A scheduling method, device, and medium for multi-core cloud computing

Abstract

This application relates to the field of cloud platform technology and discloses a scheduling method, device, and medium for multi-core cloud computing. When scheduling cloud hosts, each physical host is first scored, and cloud hosts are prioritized for scheduling to physical hosts with higher scores. After each scheduling, the characteristic information of each physical host is updated, and the scores are recalculated to select the host with the best performance. Unscheduled cloud hosts are then assigned to this host. This process is repeated until all cloud hosts are scheduled. It can be seen that by using this technical solution, cloud hosts are always scheduled to the best-performing physical host, thus achieving balanced use of physical hosts and efficient utilization of their performance.

Description

Technical Field

[0001] This application relates to the field of cloud platform technology, and in particular to a scheduling method, apparatus and medium for one cloud and multiple chips. Background Technology

[0002] As the core product of the semiconductor industry, chips can be said to be the foundation of the entire information age. In today's world where people's lives are becoming increasingly digitalized, from small items like mobile phones and laptops used for work, to routers, refrigerators, air conditioners, and televisions used at home, to large-scale traffic signal systems and vehicle control systems, all rely on the chip as the "brain" to maintain the normal operation of the entire system by receiving information and issuing instructions.

[0003] In the field of cloud computing, especially in the general-purpose server market relied upon by hybrid cloud scenarios, there are chips with various instruction set types such as x86, ARM, RISC-V, and MIPS. Currently, some cloud vendors offer solutions by building separate clouds for each type of chip cluster, and then managing them uniformly at the software level through a cloud management platform after forming a multi-cloud ecosystem. However, due to significant differences in computing power and characteristics between server chips of different architectures or from different vendors, after scheduling cloud hosts to physical hosts, the performance differences of the physical hosts result in low-performance hosts being configured with multiple cloud hosts, while high-performance physical hosts are configured with fewer cloud hosts. This leads to uneven utilization of physical host performance, resulting in resource waste. Furthermore, the multi-cloud structure makes unified resource scheduling difficult, failing to fully utilize the peak and off-peak periods of different services for elastic usage.

[0004] Therefore, it is evident that how to fully and rationally utilize the resources of each host under the unified management of the cloud platform, and achieve balanced utilization of physical host performance and efficient scheduling of each cloud host, is a problem that urgently needs to be solved by those skilled in the art. Summary of the Invention

[0005] The purpose of this application is to provide a scheduling method, device, and medium for multi-core cloud computing, so as to make full and reasonable use of the resources of each host under the unified management of the cloud platform, and achieve balanced utilization of the performance of physical hosts and efficient scheduling of each cloud host.

[0006] To address the aforementioned technical problems, this application provides a scheduling method for multiple chips within a single cloud, comprising:

[0007] Confirm the number of cloud servers to be created based on user instructions;

[0008] Obtain performance characteristic information of each physical host in the cluster;

[0009] Each physical host is scored based on the performance characteristics information;

[0010] The target host with the highest performance is selected based on the scoring results;

[0011] The cloud host to be scheduled is scheduled to the target host;

[0012] Update the performance characteristic information of the physical host; return to the step of scoring each physical host based on the performance characteristic information, until the scheduling of all cloud hosts is completed.

[0013] On the one hand, the performance characteristics information includes the computing power of the physical host, the allocatable central processing unit, and the allocatable memory.

[0014] On the other hand, after all cloud hosts have been scheduled, the process also includes:

[0015] The request to launch a cloud host from an image is sent to the computing service of each physical host, so that after receiving the request, the computing service requests the image management service to obtain the corresponding image and launches the cloud host on the physical host according to the corresponding image.

[0016] On the other hand, it also includes:

[0017] Define image names so that the image management service can manage images from all vendors based on the image names;

[0018] Furthermore, prior to the step of obtaining the performance characteristic information of each physical host within the cluster, the following steps are also included:

[0019] Confirm the image name according to user instructions.

[0020] On the one hand, requesting the image management service to obtain the corresponding image includes:

[0021] Obtain the vendor and architecture information of the physical host where the computing service is located;

[0022] Based on the vendor and architecture information and the image name, request the image management service to obtain the corresponding image.

