Resource allocation for virtual machines

By using pseudo-random selection and resource pool management, the uneven distribution of virtual machine resources is solved, achieving uniform resource utilization and extending lifespan, thereby improving system stability and efficiency.

CN115398400BActive Publication Date: 2026-11-10MICROSOFT TECHNOLOGY LICENSING LLC
View PDF 2 Cites 0 Cited by

Patent Information

Application Number
CN202180026793.2
Authority / Receiving Office
CN · China
Patent Type
Patents(China)
Current Assignee / Owner
Priority Date
2020-04-01
Filing Date
2021-02-09
Publication Date
2026-11-10
Estimated Expiration
2041-02-09

AI Technical Summary

Technical Problem

Traditional virtual machine resource allocation methods lead to uneven use of physical resources and cannot effectively account for the degradation of individual physical resources over time, necessitating an improved resource allocation system.

Method used

A pseudo-random selection method is used to allocate the physical resources of virtual machines. By tracking the usage information and damage level of each resource, a resource pool is formed to avoid selecting resources that are damaged or nearing the end of their lifespan, thus achieving uniform utilization of resources.

Benefits of technology

This achieves even distribution of virtual machine resources, extends resource lifespan, and improves resource utilization efficiency and system stability.

✦ Generated by Eureka AI based on patent content.

Smart Images

  • Figure CN115398400B_ABST
    Figure CN115398400B_ABST
Patent Text Reader

Abstract

A system and method include receiving a request to create a virtual machine, the virtual machine being associated with a requested number of resource units for each of a plurality of resource types; determining a pool of available resource units for each of the plurality of resource types; randomly selecting a requested number of resource units from the pool of available resource units for each of the plurality of resource types; and allocating the selected resource units of each of the plurality of resource types to the virtual machine.
Need to check novelty before this filing date? Find Prior Art

Description

Background Technology

[0001] Traditional computer servers include physical resources for computing (e.g., one or more central processing units (CPUs)), data storage (e.g., one or more random access memory (RAM) modules and one or more disk drives), and other server functions. As is known in the art, these physical resources or physical devices can be allocated to one or more logical servers or virtual machines. Virtual machines typically use the physical resources allocated to them to execute operating systems and applications, without relying on other physical resources that can be presented on the server.

[0002] The server executes a virtual machine monitor or hypervisor to manage any virtual machines running within it. The hypervisor loads the operating system for each virtual machine and allocates CPU resources, RAM, and disk storage space to each virtual machine. Each virtual machine then interacts with the hypervisor to utilize its allocated resources.

[0003] Physical resources are typically allocated to virtual machines using a first-in, first-out (FIFO) protocol. Such protocols can lead to uneven use of physical resources for a given resource type. Furthermore, traditional resource allocation cannot effectively account for the degradation of individual physical resources over time. Therefore, an improved system for allocating physical resources to virtual machines is needed. Attached Figure Description

[0004] Figure 1 This is a block diagram of a computer server that includes physical resources and executes management programs, according to some embodiments.

[0005] Figure 2 This is a block diagram of a computer server that includes physical resources and executes management programs and virtual machines, according to some embodiments.

[0006] Figure 3 This is a flowchart of the process of allocating resource units of physical resource types to a virtual machine according to some embodiments.

[0007] Figure 4 This is a block diagram of a computer server that includes physical resources and executes a management program and two virtual machines, according to some embodiments.

[0008] Figure 5 This is a block diagram of a computer server that includes physical resources and executes a management program and two virtual machines, according to some embodiments.

[0009] Figure 6 This is a block diagram of a computer server that includes physical resources and executes management programs and virtual machines, according to some embodiments.

[0010] Figure 7This is a block diagram of a computer server that includes physical resources and executes a management program and two virtual machines, according to some embodiments.

[0011] Figure 8 This is a flowchart of the process of allocating resource units of physical resource types to a virtual machine according to some embodiments.

[0012] Figure 9 The illustration depicts an architecture for allocating physical resources of multiple computer servers to corresponding virtual machines, according to some embodiments.

