WO2018199415A1 - Procédé permettant de limiter les performances d'entrée et de sortie d'une machine virtuelle - Google Patents

Procédé permettant de limiter les performances d'entrée et de sortie d'une machine virtuelle Download PDF

Info

Publication number
WO2018199415A1
WO2018199415A1 PCT/KR2017/013473 KR2017013473W WO2018199415A1 WO 2018199415 A1 WO2018199415 A1 WO 2018199415A1 KR 2017013473 W KR2017013473 W KR 2017013473W WO 2018199415 A1 WO2018199415 A1 WO 2018199415A1
Authority
WO
WIPO (PCT)
Prior art keywords
input
output speed
partition
virtual machine
performance
Prior art date
Application number
PCT/KR2017/013473
Other languages
English (en)
Korean (ko)
Inventor
노재춘
김태훈
송광혁
박정민
김영우
최훈하
박성순
김경훈
Original Assignee
(주) 글루시스
Priority date (The priority date is an assumption and is not a legal conclusion. Google has not performed a legal analysis and makes no representation as to the accuracy of the date listed.)
Filing date
Publication date
Application filed by (주) 글루시스 filed Critical (주) 글루시스
Publication of WO2018199415A1 publication Critical patent/WO2018199415A1/fr

Links

Images

Classifications

    • G—PHYSICS
    • G06—COMPUTING; CALCULATING OR COUNTING
    • G06N—COMPUTING ARRANGEMENTS BASED ON SPECIFIC COMPUTATIONAL MODELS
    • G06N20/00—Machine learning
    • G—PHYSICS
    • G06—COMPUTING; CALCULATING OR COUNTING
    • G06F—ELECTRIC DIGITAL DATA PROCESSING
    • G06F9/00—Arrangements for program control, e.g. control units
    • G06F9/06—Arrangements for program control, e.g. control units using stored programs, i.e. using an internal store of processing equipment to receive or retain programs
    • G06F9/44—Arrangements for executing specific programs
    • G06F9/455—Emulation; Interpretation; Software simulation, e.g. virtualisation or emulation of application or operating system execution engines
    • G06F9/45533—Hypervisors; Virtual machine monitors
    • G06F9/45558—Hypervisor-specific management and integration aspects
    • G—PHYSICS
    • G06—COMPUTING; CALCULATING OR COUNTING
    • G06F—ELECTRIC DIGITAL DATA PROCESSING
    • G06F9/00—Arrangements for program control, e.g. control units
    • G06F9/06—Arrangements for program control, e.g. control units using stored programs, i.e. using an internal store of processing equipment to receive or retain programs
    • G06F9/44—Arrangements for executing specific programs
    • G06F9/455—Emulation; Interpretation; Software simulation, e.g. virtualisation or emulation of application or operating system execution engines
    • G06F9/45533—Hypervisors; Virtual machine monitors
    • G06F9/45558—Hypervisor-specific management and integration aspects
    • G06F2009/45579—I/O management, e.g. providing access to device drivers or storage