[0023] On the one hand, in the image management service, the image name cannot be repeated.

[0024] On the one hand, the target host is:

[0025] The host with the highest score is ranked based on the computing power, allocable central processing unit, and allocable memory of each physical host.

[0026] To address the aforementioned technical problems, this application also provides a multi-core scheduling device for a single cloud, the device comprising:

[0027] The confirmation module is used to confirm the number of cloud servers to be created based on user instructions.

[0028] The acquisition module is used to acquire performance characteristic information of each physical host in the cluster.

[0029] The scoring module is used to score each physical host based on the performance characteristic information;

[0030] The filtering module is used to filter out the target host with the highest performance based on the scoring results;

[0031] The scheduling module is used to schedule the cloud host to be scheduled to the target host;

[0032] The processing module is used to update the performance characteristic information of the physical host; return to the step of scoring each physical host according to the performance characteristic information, until the scheduling of all cloud hosts is completed.

[0033] In some embodiments, the performance characteristic information acquired by the acquisition module includes the computing power of the physical host, the allocatable central processing unit, and the allocatable memory.

[0034] On the one hand, the scheduling device for one cloud and multiple chips also includes:

[0035] The sending module is used to send requests to launch cloud hosts from images to the computing services of each physical host, so that after receiving the request, the computing service can request the image management service to obtain the corresponding image and launch the cloud host on the physical host according to the corresponding image.

[0036] On the other hand, it also includes:

[0037] The definition module is used to define image names so that the image management service can manage images from all vendors based on the image names;

[0038] Furthermore, it also includes:

[0039] The first confirmation module is used to confirm the image name based on user instructions.

[0040] On the other hand, the sending module requests the corresponding image from the image management service, including:

[0041] Obtain the vendor and architecture information of the physical host where the computing service is located;

[0042] Based on the vendor and architecture information, as well as the image name, request the image management service to obtain the corresponding image.

[0043] To address the aforementioned technical problems, this application also provides another cloud-multi-chip scheduling device, including a memory for storing computer programs;

[0044] A processor is used to implement the steps of the one-cloud-multi-core scheduling method described above when executing the computer program.

[0045] To address the aforementioned technical problems, this application also provides a computer-readable storage medium storing a computer program, which, when executed by a processor, implements the steps of the aforementioned one-cloud-multi-core scheduling method.

[0046] The multi-chip scheduling method provided in this application determines the number of cloud hosts to be created based on user instructions; obtains the performance characteristics of each physical host in the cluster; scores each physical host based on the performance characteristics; selects the target host with the highest performance based on the scoring results; schedules the cloud hosts to be scheduled to the target host; updates the performance characteristics of the physical hosts; and returns to the step of scoring each physical host based on the performance characteristics, until the scheduling of all cloud hosts is completed. Compared with the current technology, which builds a separate cloud for each type of chip cluster, and can achieve unified scheduling of physical host performance resources, the method addresses the issue that differences in computing power and characteristics prevent the full utilization of peak and trough periods for different services, thus wasting physical host performance resources. The cloud-multi-core scheduling method provided in this application first scores each physical host during cloud host scheduling, prioritizing scheduling cloud hosts to physical hosts with higher scores. After each scheduling, the characteristic information of each physical host is updated, and the hosts are re-scored to identify the best-performing host. Unscheduled cloud hosts are then assigned to this best-performing host. This process is repeated until all cloud hosts are scheduled. It can be seen that this technical solution achieves balanced use of physical hosts and efficient utilization of their performance by consistently scheduling cloud hosts to the best-performing physical host.

[0047] Furthermore, the scheduling device and medium for one cloud and multiple cores provided in this application correspond to the scheduling method for one cloud and multiple cores described above, and have the same effect. Attached Figure Description

[0048] To more clearly illustrate the embodiments of this application, the accompanying drawings used in the embodiments will be briefly introduced below. Obviously, the drawings described below are only some embodiments of this application. For those skilled in the art, other drawings can be obtained based on these drawings without creative effort.