[0013] Figure 10 The figure illustrates a cloud service architecture that allocates physical resources of multiple computer servers to corresponding virtual machines according to some embodiments.

[0014] Figure 11 This is a block diagram of a computer server for allocating physical resources of multiple computer servers to corresponding virtual machines, according to some embodiments. Detailed Implementation

[0015] The following description is provided to enable anyone in the art to make and use the described embodiments. However, various modifications will still be apparent to those skilled in the art.

[0016] Some embodiments employ a pseudo-random selection of the physical resources to be allocated to virtual machines. For example, a request may be received to create a virtual machine comprising three processor cores in a server comprising eight processor cores. In response, embodiments may pseudo-randomly select three processor cores from the eight processor cores for allocation to the virtual machine. Since each of the eight processor cores is associated with a substantially equal probability of being selected, each core may be selected substantially equal times during operation. In some cases, particularly where resource allocation is optimized for each created virtual machine and the lifecycle of each created virtual machine is fairly consistent, each of the eight processor cores will experience substantially equal amounts of degradation. As will be described below, embodiments are not limited to processor cores.

[0017] The pool of resources from which resources are pseudo-randomly selected can be determined based on degradation. For example, some embodiments may track usage information associated with each resource. If the usage information indicates that a resource is nearing or has exceeded the end of its lifecycle from which it can be assumed to operate reliably, then that resource is not included in the pool of resources from which resources are randomly selected. For example, if one of the eight processor cores in the previous example is deemed inappropriate due to its associated usage information, then three requesting processor cores are randomly selected from the remaining seven processor cores.

[0018] In the case of processor core resource types, usage information may include the number and / or intensity of computations, and in the case of disk or RAM resource types, it may include the number of reads / writes. In some embodiments, usage information includes the total uptime of virtual machines to which a particular resource is allocated. When resource allocation is optimized for each created virtual machine, total uptime can be a particularly effective proxy for resource deficiencies. Furthermore, existing scenarios for deploying virtual machines may already include mechanisms for monitoring total uptime metrics.

[0019] Figure 1 This is a block diagram of a computer server 100 according to some embodiments, the computer server 100 including physical resources 110 and an executive management program 120. The embodiments are not limited to the computer server 100 or the foregoing description thereof.

[0020] Computer server 100 may include any type of server hardware known or known to exist. Components of computer server 100 may be located within the same physical rack or within multiple independent physical units, which may be local or geographically dispersed. Such multiple independent physical units may be “striped” to function as a single logical monolithic unit. Computer server 100 may include additional hardware and software necessary to perform the functions attributed to computer server 100 herein, as well as any other functions performed by computer server 100.

[0021] Physical resources 110 include processing (i.e., computing) resources 112, RAM storage 114, and disk storage 116. Each of these resources will be referred to herein as a resource type. Embodiments are not limited to the aforementioned resource types.

[0022] The resources 110 of computer server 100 run to execute the program code of hypervisor 120. As described above, hypervisor 120 can allocate resources 110 to virtual machines running within computer server 100, instantiate virtual machines, execute guest operating systems for each virtual machine, and allow virtual machines to access the resources 110 allocated to them.

[0023] Hypervisor 120 may include a native hypervisor or a managed hypervisor. A native hypervisor runs directly on the host system hardware, rather than on the underlying operating system. Therefore, the native hypervisor directly accesses the host system's physical resources. A managed hypervisor is installed on the host system's underlying operating system. Therefore, the managed hypervisor sends requests to the underlying operating system to access the host system's physical resources.

[0024] Hypervisor 120 allocates a subset of each resource type to each virtual machine created within server 100. Each individually allocable subset of resource type will be referred to herein as a resource unit. For example, if hypervisor 120 is able to allocate a separate processor core processing resource 112 to a virtual machine executing within server 100, then each processor core is considered a unique resource unit. Each dual in-line memory module (DIMM) of RAM resource 114 can be considered a single resource unit, and / or each gigabyte of storage of RAM resource 114 can be considered a single resource unit, depending on the capabilities of hypervisor 120, computer server 100, and RAM design to directly address and allocate these unique subsets. Fixed disk 116 can similarly consist of resource units defined by unique drives, platters, sectors, or any other suitable subset.

