of a cloud server of a specified size and region via API, the platform operates according to the following steps: 1. Receiving Creation Requests. The management console or API gateway receives user requests to create cloud machines. These requests include information such as the required image type, computing specifications (e.g., 4 vCPUs / 8GB memory), desired region (optional), and data disk size.

[0032] Second: Request scheduling decision. After receiving the internal processing request, the cloud machine lifecycle service initiates a scheduling decision request to the scheduling policy service, passing in cloud machine specifications, regional preferences, and other constraints.

[0033] 3. Execute the scheduling strategy. The scheduling strategy service calls the resource pool management service to obtain a global resource view and inventory information for all current resource pools. The strategy engine then filters and makes decisions based on preset strategies: If a region is specified, resources from that region's resource pool will be selected first.

[0034] Based on the load balancing strategy, resource pools with low CPU / memory resource utilization are selected.

[0035] Within the target resource pool, a list of physical servers that have not reached the streaming threshold and have sufficient resources is selected by combining the current number of streaming instances and the threshold value provided by the threshold control service.

[0036] Finally, the policy engine outputs an optimal target resource pool ID and target physical host IP.

[0037] 4. Allocate resources and generate configurations. The resource pool management service locks the allocated resources in the global view (reducing available inventory). The cloud machine lifecycle service generates specific cloud machine configurations based on the image specifications and notifies the storage management service to create a data volume in the target resource pool's storage system according to the requested size (e.g., 100GB).

[0038] 5. Issuing Creation Instructions. The cloud machine lifecycle service will issue detailed cloud machine creation instructions (including image URL, computing specifications, network configuration, and data volume ID) to the data plane agent on the target physical host.

[0039] 6. Perform local creation. The data plane agent on the target host calls the local virtualization interface, mounts the specified image and storage volume, and starts the cloud instance according to the configuration parameters.

[0040] 7. Update Status and Complete. After the cloud server starts successfully, the data plane agent reports the cloud server status as "Running" to the monitoring and reporting service. The cloud server lifecycle service updates the cloud server status and returns the cloud server's access information (such as IP address and connection port) to the user, thus completing the entire creation process. From point one to point seven, the platform ensures that, with sufficient resources and clear policies, the overall response time is completed within the designed target (5 minutes).

[0041] It should be emphasized that, in this embodiment, the response time from responding to the cloud machine creation request to completing the creation of the cloud machine instance on the target physical host does not exceed a preset time threshold, and the online upgrade or downgrade is completed without restarting the target cloud machine instance.

[0042] Step S30: During the operation of the cloud machine instance, monitor the streaming session status of each cloud machine instance in real time.

[0043] In the specific implementation, during the operation of the cloud machine instance, this embodiment needs to perform streaming status awareness, specifically by monitoring the streaming session status of each cloud machine instance in real time. The streaming session status includes the streaming start status and the streaming end status.

[0044] It should be noted that for the "streaming start" event: the decision logic checks whether the cloud machine's current specifications are "basic specifications". If so, an "upgrade task" is automatically generated, with the target specifications possibly being the preset "streaming enhancement specifications" (e.g., increasing the number of vCPU cores or allocating additional GPU resources to improve encoding performance). For the "streaming end" event: the decision logic starts a timer. If the cloud machine does not report a "streaming start" event within a preset quiet period (e.g., 30 minutes), a "downgrade task" is generated, reverting its specifications from "streaming enhancement specifications" to "basic specifications" to conserve resources.

[0045] Step S40: When the target cloud machine instance is detected to be in the streaming start state, the computing resources of the target cloud machine instance are upgraded online.

[0046] Step S50: When it is detected that the target cloud machine instance is in the stream termination state and meets the preset conditions, the computing resources of the target cloud machine instance are downgraded online.

[0047] It should be noted that after performing stream sensing, this embodiment will also dynamically adjust resources based on the results of stream sensing. Specifically, this includes online upgrades, which adjust the computing specifications from the basic configuration to the stream-enhanced configuration, and online downgrades, which adjust the computing specifications from the stream-enhanced configuration back to the basic configuration.