[0049] Figure 1 A flowchart illustrating a scheduling method for a single cloud with multiple cores, provided in an embodiment of this application;

[0050] Figure 2 A schematic diagram illustrating a scheduling method for a single cloud and multiple chips provided in an embodiment of this application;

[0051] Figure 3 A structural diagram of a multi-core scheduling device for a single cloud provided in an embodiment of this application;

[0052] Figure 4 This is a structural diagram of another cloud-multi-core scheduling device provided in an embodiment of this application. Detailed Implementation

[0053] The technical solutions of the embodiments of this application will be clearly and completely 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 protection scope of this application.

[0054] As the core product of the semiconductor industry, chips can be said to be the foundation of the entire information age. In today's world where people's lives are becoming increasingly digitalized, from small items like mobile phones and laptops used for work, to routers, refrigerators, air conditioners, and televisions used at home, to large-scale traffic signal systems and vehicle control systems, all rely on the chip as the "brain" to maintain the normal operation of the entire system by receiving information and issuing instructions.

[0055] In the field of cloud computing, especially in the general-purpose server market relied upon by hybrid cloud scenarios, there are chips with various instruction set types such as x86, ARM, RISC-V, and MIPS. Currently, some cloud vendors offer solutions that build separate clouds for each type of chip cluster, then manage them uniformly at the software level through a cloud management platform after forming a multi-cloud ecosystem. However, due to significant differences in computing power and characteristics between server chips of different architectures or from different vendors, the multi-cloud structure makes it difficult to uniformly schedule resources and fully utilize the peak and off-peak periods of different services for elastic usage.

[0056] Therefore, it is evident that how to fully and rationally utilize the resources of each host under the unified management of the cloud platform, and achieve efficient scheduling and balanced utilization of computing resources, is a problem that urgently needs to be solved by those skilled in the art.

[0057] The core of this application is to provide a scheduling method, device, and medium for multi-core cloud computing, so as to make full and reasonable use of the resources of each host under the unified management of the cloud platform, and achieve balanced utilization of the performance of physical hosts and efficient scheduling of each cloud host.

[0058] The term "one cloud, multiple cores" refers to using a single cloud operating system to manage a cluster of hardware servers with different architectures. First, it should be noted that the one cloud, multiple cores scheduling method provided in this application can be applied to a one cloud, multiple cores scheduling device. This scheduling device can be a computer with a cloud management platform deployed on it. This allows for software-level scheduling of the one cloud, multiple cores structure provided in this application through the cloud management platform, enabling the batch creation and scheduling of cloud hosts based on user instructions.

[0059] To enable those skilled in the art to better understand the present application, the present application will be further described in detail below with reference to the accompanying drawings and specific embodiments.

[0060] Figure 1 A flowchart illustrating a scheduling method for a multi-core cloud architecture provided in this application is shown below. Figure 1 As shown, the method includes:

[0061] S10: Confirm the number of cloud servers to be created based on user instructions;

[0062] S11: Obtain performance characteristic information of each physical host in the cluster;

[0063] S12: Score each physical host based on performance characteristics;

[0064] S13: Select the target host with the highest performance based on the scoring results;

[0065] S14: Schedule the cloud host to be scheduled to the target host;

[0066] S15: Update the performance characteristics information of the physical host;

[0067] S16: Determine whether the scheduling of all cloud hosts has been completed. If yes, end the process; otherwise, return to step S12.

[0068] In this application, the user first needs to input user commands into the cloud management platform, i.e., the API service, to request the number of cloud hosts to be created from the control service on the multi-core scheduling device. After receiving the request, the control service will request scheduling from the scheduling service. The scheduling service will then obtain the performance characteristics information of all physical hosts in the cluster. This performance characteristics information reflects the performance indicators of each physical host, including computing power, allocable central processing units (CPUs), and other information. This completes the scheduling preparation work.

