VDI RESOURCE PLACEMENT DECISION DEVICE, VDI RESOURCE PLACEMENT DECISION METHOD, AND VDI RESOURCE PLACEMENT DECISION PROGRAM

The VDI resource placement determination device addresses the challenge of optimizing resource allocation in VDI systems by predicting future workload and performance needs, ensuring efficient resource utilization and enhanced user experience.

JP7593408B2Active Publication Date: 2024-12-03NIPPON TELEGRAPH & TELEPHONE CORP
View PDF 3 Cites 0 Cited by

Patent Information

Application Number
JP2022553289
Authority / Receiving Office
JP · JP
Patent Type
Patents
Current Assignee / Owner
Filing Date
2020-09-30
Publication Date
2024-12-03
Estimated Expiration
2040-09-30

AI Technical Summary

Technical Problem

Existing methods do not adequately address resource placement to satisfy performance requirements in Virtual Desktop Infrastructure (VDI) devices, impacting user experience.

Method used

A VDI resource placement determination device that collects workload and performance data, applies workload and performance models to predict future resource needs, and calculates resource allocations to meet specified performance requirements.

Benefits of technology

Enables determination of resource placements that satisfy performance requirements, improving user experience and resource utilization efficiency in VDI systems.

✦ Generated by Eureka AI based on patent content.

Smart Images

  • Figure 0007593408000001
    Figure 0007593408000001
  • Figure 0007593408000002
    Figure 0007593408000002
  • Figure 0007593408000003
    Figure 0007593408000003
Patent Text Reader

Abstract

A virtual desktop (VDI) resource allocation determination device for determining the resource allocation of a VDI device according to an embodiment is provided with: an input unit which receives performance requirements that define the performance to be met by a VDI device, and a specified time period that indicates the time period for which the performance requirements are to be met; a resource allocation collection and storage unit which stores a current resource allocation that includes the number of virtual machines owned by the host of the VDI device and virtual computational resources allocated to the virtual machines; a workload collection and storage unit which stores workloads for the VDI device; and a computational function unit which applies the specified time period and the workloads to a workload prediction model to calculate the workload for the specified time period, then applies the workload for the specified time period and the current resource allocation to a performance model to calculate a first performance of the VDI device, and determines whether or not the first performance meets the performance requirements.
Need to check novelty before this filing date? Find Prior Art

Description

[Technical field]

[0001] The present invention relates to a VDI resource placement determination device, a VDI resource placement determination method, and a VDI resource placement determination program. [Background technology]

[0002] In recent years, virtual desktop infrastructure (VDI) devices have become more widespread due to cost reduction, work style reform, etc. A VDI device is used to virtualize the desktop environment of a user terminal, and, for example, processes such as data that should be performed by a user terminal are performed on the VDI device, and processes performed by the user terminal are consolidated on the VDI device.

[0003] In order to improve the quality of experience (QoE) for users who use VDI devices, it is necessary to quantitatively satisfy various performance requirements (see, for example, Non-Patent Document 1). [Prior art documents] [Non-patent literature]

[0004] [Non-Patent Document 1] “VDC-Analyst: Design and verification of virtual desktop cloud resource allocations”, P. Calyam et al. Computer Networks 68 (2014) 110-122 Summary of the Invention [Problem to be solved by the invention]

[0005] However, Non-Patent Document 1 does not consider a method for determining resource allocation to satisfy performance requirements.

[0006] An object of the present invention is to make it possible to determine resource allocation for satisfying performance requirements. [Means for solving the problem]

[0007] In order to solve the above problems, the VDI resource allocation determination device of the present invention, which determines resource allocation for a virtual desktop (VDI) device, comprises an input unit that receives performance requirements that define the performance that the VDI device should satisfy and a designated time period indicating a time period that satisfies the performance requirements, a resource allocation collection and storage unit that saves a current resource allocation including the number of virtual machines owned by the host of the VDI device and the virtual computing resources to be allocated to the virtual machines, a workload collection and storage unit that saves the workload of the VDI device, and a calculation function unit that applies the designated time period and the workload to a workload prediction model to calculate the workload for the designated time period, applies the workload for the designated time period and the current resource allocation to a performance model to calculate a first performance of the VDI device, and determines whether the first performance satisfies the performance requirements. Effect of the Invention

[0008] According to one aspect of the present invention, a resource allocation that satisfies a performance requirement can be determined. [Brief description of the drawings]