[0048] Furthermore, for ease of understanding, this embodiment combines... Figure 4 As shown, the dynamic allocation process of stream status awareness in this embodiment is explained.

[0049] Reference Figure 4 As shown, the process is triggered by event reporting from the data plane proxy: 1. Monitoring and Reporting Streaming Events. The data plane agent monitors the streaming port connections of the cloud machines under its jurisdiction in real time. When a new streaming session is detected to have been established on a cloud machine, a "Streaming Start" event is immediately generated and reported to the monitoring and reporting service. The event includes information such as the cloud machine ID and timestamp. Similarly, when a streaming session ends, a "Streaming End" event is reported.

[0050] 2. Event Analysis and Decision-Making. After receiving an event, the monitoring and reporting service persists it to the database and triggers dynamic allocation decision-making logic.

[0051] For the "streaming start" event: the decision logic checks whether the current specifications of the cloud machine are "basic specifications". If so, an "upgrade task" is automatically generated, and the target specifications may be the preset "streaming enhancement specifications" (e.g., increasing the number of vCPU cores or allocating additional GPU resources to improve encoding performance).

[0052] For the "streaming end" event: the decision logic will start a timer. If the cloud server does not report the "streaming start" event within a preset quiet period (e.g., 30 minutes), a "downgrade task" will be generated to revert its specifications from "streaming enhancement specifications" to "basic specifications" to save resources.

[0053] 3. Perform resource adjustments. The upgrade / downgrade task is submitted to the cloud machine lifecycle service. This service first calls the resource pool management service and threshold control service to confirm that the target host has sufficient resources to complete the specification change (upgrade requires checking, downgrade releases resources). After confirmation, the specification change instruction is issued to the data plane agent of the physical host where the cloud machine is located.

[0054] 4. Online adjustment of cloud machine specifications. The data plane agent utilizes the hot-sizing capabilities of virtualization technology (such as KVM's hot-swappable CPU / memory) to dynamically adjust computing resources without restarting the cloud machine, without affecting its operational status (for upgrades) or with only a temporary impact (for downgrades). After the adjustment is completed, the new cloud machine specification information is reported.

[0055] Step S60: In response to the request to expand the data disk bound to the running cloud machine instance, the corresponding storage volume in the underlying storage system is expanded online, and the operating system inside the cloud machine instance is triggered to expand the file system online to utilize the new capacity.

[0056] In its specific implementation, this embodiment not only enables dynamic allocation but also online expansion of the cloud disk. This includes online capacity expansion of the corresponding storage volume in the underlying storage system, comprising: calling the interface of the underlying storage system to expand the capacity of the storage volume corresponding to the data disk; notifying the physical host where the cloud instance resides; sending a disk capacity change notification to the operating system of the cloud instance via a virtualization channel through an agent on the physical host; and the operating system of the cloud instance responding to the notification by online expanding its file system to cover the newly added storage space.

[0057] For ease of understanding, this embodiment combines... Figure 5 The online cloud disk expansion process in this embodiment is illustrated below.

[0058] See Figure 5 , Figure 5This is a flowchart of the online expansion process for cloud-based mobile cloud storage provided in an embodiment of the present invention.

[0059] 1. Receiving expansion requests. Through the management console, users can select the data disk bound to a cloud machine that is in a "running" state and initiate an online expansion operation, specifying the amount of new capacity to be added (e.g., adding 50GB).

[0060] Second: Storage Layer Expansion. After receiving the expansion command (including the cloud disk volume ID and the new capacity), the storage management service calls the API of the underlying storage system (such as Ceph or distributed block storage) to expand the backend storage volume corresponding to the cloud disk online. This operation is only performed at the storage layer and is transparent to the front-end cloud machine; the cloud machine's operating system is not yet aware of the disk capacity change.

[0061] 3. Notify the cloud machine to refresh. After the storage volume expansion is successful, the storage management service notifies the cloud machine lifecycle service. The lifecycle service then sends a "disk expansion notification" command to the data plane agent of the host where the target cloud machine is located.