[0069] Then, the scheduling service scores each physical host based on performance characteristics. In practice, these characteristics include the physical host's computing power, available CPU, and available memory. By using these metrics as scoring criteria, the host with the highest score can be selected as the target host. This target host can be considered the highest-performing physical host among all physical hosts, and in this application, the cloud host to be scheduled will be scheduled to this target host. It is understandable that, during cloud server scheduling, to better achieve balanced utilization of physical server performance and efficient scheduling of each cloud server, cloud servers can be scheduled to physical servers one by one. Therefore, in this embodiment, after selecting the target server with the highest performance based on the scoring results, an unscheduled cloud server is scheduled to that physical server. It is understood that after a cloud server is scheduled to a physical server, the performance characteristics of the physical server, such as computing power, allocable CPU, and allocable memory, will change. Therefore, it is necessary to update the performance characteristics of the physical server, return to the step of scoring each physical server based on the performance characteristics, re-score, select the target server with the highest performance at that moment, and then schedule the next unscheduled cloud server to this physical server. This process is repeated until all cloud servers are scheduled. To achieve better allocation results, cloud servers in this embodiment can be scheduled one by one; in other embodiments, several cloud servers can be scheduled together at a time.

[0070] As can be seen, by adopting this technical solution, cloud hosts are scheduled to the highest-performing physical hosts in batches, making full use of the performance of the physical hosts. Under the unified management of the cloud platform, the resources of each host are fully and rationally utilized, achieving balanced utilization of physical host performance and efficient scheduling of each cloud host.

[0071] In addition, when determining whether the scheduling of all cloud hosts has been completed, it can be confirmed by checking whether the number of cloud hosts created and the number of cloud hosts that have been scheduled are consistent.

[0072] The multi-chip scheduling method for a single cloud provided in this application involves: confirming the number of cloud hosts to be created based on user instructions; obtaining performance characteristic information of each physical host within the cluster; scoring each physical host based on the performance characteristic information; selecting the target host with the highest performance based on the scoring results; scheduling the cloud hosts to be scheduled to the target host; updating the performance characteristic information of the physical hosts; and returning to the step of scoring each physical host based on the performance characteristic information, until the scheduling of all cloud hosts is completed. Compared to the current technology where each type of chip cluster is built into a separate cloud, which can achieve unified scheduling of physical host performance resources, the differences in computing power and characteristics prevent the full utilization of peak and off-peak periods for different services, thus wasting physical host performance resources. The cloud-multi-core scheduling method provided in this application first scores each physical host during cloud host scheduling, prioritizing scheduling cloud hosts to physical hosts with higher scores. After each scheduling, the characteristic information of each physical host is updated, and the hosts are re-scored to identify the best-performing host. Unscheduled cloud hosts are then assigned to this best-performing host. This process is repeated until all cloud hosts are scheduled. It can be seen that this technical solution achieves balanced use of physical hosts and efficient utilization of their performance by consistently scheduling cloud hosts to the best-performing physical host.

[0073] It is understandable that creating a cloud server requires not only scheduling it to a physical server but also allocating an image. Therefore, in this embodiment, after scheduling all cloud servers is complete, the process further includes:

[0074] The request to start a cloud host from an image is sent to the computing service of each physical host. Upon receiving the request, the computing service requests the image management service of the cloud platform to obtain the corresponding image and starts the cloud host on the physical host based on the corresponding image.

[0075] In practice, different vendors use different images, and different cloud-multi-chip architectures also require different images. Therefore, how to achieve efficient image allocation based on the cloud-multi-chip scheduling method provided in this application is a problem that needs to be solved.

[0076] Therefore, in this embodiment, the image management service also needs to be modified. Specifically, this includes:

[0077] Define image names so that the image management service can manage images from all vendors based on the image names;

[0078] Furthermore, before the step of obtaining the performance characteristics information of each physical host within the cluster, the following steps are also included:

[0079] Confirm the image name according to user instructions.

[0080] Before the transformation, many vendors' images were uniquely identified by an image ID, and image names could be duplicated. After the transformation, image names are no longer duplicated, and images from different vendors and with different architectures can be managed under the same image name. It can be understood that there is a one-to-one correspondence between the image name and the "one cloud, multiple chips" structure; that is, one image name corresponds to one cloud, and the images required for various chips within that multi-cloud can be obtained through that image name. Furthermore, requesting the image management service to obtain the corresponding image includes:

[0081] Obtain the vendor and architecture information of the physical host where the computing service is located;

[0082] Based on the vendor and architecture information, as well as the image name, request the image management service to obtain the corresponding image.

[0083] In summary, as an embodiment, this embodiment provides a specific scenario as an example for easy understanding.