[0025] Each resource type of resource 110 is associated with corresponding metadata 113, 115, and 117. Metadata 113, 115, and 117 may simply represent resource units of resource 110 that can be allocated to virtual machines, while other resource units of resource 110 may be specifically designated for implementing other operations of computer server 110 (e.g., execution of hypervisor 120).

[0026] For each resource unit of the corresponding resource type, metadata 113, 115, and 117 identifies the resource unit's identifier, the identifier of the virtual machine currently associated with the resource unit, and usage information. Usage information may include any appropriate indicators of the degree of degradation, such as, but not limited to, the total number of operations performed, the percentage of expected lifespan operations used, the total usage time, and the percentage of expected lifespan usage elapsed. Computer server 100 may track, update, and store metadata 113, 115, and 117, but embodiments are not limited thereto.

[0027] Figure 1 The diagram also illustrates receiving a request 130 to create a virtual machine. Request 130 can be received from any client or control system; specific examples will be discussed below. Among other details, request 130 can specify the number of resource units to be allocated to the virtual machine for each of several resource types. For example, request 130 may include instructions to create a virtual machine allocated 2Gb of RAM, one processor core, and 512Gb of fixed disk space.

[0028] According to some embodiments, hypervisor 120 receives request 130 and instantiates a virtual machine in accordance with request 130. To do this, hypervisor 120 may first determine the pool of resources for each specified resource type in the request. The example request above requests resource units of processor cores, RAM, and fixed disk resource types. Therefore, the pool of resources for each specified resource type can be determined based on metadata 113, 115, and 117 associated with the resource type.

[0029] Depend on Figure 1 The resource units represented by metadata 113, 115, and 117 have not been assigned to any virtual machine. Furthermore, the resource units represented by metadata 113, 115, and 117 are not associated with any use or corruption. Therefore, hypervisor 120 can determine that each of these resource units is available for inclusion in a pool of resource units of its corresponding resource type.

[0030] Then, resource units for the number of requests for each resource type are randomly selected from a pool of available resource units for that resource type. It should be understood that true randomness may not be achievable in practice, and the term random as used herein is intended to encompass pseudo-random selection and any other selection technique that substantially simulates true randomness.

[0031] According to some embodiments, random selection of resource units for a given resource type includes applying a random algorithm to an available pool of resource units for that resource type to generate a random set of resource units of a requested number. Any suitable system for random selection may be employed in some embodiments.

[0032] Next, the hypervisor 120 allocates the selected resource units for each resource type to the virtual machine. Figure 2 The illustration shows metadata 113, 115, and 117 following allocation in response to request 130, according to some embodiments. As shown, a processor core C1, two RAM DIMMs, R1 and R4 (each 1Gb), and two fixed disks, D0 and D4 (each 256Gb), have been allocated to virtual machine VM1. As is known in the art, hypervisor 120 can then instantiate VM1 140 based on the resource units allocated to it.

[0033] Figure 3This includes a flowchart of a process 300 for allocating resource units of physical resource types to virtual machines according to some embodiments. In some embodiments, a processing unit (e.g., one or more processors, processing cores, processor threads) or a host server executes software program code of a hypervisor to perform process 300. In other embodiments, process 300 is performed by a cloud service provider and / or a distributed resource manager. Process 300 and all other processes mentioned herein may be embodied in processor-executable program code that is read from one or more non-transitory computer-readable media, such as volatile or non-volatile random access memory, fixed disk drives, DVD-ROMs, flash drives, and magnetic tape, and then stored in a compressed, uncompiled, and / or encrypted format. Therefore, the embodiments are not limited to any specific combination of hardware and software.

[0034] Initially, in S310, a request to create a virtual machine is received. The requested virtual machine is associated with a number of resource units for each of a plurality of resource types. The request may be received from a client system intending to use the virtual machine, from a resource manager used to provide the virtual machine to one or more client systems, or from any other appropriate entity.