[0062] 4. Triggering the identification of new capacity within the cloud server. The data plane agent notifies the operating system within the cloud server through virtualization channels (such as injecting a specific instruction into a virtual machine or sending a SCSI command through the Virtio-SCSI interface) that the physical size of a disk it is connected to has changed.

[0063] 5. Expanding the file system within the cloud server. After receiving a notification of a disk capacity change, the operating system (usually Android / Linux) within the cloud server can call operating system commands (such as resize2fs for the ext4 file system) to expand the file system to the new disk boundary without unmounting the file system or restarting, thus making the newly added 50GB of capacity immediately available.

[0064] Through the above embodiments, the present invention realizes elastic computing resources based on business status and elastic storage resources driven by users. Together, they constitute a refined and multi-dimensional dynamic resource management capability for the "cloud phone" scenario.

[0065] This embodiment maintains and summarizes the inventory information of each resource pool and sets a global resource view for each physical server, based on a threshold for the maximum number of concurrent streaming instances based on its streaming processing capacity. In response to cloud machine creation requests, it decides and creates cloud machine instances based on the thresholds, scheduling policies, and the current load of the target host. During cloud machine operation, its streaming session status is monitored in real time, and online upgrades or downgrades of computing resources are automatically triggered based on the start / end status of the streaming. In response to requests to expand the data disk of a running cloud machine, the underlying storage volume is expanded online, and the file system within the cloud machine is expanded to utilize the increased capacity. This approach achieves unified and efficient scheduling of multiple resource pools, refined resource management of streaming services, and flexible and elastic supply of storage resources.

[0066] Furthermore, embodiments of the present invention also propose a dynamic scheduling and operation management platform for cloud machine resources, characterized in that the platform is configured to execute the methods described above.

[0067] Furthermore, this embodiment of the invention also proposes a storage medium storing a dynamic scheduling and operation management program for cloud server resources. When the dynamic scheduling and operation management program for cloud server resources is executed by a processor, it implements the steps of the dynamic scheduling and operation management method for cloud server resources as described above.

[0068] Furthermore, this embodiment of the invention also proposes a computer program product, which stores a dynamic scheduling and operation management program for cloud server resources. When the dynamic scheduling and operation management program for cloud server resources is executed by a processor, it implements the steps of the dynamic scheduling and operation management method for cloud server resources as described above.

[0069] This application embodiment also provides a dynamic scheduling and operation management device for cloud server resources, including a processor, a communication interface, a memory, and a communication bus. The processor, communication interface, and memory communicate with each other through the communication bus. The memory is used to store the dynamic scheduling and operation management program for cloud server resources. When the processor executes the program stored in the memory, it implements the above-mentioned dynamic scheduling and operation management method for cloud server resources.

[0070] The communication bus mentioned in the aforementioned cloud server resource dynamic scheduling and operation management equipment can be a Peripheral Component Interconnect (PCI) bus or an Extended Industry Standard Architecture (EISA) bus, etc. This communication bus can be divided into address bus, data bus, control bus, etc.

[0071] The communication interface is used for communication between the aforementioned cloud machine resources dynamic scheduling and operation management equipment and other equipment.

[0072] The memory may include random access memory (RAM) or non-volatile memory (NVM), such as at least one disk storage device. Optionally, the memory may also be at least one storage device located remotely from the aforementioned processor.

[0073] The processors mentioned above can be general-purpose processors, including central processing units (CPUs), network processors (NPs), etc.; they can also be digital signal processors (DSPs), application-specific integrated circuits (ASICs), field-programmable gate arrays (FPGAs), or other programmable logic devices, discrete gate or transistor logic devices, or discrete hardware components.