[0084] For example, after deploying a multi-core cloud environment, cloud hosts are distributed to users in batches. Assume the cloud environment has one Phytium physical host (FT), one Hygon physical host (HG), and one Kunpeng physical host (KP); assume each Phytium physical host has a computing power of s1, each Hygon physical host has a computing power of s2, and each Kunpeng physical host has a computing power of s3; assume each Phytium physical host can be allocated m1 memory and c1 CPU cores, each Hygon physical host can be allocated m2 memory and c2 CPU cores, and each Kunpeng physical host can be allocated m3 memory and c3 CPU cores. And when creating cloud hosts in batches, the image name selected is img_test.

[0085] First, the cloud platform user selects the image img_test and distributes it to n cloud hosts in batches; the cloud platform scheduling service receives the batch creation request for cloud hosts and obtains the performance characteristics information of all physical hosts in the cluster to complete the preparatory work for scheduling.

[0086] The cloud platform scheduling service obtains information such as computing power, available CPUs, and available memory for each host based on performance characteristics. Then, it scores each host according to the current conditions, currently using only the physical host's standard computing power, CPU, and memory as scoring criteria.

[0087] The scheduling service initially scores the Phytium, Hygon, and Kunpeng hosts, with their scores represented by Score1, Score2, and Score3 respectively. Then:

[0088] Score1 = s1*w1 + m1*w2 + c1*w3;

[0089] Score2 = s2*w1 + m2*w2 + c2*w3;

[0090] Score3 = s3*w1 + m3*w2 + c3*w3;

[0091] Where w1, w2, and w3 are the weight values ​​for physical host computing power, allocable memory, and allocable central processing unit, respectively.

[0092] Obtain the score for each physical host. Assuming Score1 > Score2 > Score3, the best host for the first selection is FT. Unscheduled cloud hosts are then scheduled to FT. Simultaneously, the computing power, allocable CPUs, and allocable memory of the FT host are updated to s1', c1', and m1', respectively. Then, a second round of scoring is performed to select the next target host. This process is repeated until all cloud hosts are scheduled. Once the scheduling service has selected hosts for all n cloud hosts, if n1 cloud hosts are scheduled to the FT physical host, n2 cloud hosts to the HG physical host, and n3 cloud hosts to the KP physical host (n = n1 + n2 + n3), the control service will send the requests to start cloud hosts from the img_test image to the computing services of each physical host.

[0093] After receiving a request to launch a cloud server from the `img_test` image, the compute service retrieves the vendor and architecture information of the physical host. Based on the vendor and architecture information and the image name, it requests the corresponding image from the image management service. For example, the cloud server image for a Phytium host is `image_test_FT_arrch64`, for a Hygon host it's `image_test_HG_x86`, and for a Kunpeng host it's `image_test_KP_arrch64`. Once the compute service obtains the cloud server image corresponding to the vendor's architecture, it can launch the cloud server on the physical host using the corresponding image, enabling batch creation of cloud servers. Figure 2 This is a schematic diagram of a scheduling method for one cloud and multiple chips provided in an embodiment of this application.

[0094] The above embodiments have described the scheduling method for a single cloud and multiple chips in detail. This application also provides embodiments of a scheduling device for a single cloud and multiple chips. It should be noted that this application describes the embodiments of the device part from two perspectives: one is based on functional modules, and the other is based on hardware.

[0095] Figure 3 A structural diagram of a multi-core scheduling device for a single cloud provided in this application embodiment is shown below. Figure 3 As shown, the device includes:

[0096] Confirmation module 10 is used to confirm the number of cloud hosts to be created based on user instructions;

[0097] Module 11 is used to obtain performance characteristic information of each physical host in the cluster;

[0098] The scoring module 12 is used to score each physical host based on performance characteristic information;

[0099] Filtering module 13 is used to filter out the target host with the highest performance based on the scoring results;

[0100] Scheduling module 14 is used to schedule the cloud host to be scheduled to the target host;

[0101] Processing module 15 is used to update the performance characteristic information of physical hosts; return the steps of scoring each physical host based on the performance characteristic information, until the scheduling of all cloud hosts is completed.

[0102] In some embodiments, the performance characteristic information acquired by the acquisition module includes the computing power of the physical host, the allocatable central processing unit, and the allocatable memory.