[0035] In S320, the pool of available resource units is determined for each requested resource type. The pool of available resource units may include all resource units that can be used by virtual machines on the host platform, excluding those currently allocated to virtual machines.

[0036] Figure 4 The illustration shows that in Figure 2 Shortly after the description, in S310, an example of receiving request 150 at server 100. Assume request 150 requests one resource unit of each resource type of resource 100. At this time, metadata 113, 115, and 117 show... Figure 2 The allocation is shown, and usage information associated with each allocated resource unit is also displayed. In S320, a resource pool is determined for each resource type, including each resource unit that is not currently allocated to a virtual machine (i.e., VM1). Specifically, the pool for the processor core resource type includes processor cores C0, C2, C3, and C4; the pool for the RAM storage resource type includes RAM DIMMs R0, R2, and R3; and the pool for the fixed disk resource type includes fixed disks D1, D2, and D3.

[0037] Then, in S330, the resource units for the number of requests for each resource type are randomly selected from the pool of available resource units for that resource type. According to this example, the processor core resource unit C4 is selected from the pool including processor cores C0, C2, and C3; the pool for RAM storage resource unit R3 is selected from the pool including RAM DIMMs R0, R2, and R3; and the fixed disk resource unit D2 is selected from the pool including fixed disks D1, D2, and D3.

[0038] Next, and as Figure 4 As reflected in the metadata, in S340, the selected resource unit of each resource type in the resource type is allocated to the requested virtual machine. Figure 4 The instantiation of the corresponding VM2 160 is further illustrated based on the resource units allocated to VM2 160.

[0039] Figure 5 The illustration shows server 100 after running virtual machines VM1 140 and VM2 160 for a period of time. Metadata 113, 115, and 117 have been updated to reflect the usage of resource units associated with each virtual machine. Metadata 113, 115, and 117 can be updated based on usage information collected by hypervisor 120, the underlying operating system of server 100, and / or by any other monitoring mechanism. As shown, a usage percentage (e.g., percentage of total recommended operating lifecycle consumed) associated with the processor cores and RAM resource units of virtual machine VM1 140 is 90% or higher, indicating that these resource units may be reaching the end of their recommended lifespan. Figure 5 The usage depicted is for illustrative purposes only and is expected to occur over time due to the instantiation, operation, and deallocation of many virtual machines, not just the two virtual machines in the example.

[0040] We will now assume that the hypervisor terminates virtual machine VM1 140, for example, in response to an instruction to terminate VM1 140. Figure 6 Server 100 is shown, where VM1 140 is no longer running. Figure 6 Metadata 113, 115, and 117 therefore indicate that there are no resource units of resource 110 associated with VM1140. Therefore, the resource units originally associated with VM1140 can be considered available for allocation to another virtual machine.

[0041] Continuing with this example, Figure 7The diagram illustrates server 100 after receiving request 170 to create another virtual machine. As described above, hypervisor 120 can respond to request 170 by determining a pool of resources for each requested resource type. A pool for a given resource type may include each resource unit of that resource type that is not currently allocated to a virtual machine and is associated with a usage percentage below a threshold. The threshold may be different for different resource types and / or in any other suitable manner.

[0042] In the following description, it will be assumed that the threshold for each resource type is 85%. Accordingly, the pools determined for the processor core resource type include processor cores C0, C2, and C3; the pools for the RAM storage resource type include RAM DIMMR0 and R2; and the pools for the fixed disk resource type include fixed disks D0, D1, D3, and D4.

[0043] Then in S330, resource units for the number of requests for each resource type are randomly selected from a pool of available resource units for that resource type. According to this example, processor core resource unit C2 is selected from a pool including processor cores C0, C2, and C3; RAM storage resource units R0 and R2 are selected from a pool including RAM DIMMs R0, R2, and R3; and fixed disk resource units D0, D3, and D4 are selected from a pool including fixed disks D0, D1, D3, and D4.