[0009] [Figure 1] FIG. 1 is a configuration diagram showing the entire VDI resource allocation system. [Diagram 2] FIG. 2 is a block diagram showing an example of a hardware configuration of a VDI device. [Diagram 3] FIG. 3 is a block diagram illustrating an example of a hardware configuration of a VDI resource allocation determination device according to an embodiment. [Figure 4] FIG. 4 is a block diagram showing an example of a hardware configuration of a user terminal. [Diagram 5] FIG. 5 shows a functional configuration diagram of a VDI resource placement determination device. [Figure 6] FIG. 6 is a diagram illustrating an example of a workload collected by the workload collection and storage unit from a VDI device. [Figure 7] FIG. 7 is a diagram showing an example of performance collected by the performance collection and storage unit from a VDI device. [Figure 8] FIG. 8 is a diagram illustrating an example of generation of a workload prediction model by the calculation function unit. [Figure 9] FIG. 9 is a diagram illustrating an example of a workload calculation in a specified time period by the calculation function unit using a workload prediction model. [Figure 10] FIG. 10 is a diagram illustrating an example of performance model generation by the calculation function unit. [Figure 11] FIG. 11 is a diagram illustrating an example of a workload calculation in a specified time period by a calculation function unit using a performance model. [Figure 12] FIG. 12 is a flowchart showing an example of a VDI resource allocation determination operation of the VDI resource allocation determination apparatus in the embodiment. [Figure 13] FIG. 13 is a diagram showing an example of an input screen presented to the VDI operator. [Figure 14] FIG. 14 is a diagram showing an example of the current resource allocation of a VDI system. [Figure 15] FIG. 15 is a diagram illustrating an example of a workload in a specified time period. [Figure 16] FIG. 16 is a diagram showing an example of a presentation screen on which the performance calculated in step S104 in FIG. 12 is presented with a performance requirement superimposed thereon. [Figure 17] FIG. 17 is a diagram illustrating an example of a configurable resource arrangement. [Figure 18] FIG. 18 is a diagram showing an example of a presentation screen on which the performance calculated in step S108 in FIG. 12 is presented with a performance requirement superimposed thereon. DETAILED DESCRIPTION OF THE PREFERRED EMBODIMENTS

[0010] Hereinafter, an embodiment of the present invention will be described with reference to the drawings. [composition] FIG. 1 is a configuration diagram showing the entire VDI resource allocation system.

[0011] The VDI resource allocation system in Fig. 1 includes a VDI device 1, a VDI resource allocation determination device 2 according to an embodiment of the present invention, and a user terminal 3. Fig. 1 also shows a VDI operator 4 for managing and operating the VDI resource allocation determination device 2.

[0012] The VDI device 1 comprises a VDI system 110 and a virtual infrastructure management frame unit 120. The VDI system 110 comprises a host 111, a virtual machine 112 (shown as VM in FIG. 1), and an operating system 113 (shown as OS in FIG. 1). The VDI device 1 is connected to a VDI resource allocation determination device 2 wirelessly or by wire via a network using a communication protocol such as HTTP.

[0013] The host 111 is composed of one or more physical servers, and performs processing requested by the user terminal 3. Virtualization software is installed on the host 111, and virtual machines 112 and an operating system 113 used by each of the user terminals 3 are constructed. The virtual machines 112 are connected to each of the user terminals 3 wirelessly or wired via an IP network using a screen transfer protocol such as remote desktop or PCoIP. The virtual infrastructure management frame unit 120 is used to manage the VDI system 110. The virtual infrastructure management frame unit 120 manages the resource allocation of the VDI system 110, for example, the number of virtual machines 112 to be allocated to one host 111, and the number of vCPUs (virtual CPUs), memory amount, disk capacity, etc., which are virtual computing resources to be allocated to each virtual machine 112.

[0014] The VDI resource allocation determination device 2 collects the workload, performance, and current resource allocation from the VDI device 1. Furthermore, the VDI resource allocation determination device 2 holds a workload prediction model and a performance model generated from the collected workload, performance, and current resource allocation. The workload prediction model is a model that outputs the workload in a specified time period when a specified time period is input. Furthermore, the performance model is a model that outputs the performance in a specified time period when the workload and resource allocation in a specified time period output from the workload prediction model are input. Furthermore, the VDI resource allocation determination device 2 accepts input of performance requirements from the VDI operator 4. Then, the VDI resource allocation determination device 2 calculates a resource allocation that satisfies this performance requirement, and presents the calculated resource allocation to the VDI operator 4, or directly sends an instruction to the virtual infrastructure management frame unit 120 so that the resource allocation of the VDI device 1 becomes the calculated resource allocation. Furthermore, a method of calculating a resource allocation that satisfies a performance requirement will be described later.

[0015] The user terminal 3 is a terminal used by a user, and does not need to have processing power and storage in the user terminal 3 like a thin client. In other words, it is sufficient for the user terminal 3 to display information obtained from the virtual machine 112 to the user and to send input from the user to the virtual machine 112.