[0103] In other embodiments, the scheduling device for one cloud and multiple chips further includes:

[0104] The sending module is used to send requests to launch cloud hosts from images to the computing services of each physical host, so that after receiving the request, the computing service can request the image management service to obtain the corresponding image and launch the cloud host on the physical host according to the corresponding image.

[0105] Also includes:

[0106] The definition module is used to define image names so that the image management service can manage images from all vendors based on the image names;

[0107] Furthermore, it also includes:

[0108] The first confirmation module is used to confirm the image name based on user instructions.

[0109] The sending module requests the corresponding image from the image management service, including:

[0110] Obtain the vendor and architecture information of the physical host where the computing service is located;

[0111] Based on the vendor and architecture information, as well as the image name, request the image management service to obtain the corresponding image.

[0112] Since the embodiments of the apparatus and the embodiments of the method correspond to each other, please refer to the description of the embodiments of the method for the embodiments of the apparatus, which will not be repeated here.

[0113] The multi-chip scheduling device provided in this application confirms the number of cloud hosts to be created according to user instructions; obtains the performance characteristic information of each physical host in the cluster; scores each physical host according to the performance characteristic information; selects the target host with the highest performance based on the scoring results; schedules the cloud hosts to be scheduled to the target host; updates the performance characteristic information of the physical hosts; and returns to the step of scoring each physical host according to the performance characteristic information, until the scheduling of all cloud hosts is completed. Compared with the current technology, where each type of chip cluster builds a separate cloud, which can achieve unified scheduling of physical host performance resources, the differences in computing power and characteristics make it impossible to fully utilize the peak and trough times of different services for elastic use, thus wasting the performance resources of physical hosts. The cloud-multi-core scheduling device provided in this application first scores each physical host when scheduling cloud hosts. Cloud hosts are prioritized for scheduling to physical hosts with higher scores. After each scheduling, the feature information of each physical host is updated, and the hosts are re-scored. The host with the best performance is then selected, and any unscheduled cloud hosts are assigned to it. This process is repeated until all cloud hosts are scheduled. It can be seen that by using this technical solution, cloud hosts are always scheduled to the best-performing physical host, thus achieving balanced use of physical hosts and efficient utilization of their performance.

[0114] Figure 4 A structural diagram of another cloud-multi-core scheduling device provided in this application embodiment is shown below. Figure 4 As shown, the device includes: a memory 20 for storing computer programs;

[0115] The processor 21 is used to implement the steps of the one-cloud-multi-core scheduling method as described in the above embodiments when executing a computer program.

[0116] The multi-chip scheduling device provided in this embodiment can include, but is not limited to, smartphones, tablets, laptops, or desktop computers.

[0117] The processor 21 may include one or more processing cores, such as a quad-core processor or an octa-core processor. The processor 21 may be implemented using at least one of the following hardware forms: Digital Signal Processor (DSP), Field-Programmable Gate Array (FPGA), or Programmable Logic Array (PLA). The processor 21 may also include a main processor and a coprocessor. The main processor, also known as the Central Processing Unit (CPU), is used to process data in the wake-up state; the coprocessor is a low-power processor used to process data in the standby state. In some embodiments, the processor 21 may integrate a Graphics Processing Unit (GPU), which is responsible for rendering and drawing the content to be displayed on the screen. In some embodiments, the processor 21 may also include an Artificial Intelligence (AI) processor, which is used to handle computational operations related to machine learning.

[0118] The memory 20 may include one or more computer-readable storage media, which may be non-transitory. The memory 20 may also include high-speed random access memory and non-volatile memory, such as one or more disk storage devices or flash memory devices. In this embodiment, the memory 20 is used to store at least the following computer program 201, which, after being loaded and executed by the processor 21, is capable of implementing the relevant steps of the one-cloud-multi-core scheduling method disclosed in any of the foregoing embodiments. In addition, the resources stored in the memory 20 may also include an operating system 202 and data 203, and the storage method may be temporary or permanent storage. The operating system 202 may include Windows, Unix, Linux, etc. The data 203 may include, but is not limited to, performance characteristic information.

[0119] In some embodiments, the multi-core scheduling device may further include a display screen 22, an input / output interface 23, a communication interface 24, a power supply 25, and a communication bus 26.