[0044] Next, and as Figure 7 The metadata in the resource type reflects that the selected resource unit of each resource type is allocated to the requested virtual machine. Figure 7 The diagram also illustrates the instantiation of the corresponding VM3 180 based on the resource units allocated to VM3 180.

[0045] Figure 8 Including about Figure 7 A flowchart of a process 800 for allocating resource units of physical resource types to a virtual machine, as described and according to some embodiments.

[0046] As described above regarding S310, a request to create a virtual machine is received in S810. The requested virtual machine is associated with a requested number of resource units for each of a plurality of resource types. In S820, a pool of available resource units is determined for each resource type of the request. The pool of available resource units may include all resource units that can be used by the virtual machine executing on the host platform, excluding those resource units currently allocated to the virtual machine.

[0047] In S830, broken resource units are identified from the pool of each resource unit determined in S820. Broken resource units can be identified based on usage information associated with each resource unit. Usage information associated with a resource unit may include the number of computations, computation intensity, device temperature, operating time, total operating time of the virtual machines to which the resource unit is allocated, and / or any other appropriate metrics. (Reference) Figure 7 For example, resource unit C1, resource unit R1, and resource unit R4 are identified as broken in S830 and are therefore removed from the pool of their respective available resource units in S840.

[0048] Then, S850 randomly selects resource units from the pool of available resource units for each resource type, corresponding to the number of requests for that resource type. In S860, the selected resource units for each resource type are allocated to the requested virtual machine. See again... Figure 7 In the S860, processor core resource unit C2, processor core resource unit C3, RAM storage resource unit R0, RAM storage resource unit R2, and fixed disk resource unit D0, fixed disk resource unit D3, and fixed disk resource unit D4 are allocated to the requested virtual machine.

[0049] Figure 9 The diagram illustrates an architecture in which server 900 executes resource manager 910 to allocate resource units to virtual machines as described above. Client system 920 sends a request 925 to server 900 to create virtual machines. Client system 920 and server 900 can each comprise any standalone or distributed computing system.

[0050] Server 900 can send request 925 to host servers 940, 950, and 960. The host servers can then execute their respective hypervisors 944, 954, and 964 as described above to allocate resource units to the virtual machine and instantiate it. A priority / load balancing algorithm can be used to determine which of the host servers 940, 950, and 960 should be used to instantiate the virtual machine.

[0051] In some embodiments, resource manager 910 maintains metadata, such as metadata 113, 115, and 117, associated with each resource unit in resources 942, 952, and 962 of host servers 940, 950, and 960. Accordingly, resource manager 910 may be executed to identify host servers in host servers 940, 950, and 960, determine a pool of available resource units for each identified host server for each resource type, and randomly select resource units from that pool as described herein. Resource manager 910 may then instruct the hypervisor associated with the identified host server to allocate the selected resource units to a virtual machine and instantiate the virtual machine.

[0052] Figure 10 The diagram illustrates an architecture for providing usage information to a resource manager, according to some embodiments. Figure 10 The architecture enables the deployment of cloud-based resources. Cloud-based resources are hosted and manageable projects that can be used by external devices via a network protocol stack. Cloud-based resources can include, but are not limited to, servers, virtual machines, storage accounts, web applications, databases, virtual networks, software, analytics, and intelligence. Customers typically only pay for the resources they use, which reduces operating costs while providing scalability to meet increasing demand.

[0053] Each of client systems 1015, 1016, and 1017 may include any system for communicating with the resource manager 1010 of server 1000. In some embodiments, each of client systems 1015, 1016, and 1017 may include a desktop computing system and may be operated by a client to access an interface provided by resource manager 1010 to create, update, and delete cloud-based resources. In some embodiments, a client establishes a subscription with the organization hosting server 1000, and cloud-based resources managed by the client are associated with that subscription.

[0054] These cloud-based resources can be accessed by customers, their customers' customers, or any appropriate entity. For example, a customer can define the compute and storage resources of a virtual machine to provide a publicly accessible online store. In another example, an instantiated virtual machine performs background functions for the customer.