[0016] The VDI operator 4 may be an operator who operates the entire VDI resource allocation system, and inputs performance requirements to be satisfied for the user terminal 3 to the VDI resource allocation determination device 2. Furthermore, the VDI operator 4 can instruct the resource allocation of the VDI device 1 according to the resource allocation presented by the VDI resource allocation determination device 2.

[0017] FIG. 2 is a block diagram showing an example of the hardware configuration of the virtual infrastructure management frame unit 120 of the VDI device 1. As shown in FIG.

[0018] 2, the virtual infrastructure management frame unit 120 is configured by, for example, a server computer, and has a hardware processor 11 such as a CPU (Central Processing Unit). The hardware processor 11 may be various integrated circuits such as an ASIC (Application Specific Integrated Circuit), an FPGA (Field Programmable Gate Array), or a GPU (Graphics Processing Unit) instead of or together with the CPU. In the VDI device 1, a program memory 12, a data memory 13, a communication interface 14, and an input / output interface (referred to as input / output IF in FIG. 2) 15 are connected to the hardware processor 11 via a bus 16.

[0019] The communication interface 14 may include, for example, one or more wired or wireless communication modules. The communication interface 14 communicates between the VDI resource allocation determination device 2 and each host 111 (one or more physical servers constituting each host 111) in the VDI system 110, and enables information exchange with the VDI resource allocation determination device 2 and each host 111.

[0020] The input unit 17 and the display unit 18 are connected to the input / output interface 15. The input unit 17 and the display unit 18 are so-called tablet-type input / display devices in which an input detection sheet using an electrostatic method or a pressure method is arranged on a display screen of a display device using, for example, liquid crystal or organic EL (Electro Luminescence). The input unit 17 and the display unit 18 may be configured as independent devices. The input / output interface 15 inputs operation information inputted in the input unit 17 to the processor 11, and causes the display unit 18 to display display information generated by the processor 11.

[0021] The input unit 17 and the display unit 18 do not have to be connected to the input / output interface 15. The input unit 17 and the display unit 18 can transmit and receive information to and from the processor 11 by being provided with a communication unit for connecting to the communication interface 24 directly or via a network.

[0022] The program memory 12 is a non-transient tangible computer-readable storage medium that is a combination of a non-volatile memory that can be written to and read from at any time, such as a hard disk drive (HDD) or a solid state drive (SSD), and a non-volatile memory such as a ROM. The program memory 12 stores programs required for the processor 11 to execute various control processes according to this embodiment.

[0023] The data memory 13 is a tangible computer-readable storage medium, for example, a combination of the above-mentioned non-volatile memory and a volatile memory such as a RAM (Random Access Memory). This data memory 13 is used to store various data acquired and created in the course of various processes.

[0024] Incidentally, the one or more physical servers constituting each host 111 may also have a configuration similar to that of the server computer constituting this virtual infrastructure management frame unit 120. Furthermore, the server computer constituting this virtual infrastructure management frame unit 120 and the server computer constituting any one of the physical servers of any one of the hosts 111 may be shared by a single server computer.

[0025] FIG. 3 is a block diagram showing an example of a hardware configuration of the VDI resource allocation determination device 2. As shown in FIG.

[0026] 3, the VDI resource allocation determination device 2 is configured by, for example, a server computer or a personal computer, and has a hardware processor 21 such as a CPU. In the VDI resource allocation determination device 2, a program memory 22, a data memory 23, a communication interface 24, and an input / output interface (referred to as input / output IF in FIG. 3) 25 are connected to the hardware processor 21 via a bus 26.

[0027] The communication interface 24 may include, for example, one or more wired or wireless communication modules. The communication interface 24 communicates with the VDI device 1, enabling information exchange with the VDI device 1.

[0028] An input unit 27 and a display unit 28 are connected to the input / output interface 25. For example, a tablet-type input / display device is used as the input unit 27 and the display unit 28. The input unit 27 and the display unit 28 may be configured as independent devices. The input / output interface 25 inputs operation information inputted in the input unit 27 to the processor 21, and causes the display unit 28 to display display information generated by the processor 21.

[0029] The input unit 27 and the display unit 28 do not have to be connected to the input / output interface 25. The input unit 27 and the display unit 28 can transmit and receive information to and from the processor 21 by being provided with a communication unit for connecting to the communication interface 24 directly or via a network. In this case, the input unit 27 and the display unit 28 may function as the input unit 17 and the display unit 18 of the VDI device 1. In other words, the input unit 17 and the display unit 18 of the VDI device 1 and the input unit 27 and the display unit 28 of the VDI resource allocation determination device 2 may be combined into a single input unit and display unit.

[0030] The program memory 22 is a non-transient tangible computer-readable storage medium that is a combination of a non-volatile memory such as an HDD or SSD that can be written to and read from at any time, and a non-volatile memory such as a ROM. The program memory 22 stores programs required for the processor 21 to execute various control processes according to this embodiment.