[0120] Those skilled in the art will understand that Figure 4 The structure shown does not constitute a limitation on the scheduling device for a multi-core cloud, and may include more or fewer components than shown.

[0121] The multi-core scheduling device for a single cloud provided in this application includes a memory and a processor. When the processor executes the program stored in the memory, it can implement the following methods: confirm the number of cloud hosts to be created according to user instructions; obtain the performance characteristic information of each physical host in the cluster; score each physical host according to the performance characteristic information; select the target host with the highest performance according to the scoring results; schedule the cloud host to be scheduled to the target host; update the performance characteristic information of the physical host; return to the step of scoring each physical host according to the performance characteristic information, until the scheduling of all cloud hosts is completed.

[0122] The multi-chip scheduling device provided in this application confirms the number of cloud hosts to be created according to user instructions; obtains the performance characteristic information of each physical host in the cluster; scores each physical host according to the performance characteristic information; selects the target host with the highest performance based on the scoring results; schedules the cloud hosts to be scheduled to the target host; updates the performance characteristic information of the physical hosts; and returns to the step of scoring each physical host according to the performance characteristic information, until the scheduling of all cloud hosts is completed. Compared with the current technology, where each type of chip cluster builds a separate cloud, which can achieve unified scheduling of physical host performance resources, the differences in computing power and characteristics make it impossible to fully utilize the peak and trough times of different services for elastic use, thus wasting the performance resources of physical hosts. The cloud-multi-core scheduling device provided in this application first scores each physical host when scheduling cloud hosts. Cloud hosts are prioritized for scheduling to physical hosts with higher scores. After each scheduling, the feature information of each physical host is updated, and the hosts are re-scored. The host with the best performance is then selected, and any unscheduled cloud hosts are assigned to it. This process is repeated until all cloud hosts are scheduled. It can be seen that by using this technical solution, cloud hosts are always scheduled to the best-performing physical host, thus achieving balanced use of physical hosts and efficient utilization of their performance.

[0123] Finally, this application also provides an embodiment corresponding to a computer-readable storage medium. The computer-readable storage medium stores a computer program, which, when executed by a processor, implements the steps described in the above method embodiments.

[0124] It is understood that if the methods in the above embodiments are implemented as software functional units and sold or used as independent products, they can be stored in a computer-readable storage medium. Based on this understanding, the technical solution of this application, in essence, or the part that contributes to the prior art, or all or part of the technical solution, can be embodied in the form of a software product. This computer software product is stored in a storage medium and executes all or part of the steps of the methods described in the various embodiments of this application. The aforementioned storage medium includes various media capable of storing program code, such as USB flash drives, portable hard drives, read-only memory (ROM), random access memory (RAM), magnetic disks, or optical disks.

[0125] The computer-readable storage medium provided in this application embodiment confirms the number of cloud hosts to be created according to user instructions; obtains the performance characteristic information of each physical host in the cluster; scores each physical host according to the performance characteristic information; selects the target host with the highest performance based on the scoring results; schedules the cloud hosts to be scheduled to the target host; updates the performance characteristic information of the physical hosts; and returns to the step of scoring each physical host according to the performance characteristic information, until the scheduling of all cloud hosts is completed. Compared with the current technology, where each type of chip cluster builds a separate cloud, which can achieve unified scheduling of physical host performance resources, the differences in computing power and characteristics make it impossible to fully utilize the peak and trough times of different services for elastic use, thus wasting the performance resources of physical hosts. The computer-readable storage medium provided in this application, when scheduling cloud hosts, first scores each physical host and prioritizes scheduling cloud hosts to physical hosts with higher scores. Each time a cloud host is scheduled, the characteristic information of each physical host is updated, and each physical host is re-scored. The host with the best performance among all physical hosts at that time is selected, and the unscheduled cloud hosts are scheduled to that physical host. The above steps are repeated until all cloud hosts are scheduled. It can be seen that by adopting this technical solution, cloud hosts are scheduled to the best-performing physical host each time, thereby achieving balanced use of physical hosts and efficient utilization of their performance.