[0055] According to some embodiments, in response to a received request to create a virtual machine, resource manager 1010 operates as described above. Specifically, resource manager 1010 determines a pool of resource units for each resource type from one of servers 1020, 1022, and 1024, which may exclude broken resource units. Resource units that satisfy the request are then randomly selected from the corresponding pool and allocated to the virtual machine, which is then instantiated.

[0056] Figure 10 The diagram also illustrates the usage server 1030 executing the usage monitor 1032. The hosting entity can operate the usage monitor 1032 to monitor the usage of virtual machines and / or the constituent resource units of servers 1020, 1022, and 1024. The usage monitor 1032 can track the total time each resource unit and / or virtual machine is being used, as well as the time spent in various states (e.g., turbo state, idle state). Therefore, as shown, the usage monitor can send the total usage time 1040 to the resource manager 1010. The resource manager 1010 can use the total usage time 1040 to maintain usage metadata associated with each resource unit of the servers 1020, 1022, and 1024 that can be used.

[0057] The monitor 1032 can also send total usage time 1040 to the reporting system 1050 of the hosted entity. Typically, the reporting system 1050 can operate to generate a usage report 1055. The usage report 1055 can include a list of resources associated with a given subscription. The report 1055 can also associate each instantiated virtual machine (and any other cloud-based resources subscribed to by the customer) with total usage time and a cost calculated based on the total usage time (e.g., cost per hour × total usage time in hours).

[0058] Figure 11 This is a block diagram of apparatus 1100 according to some embodiments. Apparatus 1100 may include a general-purpose computing device and may execute program code to perform any of the functions described herein. In some embodiments, apparatus 1100 may include an implementation of server 1000. According to some embodiments, apparatus 1100 may include other elements not shown.

[0059] Device 1100 includes a processing unit 1110 operatively coupled to a communication device 1120, a data storage device 1130, one or more input devices 1140, one or more output devices 1150, and a memory 1160. The communication interface 1120 facilitates communication with a network and ultimately facilitates communication with a client system as described above. The input devices 1140 may include, for example, a keyboard, keypad, mouse, or other pointing device, a microphone, knobs or switches, an infrared (IR) port, a docking station, and / or a touchscreen. The input devices 1140 may be used, for example, to input information into device 1100. The output devices 1150 may include, for example, a display and a speaker.

[0060] Data storage device 1130 may include any suitable persistent storage device, including combinations of the following: magnetic storage devices (e.g., magnetic tape, hard disk drives, and flash memory), optical storage devices, read-only memory (ROM) devices, etc., while memory 1160 may include random access memory (RAM), storage class memory (SCM), or any other fast access memory. Data storage device 1130 may be implemented using a distributed storage system.

[0061] Resource manager 1132 may include program code that can be executed by processing unit 1110 to cause device 1100 to perform any or more processes described herein. Embodiments are not limited to the execution of these processes by a single device. For example, processes categorized herein as resource manager, management program, usage monitor, and reporting server may be executed by one or more nodes of a distributed computing system.

[0062] Resource unit metadata 1134 can associate each resource unit managed by resource manager 1132 with the virtual machine (if any) to which that resource unit is allocated. Usage data 1136 can associate usage information (such as usage hours) with resource units and is used to determine the pool of resource units as described herein. Data storage device 1130 can also store data and other program code, such as device drivers, operating system files, etc., necessary to provide additional functionality and / or for the operation of device 1100.

[0063] Each functional component described herein may be implemented in computer hardware (integrated and / or discrete circuit components), in program code, and / or in one or more computing systems that execute such program code known in the art. Such a computing system may include one or more processing units that execute processor-executable program code stored in a memory system.

[0064] The diagrams described above represent a logical architecture for illustrating processes according to some embodiments, and actual implementations may include more or different components arranged in other ways. Other topologies may be used in conjunction with other embodiments. Furthermore, each component or device described herein may be implemented by any number of devices communicating via any number of other public and / or private networks. Two or more such computing devices may be geographically separated from each other and may communicate with each other via any number of networks and / or private connections of any known manner. Each component or device may include any number of hardware and / or software elements suitable for providing the functionality described herein as well as any other functionality.