[0031] The data memory 23 is a tangible computer-readable storage medium, for example, a combination of the above-mentioned non-volatile memory and the above-mentioned volatile memory such as RAM, etc. This data memory 23 is used to store various data acquired and created in the course of various processes.

[0032] FIG. 4 is a block diagram showing an example of a hardware configuration of the user terminal 3. As shown in FIG.

[0033] As shown in Fig. 4, the user terminal 3 has, for example, a communication interface 31 and an input / output interface (referred to as input / output IF in Fig. 4) 32 connected via a bus 35. That is, the user terminal 3 may be a thin client, and may not have a processor and various memories as shown in Figs. 2 and 3. Although not shown in Fig. 4, the user terminal 3 may have a volatile memory such as a RAM for temporarily storing data during the course of various processes.

[0034] The communication interface 31 may include, for example, one or more wired or wireless communication modules. The communication interface 31 communicates with the VDI device 1, and enables the exchange of information with the VDI device 1 regarding processing performed by the VDI device 1 or input from a user, etc.

[0035] An input unit 33 and a display unit 34 are connected to the input / output interface 32. For example, a tablet-type input / display device is used as the input unit 33 and the display unit 34. The input unit 33 and the display unit 34 may be configured as independent devices. The input / output interface 32 inputs operation information inputted in the input unit 33 to the VDI device 1 via the communication interface 31, and displays display information generated by the VDI device 1 on the display unit 34 via the communication interface 31.

[0036] FIG. 5 shows a functional configuration diagram of the VDI resource allocation determination device 2 according to this embodiment.

[0037] The VDI resource allocation determination device 2 includes a collection function group 210, a model generation function group 220, and a calculation function group 230.

[0038] The collection function group 210 includes a workload collection and storage unit 211 , a performance collection and storage unit 212 , and a resource allocation collection and storage unit 213 .

[0039] The workload collection and storage unit 211 collects and stores information about application workloads and host workloads occurring in the VDI system 110. The application workloads are workloads occurring on the virtual machines 112 and the operating system 113, such as a material viewing and editing workload, a remote conference workload, a program execution workload, an image editing workload, etc. The host workloads are workloads occurring on the host 111, such as a CPU / GPU workload, a memory workload, a network throughput workload, a virtual machine startup count and shutdown count workload, etc. of the host 111.

[0040] 6 is a diagram showing an example of a workload collected by the workload collection and storage unit 211 from the VDI device 1. The VDI device 1 collects application workloads using each application workload collection tool installed on the VDI device 1. Then, the workload collection and storage unit 211 receives and stores the collected application workloads. The VDI device 1 collects host workloads using a host workload collection tool installed on the VDI device 1. Then, the workload collection and storage unit 211 receives and stores the collected host workloads.

[0041] The application workload can directly represent the workload of the virtual machine 112, and the types of the application workload are diverse. However, because there are few dedicated tools for collecting application workloads or because of user privacy issues, the workload collection and storage unit 211 may find it difficult to collect application workloads or may only be able to collect them in a limited manner. Furthermore, the workload is essential data for generating a performance model described below. Therefore, even if it is difficult to collect application workloads or only a limited amount of application workloads can be collected, it is necessary to collect workload information that can represent the weight of VDI processing instead of application workloads. Therefore, the workload collection and storage unit 211 collects and stores host workloads that can indirectly represent VDI workloads.

[0042] When it is difficult to collect an application workload or when it is possible to collect only a limited amount of application workload, the workload collection and storage unit 211 collects a host workload that can substitute for the application workload, in order to substitute the application workload and have the performance model learn using the workload information, and sets the host workload as a workload portion of the learning data of the performance model. For example, in order to substitute a workload of viewing and editing materials, which is an application workload, the workload collection and storage unit 211 collects a CPU / GPU workload, a memory workload, and a disk workload, which are host workloads, and sets the host workload as a portion of the learning data of the performance model. Furthermore, in order to substitute a remote conference workload, which is an application workload, the workload collection and storage unit 211 collects a CPU / GPU workload and a network throughput workload, which are host workloads, and sets the host workload as a portion of the learning data of the performance model. In this way, even when it is difficult to collect an application workload or when it is possible to collect only a limited amount of application workload, the VDI resource allocation determination device 2 collects the collected host workload, thereby making it possible to provide workload information to the performance model and generate the performance model.

[0043] In addition, the workload collection and storage unit 211 collects both the application workload and the host workload even when the application workload can be collected, which allows the workload of the entire VDI system 110 to be represented more accurately and comprehensively, leading to improved accuracy of the performance model.

[0044] The performance collection and storage unit 212 collects and stores application performance and host performance. Application performance is performance that a user directly experiences, such as input response time, program startup delay, image quality, and virtual machine startup / shutdown delay time of the virtual machine 112 and the operating system 113. Host performance is performance that can be observed and evaluated by the host 111, such as CPU latency, memory overload, input response time, and network delay time of the host 111. Furthermore, the host performance can indirectly evaluate application performance.