[0126] The foregoing has provided a detailed description of the scheduling method, apparatus, and medium for multi-core cloud computing provided in this application. The various embodiments in the specification are described in a progressive manner, with each embodiment focusing on its differences from other embodiments. Similar or identical parts between embodiments can be referred to interchangeably. For the apparatus disclosed in the embodiments, since it corresponds to the method disclosed in the embodiments, the description is relatively simple; relevant parts can be referred to in the method section. It should be noted that those skilled in the art can make several improvements and modifications to this application without departing from the principles of this application, and these improvements and modifications also fall within the protection scope of the claims of this application.

[0127] It should also be noted that, in this specification, 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.

Claims

1. A cloud-multiple-core scheduling method, characterized in that, The method comprises the following steps: Confirming the number of created cloud hosts according to user instructions; Obtaining performance characteristic information of each physical host in a cluster; Scoring each physical host according to the performance characteristic information; Selecting a target host with the highest performance according to the scoring results; Scheduling a cloud host to be scheduled to the target host; Updating the performance characteristic information of the physical host; Returning to the step of scoring each physical host according to the performance characteristic information until the scheduling of all cloud hosts is completed; wherein the performance characteristic information comprises the computing power, allocable central processing unit and allocable memory of the physical host; When judging whether the scheduling of all cloud hosts is completed, confirming whether the number of created cloud hosts is consistent with the number of cloud hosts that have completed the scheduling; After the scheduling of all cloud hosts is completed, the method further comprises the following steps: Sending a request for starting a cloud host by an image to the computing service of each physical host, so that, after the computing service receives the request, the computing service requests an image management service to obtain a corresponding image and starts a cloud host on the physical host according to the corresponding image; The method further comprises the following steps: Defining an image name, so that the image management service manages images of all manufacturers according to the image name; Further, before the step of obtaining the performance characteristic information of each physical host in the cluster, the method further comprises the following steps: Confirming the image name according to user instructions; The step of requesting the image management service to obtain a corresponding image comprises the following steps: Obtaining manufacturer and architecture information of the physical host where the computing service is located; Requesting the image management service to obtain a corresponding image according to the manufacturer and architecture information and the image name; In the image management service, the image name is not repeatable. 2.The method of claim 1, wherein, The target host is: The host with the highest score in scoring according to the computing power, allocable central processing unit and allocable memory of each physical host.

3. A cloud-multiple-core scheduling device, characterized by comprising: The method comprises the following steps: A confirming module is configured to confirm the number of created cloud hosts according to user instructions; An obtaining module is configured to obtain performance characteristic information of each physical host in a cluster; A scoring module is configured to score each physical host according to the performance characteristic information; A selecting module is configured to select a target host with the highest performance according to the scoring results; A scheduling module is configured to schedule a cloud host to be scheduled to the target host; A processing module is configured to update the performance characteristic information of the physical host; Return to the step of scoring each physical host according to the performance characteristic information until the scheduling of all cloud hosts is completed; wherein the performance characteristic information comprises the computing power, allocable central processing unit and allocable memory of the physical host; wherein, when judging whether the scheduling of all cloud hosts is completed, confirming whether the number of created cloud hosts is consistent with the number of cloud hosts that have completed the scheduling; The method further comprises the following steps: A sending module is configured to send a request for starting a cloud host by an image to the computing service of each physical host, so that, after the computing service receives the request, the computing service requests an image management service to obtain a corresponding image and starts a cloud host on the physical host according to the corresponding image; The method further comprises the following steps: A defining module is configured to define an image name, so that the image management service manages images of all manufacturers according to the image name; Further, the method further comprises the following steps: The first confirmation module is configured to confirm the image name according to a user instruction. The sending module comprises: Obtaining the vendor and architecture information of the physical host where the computing service is located; According to the vendor and architecture information and the image name, the image management service is requested to obtain the corresponding image; In the image management service, the image name is not repeatable.

4. A cloud-multiple-core scheduling device, characterized by comprising: The computer program is stored in the memory, and the processor is configured to execute the computer program to realize the steps of the cloud multi-core scheduling method according to any one of claims 1-2. The computer program is stored in the memory, and the processor is configured to execute the computer program to realize the steps of the cloud multi-core scheduling method according to any one of claims 1-2.

5. A computer readable storage medium, characterized in that, ​

Images