[0065] The embodiments described herein are for illustrative purposes only. Those skilled in the art will recognize that other embodiments can be practiced using modifications and variations of the embodiments described above.

Claims

1. A system comprising: Memory, which stores processor-executable process steps; as well as Processing unit, configured to execute processor-executable processing steps to: Receive a request to create a virtual machine, the virtual machine being associated with a resource unit of the request number for each of a plurality of resource types; For each of the multiple resource types, determine a pool of available resource units; For each of the plurality of resource types, a subset of the pool of available resource units of that resource type is determined based on usage information, the usage information indicating the percentage of remaining expected lifetime operation usage for each resource unit of the pool of available resource units of that resource type and a specified indicator for each resource unit provided by metadata; For each of the plurality of resource types, the number of resource units requested is selected from a determined subset of the pool of available resource units of that resource type; as well as Allocate the selected resource unit of each of the plurality of resource types to the virtual machine.

2. The system of claim 1, wherein the usage information associated with the resource unit of the resource type further indicates at least one of the following: total operating time and computational intensity.

3. The system of claim 1, wherein the pool of available resource units is associated with a first computer server, and the processing unit performs the processor-executable processing steps to: Receive a second request to create a second virtual machine, the second virtual machine being associated with a resource unit of the number of second requests for each of the plurality of resource types; For each of the plurality of resource types, a second pool of available resource units associated with the second computer server is determined; For each of the plurality of resource types, a subset of the second pool of available resource units of that resource type is determined based on usage information, wherein the usage information indicates the total execution operation for each resource unit of the second pool of available resource units of that resource type; For each of the plurality of resource types, resource units of the second request number are randomly selected from the determined subset of the second pool of available resource units of the resource type; as well as Allocate the second virtual machine with the selected number of resource units of each of the plurality of resource types, representing the number of the second request. The allocation of selected resource units of each of the plurality of resource types to the virtual machine includes sending instructions to the first management program of the first computer server, and The resource unit that allocates the selected number of the second request for each of the plurality of resource types to the second virtual machine includes sending a second instruction to the second management program of the second computer server.

4. The system of claim 3, wherein determining the subset of the pool of available resource units and the subset of the second pool of available resource units for each of the plurality of resource types is based on at least one of the following: total operating time and computational intensity.

5. The system according to claim 3, wherein: The subset of the second pool of available resource units of the resource type includes a plurality of the available resource units of the resource type; as well as The subset of the second pool of available resource units of the resource type exceeds the number of resource units of the second request for the resource type used by the second virtual machine.

6. The system of claim 5, wherein the resource units of the determined subset of the second request number from the second pool of available resource units of the resource type are randomly selected.

7. The system of claim 1, wherein the plurality of resource types include processing resources, random access memory (RAM), and disk storage.

8. The system according to claim 7, wherein: The pool of available resource units for the processing resources corresponds to the processor kernel pool; The pool of available resource units for the RAM corresponds to a dual in-line memory module (DIMM) pool; as well as The pool of available resource units for the disk storage corresponds to a gigabyte storage pool.

9. The system of claim 1, wherein the percentage of remaining expected lifetime operation of each resource unit in the pool of available resource units of the resource type is compared with a threshold.

10. The system of claim 9, wherein when the percentage of remaining expected lifetime operations used is below the threshold, the corresponding resource unit of the pool of available resource units is not added to the subset of the pool of available resource units of the resource type.

11. The system of claim 1, wherein the subset of the pool of available resource units of the resource type comprises a plurality of the available resource units of the resource type.

12. The system of claim 11, wherein the subset of the pool of available resource units of the resource type exceeds the number of resource units requested for the resource type of the virtual machine.

13. The system of claim 12, wherein the number of resource units in the determined subset of the pool of available resource units of the resource type is randomly selected.