[0045] 7 is a diagram showing an example of performance collected by the performance collection and storage unit 212 from the VDI device 1. The VDI device 1 collects application performance using each application performance collection tool installed on the VDI device 1. The performance collection and storage unit 212 then receives and stores the collected application performance. The VDI device 1 collects host performance using a host performance collection tool installed on the VDI device 1. The performance collection and storage unit 212 then receives and stores the collected host performance.

[0046] The resource allocation collection and storage unit 213 collects and stores the cluster level and resource allocation from the virtual infrastructure management frame unit 120. Here, the cluster level indicates, for example, the number of nodes in a cluster and the total amount of computing resources, and the cluster is a grouping of a predetermined number of hosts 111. The computing resources indicate, for example, the amount of vCPU, memory, and disk in a cluster. The resource allocation indicates the number of hosts in a VM 112, the number of VMs 112 allocated to the host, and the virtual computing resources allocated to each VM 112. Here, the virtual computing resources indicate the amount of vCPU, memory, and disk allocated to one VM 112. The resource allocation collection and storage unit 213 also calculates and stores the resource allocation that can be set in the VDI system 110 from the collected cluster level and resource allocation.

[0047] The model generation function unit group 220 includes a workload prediction model generation unit 221 , a workload prediction model storage unit 222 , a performance model generation unit 223 , and a performance model storage unit 224 .

[0048] The workload prediction model generating unit 221 generates a workload prediction model based on the application workload and host workload stored in the workload collecting and storing unit 211, i.e., past workload data. The workload prediction model is a model that outputs an application workload in a specified time period and a host workload in the specified time period when a future specified time period is specified. The specified time period indicates a time period in which the performance requirements that the VDI operator 4 wants the user terminal 3 to satisfy should be satisfied. Here, the workload prediction model generating unit 221 can generate the workload prediction model using an existing method such as LSTM (Long Short Time Memory). Therefore, a detailed method of generating the workload prediction model will be omitted here.

[0049] The workload prediction model storage unit 222 stores the workload prediction model generated by the workload prediction model generation unit 221 .

[0050] The performance model generating unit 223 generates a performance model based on the application workload and host workload stored in the workload collecting and storing unit 211, i.e., past workload data, the application performance and host performance stored in the performance collecting and storing unit 212, i.e., past performance, and the current resource allocation including the cluster level and resource allocation stored in the resource allocation collecting and storing unit 213. The performance model is a model that outputs the application performance in the specified time period and the host performance in the specified time period when the application workload and host workload in the specified time period and the resource allocation are input. The performance model is a model that calculates a qualitative relationship between the workload, which is the input, and the performance, which is the output. Here, the performance model generating unit 223 can use an existing method such as neural network regression to generate the performance model. Therefore, a detailed method of generating the performance model is omitted here.

[0051] The performance model storage unit 224 stores the performance model generated by the performance model generation unit 223.

[0052] The calculation function unit group 230 includes an input function unit 231 , a calculation function unit 232 , and a presentation function unit 233 .

[0053] The input function unit 231 receives requirement information including application performance requirements, host performance requirements, and designated time periods in which these requirements should be satisfied, input by the VDI operator 4. Upon receiving the requirement information, the input function unit 231 transmits the requirement information to the calculation function unit 232.

[0054] The computing function 232 calculates a workload for the specified time period using the workload prediction model. Furthermore, the computing function 232 calculates performance for the specified time period using the performance model for the current resource settings and other configurable resource settings for the workload for the specified time period.

[0055] FIG. 8 is a diagram showing an example of workload prediction model generation by the calculation function unit 232. As shown in FIG. 8, the learning data is composed of a plurality of records, and each record is composed of a feature value which is a workload in a certain time period T in the past, and a target value which is a workload in a time period T+n which is n times later than the certain time period T in the past. Here, n represents an arbitrary time. By inputting the learning data into an unlearned model such as LSTM and having it learn, a model capable of predicting a workload in a future time period from the workload in a past time period is generated. In other words, when a future time period is specified, the generated model can predict the workload in the specified future time period.

[0056] Fig. 9 is a diagram showing an example of workload calculation for a designated time period using a workload prediction model by the calculation function unit 232. As shown in Fig. 9, when a designated time period is input as requirement information, the workload prediction model outputs an application workload for the designated time period and a host workload for the designated time period.