[0074] In the above embodiments, implementation can be achieved entirely or partially through software, hardware, firmware, or any combination thereof. When implemented using software, it can be implemented entirely or partially in the form of a computer program product. The computer program product includes one or more computer instructions. When the computer program instructions are loaded and executed on a computer, all or part of the processes or functions described in the embodiments of this application are generated. The computer can be a general-purpose computer, a special-purpose computer, a computer network, or other programmable device. The computer instructions can be stored in a computer-readable storage medium or transmitted from one computer-readable storage medium to another. For example, the computer instructions can be transmitted from one website, computer, server, or data center to another website, computer, server, or data center via wired (e.g., coaxial cable, fiber optic, digital subscriber line (DSL)) or wireless (e.g., infrared, wireless, microwave, etc.) means. The computer-readable storage medium can be any available medium that a computer can access or a data storage device such as a server or data center that integrates one or more available media. The available medium can be a magnetic medium (e.g., floppy disk, hard disk, magnetic tape), an optical medium (e.g., DVD), or a semiconductor medium (e.g., solid state disk (SSD)).

[0075] It should be noted that, in this document, relational terms such as "first" and "second" are used only to distinguish one entity or operation from another, and do not necessarily require or imply any such actual relationship or order between these entities or operations. Furthermore, the terms "comprising," "including," or any other variations thereof are intended to cover non-exclusive inclusion, such that a process, method, article, or apparatus that comprises a list of elements includes not only those elements but also other elements not expressly listed, or elements inherent to such a process, method, article, or apparatus. Without further limitations, an element defined by the phrase "comprising one..." does not exclude the presence of other identical elements in the process, method, article, or apparatus that includes said element.

[0076] The various embodiments in this specification are described in a related manner. Similar or identical parts between embodiments can be referred to mutually. Each embodiment focuses on describing the differences from other embodiments. In particular, the system embodiments are basically similar to the method embodiments, so the description is relatively simple; relevant parts can be referred to the descriptions of the method embodiments.

[0077] The above embodiments are only used to illustrate the technical solutions of the present invention, and are not intended to limit it. Although the present invention 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 spirit and scope of the technical solutions of the embodiments of the present invention.

[0078] It should be understood that the above are merely illustrative examples and do not constitute any limitation on the technical solutions of the present invention. In specific applications, those skilled in the art can make settings as needed, and the present invention does not impose any restrictions on this.

[0079] It should be noted that the workflow described above is merely illustrative and does not limit the scope of protection of this invention. In practical applications, those skilled in the art can select some or all of the workflow to achieve the purpose of this embodiment according to actual needs, and no restrictions are imposed here.

[0080] In addition, for technical details not described in detail in this embodiment, please refer to the dynamic scheduling and operation management method of cloud machine resources provided in any embodiment of the present invention, which will not be repeated here.

[0081] Furthermore, it should be noted that, in this document, the terms "comprising," "including," or any other variations thereof are intended to cover non-exclusive inclusion, such that a process, method, article, or system that comprises a list of elements includes not only those elements but also other elements not expressly listed, or elements inherent to such a process, method, article, or system. Unless otherwise specified, an element defined by the phrase "comprising one..." does not exclude the presence of other identical elements in the process, method, article, or system that includes that element.

[0082] The sequence numbers of the above embodiments of the present invention are for descriptive purposes only and do not represent the superiority or inferiority of the embodiments.

[0083] Through the above description of the embodiments, those skilled in the art can clearly understand that the methods of the above embodiments can be implemented by means of software plus necessary general-purpose hardware platforms. Of course, they can also be implemented by hardware, but in many cases the former is a better implementation method. Based on this understanding, the technical solution of the present invention, or the part that contributes to the prior art, can be embodied in the form of a software product. This computer software product is stored in a storage medium (such as read-only memory (ROM) / RAM, magnetic disk, optical disk) and includes several instructions to cause a terminal device (which may be a mobile phone, computer, server, or network device, etc.) to execute the methods described in the various embodiments of the present invention.

[0084] The above are merely preferred embodiments of the present invention and do not limit the scope of the patent. Any equivalent structural or procedural transformations made based on the description and drawings of the present invention, or direct or indirect applications in other related technical fields, are similarly included within the scope of patent protection of the present invention.

[0085] It is understood that the system provided in the embodiments of the present invention corresponds to the method provided in the embodiments of the present invention, and the explanation, examples and beneficial effects of the relevant content can be referred to the corresponding parts of the above methods.