14. A computer-implemented method, comprising: Receive a request to create a virtual machine, the virtual machine being associated with a resource unit of the request number for each of a plurality of resource types; For each of the multiple resource types, determine a pool of resource units for the computer server; For each of the plurality of resource types, a subset of the pool of resource units of that resource type is determined based on usage information, wherein the usage information indicates the percentage of remaining expected lifetime operation usage for each resource unit in the pool of resource units of that resource type and a specified indicator for each resource unit provided by metadata; For each of the plurality of resource types, select the number of resource units requested from a determined subset of the pool of resource units of that resource type; as well as A virtual machine is instantiated on the computer server based on the selected resource unit of each of the plurality of resource types.

15. The method of claim 14, wherein the usage information associated with the resource unit of the resource type further indicates at least one of the following: total operating time and computational intensity.

16. The method of claim 14, further comprising: Receive a second request to create a second virtual machine, the second virtual machine being associated with a resource unit of the number of second requests for each of the plurality of resource types; For each of the plurality of resource types, a second pool of resource units of the second computer server is determined; For each of the plurality of resource types, a subset of the second pool of resource units of that resource type is determined based on usage information, wherein the usage information indicates the total execution operation for each resource unit of the second pool of resource units of that resource type; For each of the plurality of resource types, the resource units of the second request number are randomly selected from the determined subset of the second pool of resource units of the resource type; as well as Based on the selected second resource unit of each of the plurality of resource types, a second virtual machine is instantiated on the second computer server. The instantiation of the virtual machine includes sending instructions to a first hypervisor of the computer server, and The instantiation of the second virtual machine includes sending a second instruction to the second management program of the second computer server.

17. The method of claim 16, wherein determining the subset of the pool of resource units and the subset of the second pool of resource units for each of the plurality of resource types is based on at least one of: total operating time and computational intensity.

18. A system comprising: A plurality of computer servers, each of the plurality of computer servers comprising one or more resource units of each of a plurality of resource types; File Explorer, used for: Receive a request to create a virtual machine, the virtual machine being associated with a resource unit of the request number for each of a plurality of resource types; For each of the plurality of resource types, determine a pool of available resource units for a first computer server among the plurality of computer servers; For each of the plurality of resource types, a subset of the pool of available resource units of that resource type is determined based on usage information, the usage information indicating the percentage of remaining expected lifetime operation usage for each resource unit of the pool of available resource units of that resource type and a specified indicator for each resource unit provided by metadata; For each of the plurality of resource types, the number of resource units requested is randomly selected from a determined subset of the pool of available resource units of that resource type; as well as Allocate the selected resource unit of each of the plurality of resource types to the virtual machine; as well as Use a monitor for: Monitor the use of the virtual machine; Usage information is generated based on the monitored usage. as well as The usage information is sent to the resource manager.

19. The system according to claim 18, wherein the resource manager is further configured to: Receive a request to create a second virtual machine, the second virtual machine being associated with a resource unit of a second request number for each of the plurality of resource types; For each of the plurality of resource types, a second pool of available resource units of the second computer server among the plurality of computer servers is determined; For each of the plurality of resource types, a subset of the second pool of available resource units of that resource type is determined based on usage information, wherein the usage information indicates the total execution operation for each resource unit of the second pool of available resource units of that resource type; For each of the plurality of resource types, select the resource units of the second request number from the subset of the available resource units of the second pool of the resource type; as well as Allocate resource units of the selected number of the second request for each of the plurality of resource types to the second virtual machine; and The monitor is also used for: Monitor the use of the second virtual machine; Second usage information is generated based on the monitored usage of the second virtual machine; and The second usage information is sent to the resource manager.

20. The system of claim 19, wherein determining the subset of the pool of available resource units and the subset of the second pool of available resource units for each of the plurality of resource types is based on usage information associated with one or more virtual machines, the subset of the pool of available resource units and the subset of the second pool of available resource units having been previously allocated to the virtual machines.

Citation Information

Patent Citations

  • Virtual machine management method and server

    CN107526653A

  • Unified resource scheduling coordinator, method for creating virtual machine and / or container thereof and unified resource scheduling system

    CN109885377A