[0057] Fig. 10 is a diagram showing an example of performance model generation by the calculation function unit 232. As shown in Fig. 10, the learning data is composed of a plurality of records, and each record is composed of feature values ​​which are workload and resource allocation in a certain time period T in the past, and a target value which is performance in the certain time period T in the past. By inputting the learning data into an unlearned model such as a neural network regression and having it learn, a model capable of calculating performance from the workload and resource allocation in a certain time period T is generated. In other words, when a workload and resource allocation are specified, the generated model can predict performance with the workload and resource allocation.

[0058] 11 is a diagram showing an example of workload calculation for a specified time period using a performance model by the calculation function unit 232. When the application workload for the specified time period and the host workload for the specified time period output from the workload prediction model and the current resource allocation stored in the resource allocation collection and storage unit 213 are input, the performance model outputs the application performance for the specified time period and the host performance for the specified time period.

[0059] Furthermore, the calculation function unit 232 judges whether or not the performance output from the performance model satisfies the performance requirements input by the VDI operator 4. If it does not satisfy the requirements, the calculation function unit 232 calculates the performance again using the configurable resource allocations stored in the resource allocation collection and storage unit 213, and finds a resource allocation that satisfies the performance requirements. Then, the calculation function unit 232 transmits information about the found resource allocation to the presentation function unit 233.

[0060] The presentation function unit 233 presents the received information on resource allocation to the VDI operator 4 .

[0061] [Operation] 12 is a flowchart showing an example of the VDI resource allocation decision operation of the VDI resource allocation decision device 2 in this embodiment. The operation of this flowchart is realized by the processor 21 of the VDI resource allocation decision device 2 reading and executing a VDI resource allocation decision program stored in the program memory 22. This flowchart starts when a request instruction to input performance requirements of the VDI device 1 is received from the VDI operator 4.

[0062] The input function unit 231 in the VDI resource allocation determination device 2 receives requirement information including performance requirements and a designated time period when the requirements are required from the VDI operator 4 via the input unit 27 (step S101). That is, when the input function unit 231 receives the request instruction from the VDI operator 4, it displays information for inputting performance requirements including application performance requirements and host performance requirements and a designated time period in the future when these requirements should be satisfied on the display unit 28, and prompts the VDI operator 4 to input. The input function unit 231 receives inputs of application performance requirements, host performance requirements, and a designated time period in the future when these requirements are required from the VDI operator 4 via the input unit 27. When the requirement information including the performance requirements and the designated time period when the requirements are required is input, the input function unit 231 transmits the requirement information to the calculation function unit 232.

[0063] FIG. 13 is a diagram showing an example of an input screen presented to the VDI operator 4. FIG. 13 shows an example in which the VDI operator 4 inputs a user input response as an application performance requirement, and an upper limit of a CPU wait time, a memory overload, and an I / O response time as a host performance requirement, and a future designated time period. The VDI resource allocation determination device 2 prompts the VDI operator 4 to input numbers in the boxed area. The VDI operator 4 inputs the performance requirement and the designated time period in which the requirement is required. Note that the requirements disclosed in FIG. 13 are merely examples, and it goes without saying that the application performance requirement and the host performance requirement are not limited to those shown in FIG. 13. Furthermore, the application performance requirement may be rephrased as a user experience requirement.

[0064] When the calculation function unit 232 receives the requirement information from the input function unit 231, it reads information about the current resource allocation of the VDI system 110 from the resource allocation collection and storage unit 213 (step S102).

[0065] Fig. 14 is a diagram showing an example of the current resource allocation of the VDI system 110. As shown in Fig. 14, the computing function unit 232 reads information about the current resource allocation, that is, the number of hosts in the VDI system 110, the number of VMs on each host, and the virtual computing resources allocated to each VM, that is, the number of vCPUs, the amount of memory, and the disk capacity.

[0066] The calculation function unit 232 calculates the workload in the specified time period using the workload prediction model stored in the workload prediction model storage unit 222 (step S103).

[0067] Fig. 15 is a diagram showing an example of a workload in a specified time period. In the example of Fig. 15, a disk I / O workload, a network throughput workload, and a workload of the number of VM startups and the number of shutdowns are shown as examples of predicted workloads in a specified time period in order to calculate a resource allocation corresponding to the performance requirements in the specified time period shown in Fig. 13. Note that the workloads shown in Fig. 15 are merely examples, and it goes without saying that the calculation function unit 232 can calculate other workloads in order to calculate the required resource allocation.

[0068] The calculation function unit 232 applies the performance model stored in the performance model storage unit 224 to the workload and current resource allocation for the specified time period calculated in step S103 to calculate the performance for the current resource allocation for the specified time period (step S104). That is, the calculation function unit 232 inputs the application workload for the specified time period and the host workload for the specified time period calculated above, and the current resource allocation stored in the resource allocation collection and storage unit 213, into the performance model. The performance model outputs the application performance and host performance for the specified time period. For example, when the workload shown in FIG. 15 and the current resource allocation shown in FIG. 14 are input into the performance model, the performance model outputs the user response time as the application performance for the specified time period, and outputs the CPU latency, memory overload, and disk I / O response time as the host performance for the specified time period.