Claims

1. A method for dynamic scheduling and operation management of cloud server resources, characterized in that, The dynamic scheduling and operation management method for cloud server resources is applied to a platform that manages multiple physically dispersed and logically isolated resource pools in a unified manner. The method includes: Maintain a global resource view that aggregates the resource inventory information of physical servers in each resource pool, and set a maximum concurrent streaming instance threshold for each physical server based on the streaming processing capacity of the global resource view. In response to the cloud machine creation request, a target physical host that has not exceeded the threshold and has sufficient resources is determined based on at least the maximum number of concurrent streaming instances contained in the global resource view, the preset scheduling policy, and the current number of concurrent streaming instances of the target physical server, and a cloud machine instance is created on the target physical host. During the operation of the cloud machine instance, the streaming session status of each cloud machine instance is monitored in real time. The streaming session status includes the streaming start status and the streaming end status. When the target cloud machine instance is detected to be in the streaming start state, the computing resources of the target cloud machine instance are upgraded online. When the target cloud server instance is detected to be in a streaming termination state and meets the preset conditions, the computing resources of the target cloud server instance are reduced online. In response to a request to expand the data disk bound to a running cloud instance, the corresponding storage volume in the underlying storage system is expanded online, and the operating system inside the cloud instance is triggered to expand the file system online to utilize the increased capacity.

2. The method for dynamic scheduling and operation management of cloud server resources as described in claim 1, characterized in that, The preset scheduling strategy includes at least a load balancing strategy for selecting a low-load resource pool when no region is specified, and a strategy for selecting low-load hosts that have not exceeded the threshold within the resource pool of a specified region.

3. The method for dynamic scheduling and operation management of cloud server resources as described in claim 1, characterized in that, The online upgrade refers to adjusting the computing specifications from the basic configuration to the streaming enhancement configuration, while the online downgrade refers to adjusting the computing specifications from the streaming enhancement configuration to the basic configuration.

4. The method for dynamic scheduling and operation management of cloud server resources as described in claim 1, characterized in that, The maximum number of concurrent streaming instances of the physical server is determined dynamically or statically by at least one of the following factors: server model, GPU encoding performance, or network bandwidth.

5. The method for dynamic scheduling and operation management of cloud server resources as described in claim 3, characterized in that, The response time from responding to the cloud machine creation request to completing the creation of the cloud machine instance on the target physical host does not exceed a preset time threshold, and the online upgrade or downgrade is completed without restarting the target cloud machine instance.

6. The method for dynamic scheduling and operation management of cloud server resources as described in claim 1, characterized in that, The online capacity expansion of the corresponding storage volume in the underlying storage system includes: Call the interface of the underlying storage system to expand the capacity of the storage volume corresponding to the data disk; notify the physical host where the cloud machine instance is located; The disk capacity change notification is sent to the operating system of the cloud machine instance through the agent on the physical host using the virtualization channel; In response to the notification, the operating system of the cloud instance expands its file system online to cover the newly added storage space.

7. A dynamic scheduling and operation management platform for cloud server resources, characterized in that, The platform is configured to perform the method as described in any one of claims 1 to 6.

8. A device for dynamic scheduling and operation management of cloud server resources, characterized in that, The cloud server resource dynamic scheduling and operation management device includes: a memory, a processor, and a cloud server resource dynamic scheduling and operation management program stored in the memory and executable on the processor. The cloud server resource dynamic scheduling and operation management program is configured to implement the steps of the cloud server resource dynamic scheduling and operation management method as described in any one of claims 1 to 6.

9. A storage medium, characterized in that, The storage medium stores a dynamic scheduling and operation management program for cloud server resources. When the processor executes the dynamic scheduling and operation management program for cloud server resources, it implements the steps of the dynamic scheduling and operation management method for cloud server resources as described in any one of claims 1 to 6.

10. A computer program product, characterized in that, The computer program product stores a dynamic scheduling and operation management program for cloud server resources. When the dynamic scheduling and operation management program for cloud server resources is executed by the processor, it implements the steps of the dynamic scheduling and operation management method for cloud server resources as described in any one of claims 1 to 6.

Images