[0069] The calculation function unit 232 judges whether the calculated performance satisfies the performance requirements input by the VDI operator 4 (step S105). If the application performance and host performance calculated in step S104 all satisfy the performance requirements, the calculation function unit 232 transmits information to the presentation function unit 233 that the current resource arrangement satisfies the performance requirements input by the VDI operator 4. The presentation function unit 233 presents information indicating that the current resource arrangement satisfies the performance requirements specified to the VDI operator 4. If even one of the application performance and host performance calculated in step S104 does not satisfy the performance requirement, the process proceeds to step S106.

[0070] FIG. 16 is a diagram showing an example of a presentation screen that presents the performance calculated in step S104 with the performance requirements superimposed thereon. The performance shown in FIG. 16 is the performance calculated from the resource allocation and workload shown in FIG. 14 and FIG. 15, and the performance requirements are the performance requirements shown in FIG. 13. The application performance presentation area 1401 is a diagram in which the user response time upper limit is superimposed on the user response time, which is the application performance. The diagram shown in the host performance presentation area 1402 is a diagram in which the upper limits input by the VDI operator 4 are superimposed on the CPU latency, memory overload, and I / O response time, which are the host performance. The user input response time in the application performance presentation area 1401 exceeds the upper limit of the user response time in a certain time period. The CPU latency, memory overload, and disk I / O response time in the host performance presentation area 1402 exceed the upper limits in a certain time period. Therefore, the example shown in FIG. 16 shows a case in which all the performance does not satisfy the performance requirements.

[0071] The calculation function unit 232 judges whether or not a settable resource allocation is stored in the resource allocation collection and storage unit 213 (step S106). If there is no settable resource allocation in the resource allocation collection and storage unit 213, the calculation function unit 232 transmits information indicating that there is no resource allocation that satisfies the performance requirements specified by the VDI operator 4 to the presentation function unit 233. Then, the presentation function unit 233 presents to the VDI operator 4 information indicating that there is no resource allocation that satisfies the specified performance requirements.

[0072] If there is a configurable resource allocation, the calculation function unit 232 reads one of the configurable resource allocations from the resource allocation collection and storage unit 213 (step S107).

[0073] Fig. 17 is a diagram showing an example of a configurable resource allocation. Fig. 17 shows changes from the resource allocation shown in Fig. 14. Fig. 17 shows an example of a resource allocation in which the number of virtual machines on host 1 is reduced from 10 to 8, the number of vCPUs on virtual machine 4 is increased from 1 to 2, and the memory amount of virtual machine 5 is increased from 2 GB to 4 GB.

[0074] The calculation function unit 232 applies a performance model to the read configurable resource allocation and the workload in the specified time period calculated in step S104, to calculate performance (step S108).

[0075] The calculation function unit 232 judges whether the calculated performance satisfies the performance requirements (step S109). If the result of the judgment is that the performance does not satisfy the performance requirements, that is, if any one of the application performance or the host performance does not satisfy the performance requirements, the process returns to step S106. If the result of the judgment is that the performance satisfies the performance requirements, that is, if both the application performance and the host performance satisfy the performance requirements, the calculation function unit 232 transmits information that the read resource arrangement satisfies the performance requirements input by the VDI operator 4 to the presentation function unit 233. The presentation function unit 233 presents information indicating that the read resource arrangement satisfies the performance requirements to the VDI operator 4. The calculation function unit 232 may also directly instruct the virtual infrastructure management frame unit 120 to use the read resource arrangement.

[0076] FIG. 18 is a diagram showing an example of a presentation screen that presents the performance calculated in step S108 with the performance requirements superimposed thereon. The performance shown in FIG. 18 is the performance calculated from the resource allocation and workload shown in FIG. 14 and FIG. 15, as in FIG. 16, and the performance is the performance requirement shown in FIG. 13. The application performance presentation area 1401 is a diagram in which the user response time upper limit is superimposed on the user response time, which is the application performance. The diagram shown in the host performance presentation area 1402 is a diagram in which the upper limits input by the VDI operator 4 are superimposed on the CPU latency, memory overload, and I / O response time, which are the host performance. The user input response time in the host performance presentation area 1402 does not exceed the upper limit of the user response time in any time period. Furthermore, the CPU latency, memory overload, and disk I / O response time in the host performance presentation area 1402 do not exceed the upper limit in any time period. Therefore, the example shown in FIG. 18 shows a case in which the performance satisfies all the performance requirements.

[0077] [Effects] It is possible to automatically calculate resource allocation that satisfies the application performance requirements and host performance requirements input by the VDI operator 4. In addition, by using this result to allocate resources of the VDI system 110, it is possible to improve the efficiency of resource utilization and to improve the satisfaction of users who use the user terminals 3.

[0078] [Other embodiments] It should be noted that the present invention is not limited to the above embodiment. For example, the VDI device 1 and the VDI resource allocation determination device 2 are shown as separate devices, but they may be one device in which all the functions are integrated.

[0079] The method described in the above embodiment can be stored as a program (software means) that can be executed by a calculator (computer) in a storage medium such as a magnetic disk (floppy (registered trademark) disk, hard disk, etc.), optical disk (CD-ROM, DVD, MO, etc.), semiconductor memory (ROM, RAM, flash memory, etc.), and can also be distributed by transmitting it via a communication medium. The program stored on the medium side also includes a setting program that configures the software means (including not only execution programs but also tables and data structures) that the calculator executes in the computer. The computer that realizes this device reads the program stored in the storage medium, and in some cases, configures the software means by the setting program, and executes the above-mentioned processing by controlling the operation of the software means. The storage medium referred to in this specification is not limited to a storage medium for distribution, but also includes a storage medium such as a magnetic disk or semiconductor memory provided inside the computer or in a device connected via a network.

[0080] In short, this invention is not limited to the above-mentioned embodiment, and various modifications can be made in the implementation stage without departing from the gist of the invention. Moreover, each embodiment may be implemented in combination as appropriate as possible, in which case the combined effect can be obtained. Furthermore, the above-mentioned embodiment includes inventions at various stages, and various inventions can be extracted by appropriate combinations of the disclosed constituent elements. [Explanation of symbols]

[0081] 1...VDI device 2...VDI resource placement decision device 3. User terminal 4…VDI operator 11…Processor 12...Program memory 13…Data memory 14...Communication interface 15... Input / Output Interface 17...Input section 18...Display section 21…Processor 22...Program memory 23…Data memory 24…Communication interface 25... Input / Output Interface 26…Bus 27...Input section 28…Display section 31...Communication interface 32... Input / Output Interface 33...Input section 34…Display section 110…VDI system 111…Host 112…Virtual Machine 113…Operating system 120...Virtual infrastructure management frame section 210... Collection Functions 211...Workload Collection and Storage Division 212…Performance Collection and Storage Department 213… Resource Allocation Collection and Storage Department 220...Model generation function group 221...Workload prediction model generation unit 222...Workload prediction model storage unit 223…Performance model generation unit 224…Performance Model Storage Section 230...Calculation function group 231... Input function unit 232...Calculation function section 233…Presentation function section 1401…Application performance display area 1402…Host performance presentation area

Claims

1. A VDI resource allocation determination device that determines resource allocation of a virtual desktop (VDI) device, an input unit for receiving performance requirements that define the performance that the VDI device should satisfy and a designated time period that indicates a time period during which the performance requirements are satisfied; a resource allocation collection and storage unit that stores a current resource allocation including the number of virtual machines owned by the host of the VDI device and virtual computing resources to be allocated to the virtual machines; a workload collection and storage unit that stores the workload of the VDI device; a calculation function unit that applies the designated time period and the workload to a workload prediction model to calculate a workload for the designated time period, applies the workload for the designated time period and the current resource allocation to a performance model to calculate a first performance of the VDI device, and determines whether the first performance satisfies the performance requirement; The workloads include application workloads that occur on the virtual machine and the operating system of the virtual machine, and host workloads that occur on the host, and when it is difficult to collect the application workloads or collection is limited, the workload collection and storage unit has a function of collecting the host workloads as an alternative and generating a model.

2. A VDI resource allocation determination method performed by a VDI resource allocation determination device that determines resource allocation of a virtual desktop (VDI) device, comprising: receiving performance requirements that define the performance that the VDI device should satisfy and a designated time period that indicates a time period during which the performance requirements are satisfied; Saving a current resource allocation including the number of virtual machines owned by the host of the VDI device and virtual computing resources to be allocated to the virtual machines; storing a workload of the VDI appliance, the workload comprising an application workload occurring on the virtual machine and the operating system of the virtual machine, and a host workload occurring on the host; applying the specified time period to a workload prediction model to calculate a workload during the specified time period, applying the workload during the specified time period and the current resource allocation to a performance model to calculate a first performance of the VDI device, and determining whether the first performance satisfies the performance requirement; If the application workload is difficult to collect or can be collected only with limited capacity, collecting the host workload instead to generate a model; A VDI resource placement determination method comprising:

3. A notification control program that causes a processor to function as each of the units of the VDI resource allocation determination device according to claim 1 .

Citation Information

Patent Citations

  • Virtual computer system and control method thereof

    JP2009104530A

  • Management server, virtual machine management method and virtual machine management program

    JP2011210134A

  • Device, method, and program for determining virtual machine combination

    JP2015207180A