Definitions

  • the present invention relates to a method for limiting input / output performance for a virtual machine, and more particularly, to a method for limiting input / output performance for a virtual machine for efficiently distributing resources to the virtual machines.
  • Kernel-Based Virtual Machine a Linux-based hypervisor
  • KVM Kernel-Based Virtual Machine
  • I / O performance means the maximum input / output speed that can be used per second, and can be set by dividing into KB / sec, IOPS, read and write.
  • the I / O performance limit is an option that must be set when installing a virtual machine, but unlike the storage space setting, it is not set as the default because it is not a requirement for running a computer.
  • FIG. 1 is a diagram illustrating a problem that may occur when an input / output performance limit for a virtual machine is not set.
  • the hypervisor divides the virtual machine's storage space (image) through a single storage resource.
  • the maximum performance of the storage is 1GB / sec, and when a large number of I / O operations are performed in a specific virtual machine VM1, since the I / O performance limit for the virtual machine is not set, the virtual machine VM1 ) Can exclusively use 1GB / sec, the maximum performance of the disk.
  • I / O operations of other virtual machines VM2, VM3, VM4 cannot be guaranteed stable performance.
  • FIG. 2 is a diagram illustrating a problem that may occur when setting input / output performance limits for a virtual machine.
  • the virtual machine (VM1) is maximally using resources within its limited performance of 200 MB / sec, while the other virtual machines (VM2, VM3, VM4) have very low resource utilization (10 MB / sec, 5 MB / sec, 20 MB /). sec). In this case, the storage utilization rate is 235 MB / sec, and thus about 750 MB / sec of resources are wasted than the maximum resource of 1 GB / sec.
  • the virtual machine may divide and use the allocated virtual disk into a plurality of partitions.
  • the above-described problem occurs in the same way. That is, when I / O performance is limited for each partition, idle resources may occur depending on the partition.
  • the data block processing command is stored in one queue allocated to each virtual disk regardless of the number of partitions, and the data block processing command is processed.
  • the input / output operation of the process for the first partition is being performed and the first partition is executed. If the process for the server uses the maximum I / O speed set in the virtual machine, a bottleneck may occur in which the I / O operations for other partitions cannot be processed.
  • the present invention is to provide a method for limiting input / output performance for a virtual machine for efficiently allocating resources to the virtual machines.
  • the input / output performance of the virtual machine can be adjusted in real time for each partition of the virtual machine according to the resource utilization rate of the virtual machine without manually setting the input / output performance limit for the virtual machine.
  • a bottleneck that may occur due to limiting the I / O performance of each partition can be eliminated, and the maximum I / O speed of each partition can be efficiently set.
  • FIG. 1 is a diagram illustrating a problem that may occur when an input / output performance limit for a virtual machine is not set.
  • FIG. 2 is a diagram illustrating a problem that may occur when setting input / output performance limits for a virtual machine.
  • FIG. 3 is a flowchart illustrating a method of limiting input / output performance for a virtual machine according to an embodiment of the present invention.
  • FIG. 4 is a diagram illustrating an example of adjusting a maximum input / output speed for a partition of a virtual disk.
  • FIG. 5 is a block diagram of a hypervisor according to an embodiment of the present invention.
  • FIG. 6 is a diagram illustrating an input / output performance restriction platform according to a specific embodiment of the present invention.
  • FIG. 7 is a flowchart illustrating a method of limiting input / output performance for a virtual machine according to another embodiment of the present invention.
  • the present invention proposes a method of limiting the input / output performance of a virtual machine in real time according to the resource utilization rate of the virtual machine without the user manually setting the input / output performance limitation of the virtual machine. That is, according to the present invention, the input / output performance of the virtual machine may be adjusted in real time according to the resource utilization rate of the virtual machine.
  • the present invention focuses on the fact that a virtual disk of a virtual machine is generally divided into a plurality of partitions, thereby creating a queue for storing I / O processing instructions for each partition and limiting I / O performance for each partition by using such a queue. do.
  • a queue that stores I / O processing instructions for each partition bottlenecks caused by limiting I / O performance for each partition can be eliminated, and the maximum I / O speed for each partition can be efficiently set.
  • the method for limiting input / output performance according to the present invention may be performed in a hypervisor, which is a logical platform on which a virtual machine is driven.
  • FIG. 3 is a flowchart illustrating a method of limiting input / output performance for a virtual machine according to an embodiment of the present invention
  • FIG. 4 is a diagram illustrating an example of adjusting a maximum input / output speed for a partition of a virtual disk
  • 5 is a block diagram of a hypervisor according to an embodiment of the present invention.
  • the hypervisor stores an input / output processing command in a queue set for each partition of a virtual disk for a virtual machine (S310).
  • the maximum input / output speed for each partition is adjusted according to the resource usage rate of the partition (S320).
  • the resource may be a CPU, a memory, a network, storage, an input / output speed, and the like.
  • the hypervisor may adjust the maximum input / output speed for each partition according to the utilization rate of the input / output speed resource.
  • the resource utilization rate refers to the ratio of the input / output speed of the partition to the maximum input / output speed.
  • the hypervisor increases the maximum input / output speed for the first partition, and increases the maximum input / output speed for the second partition for which the input / output speed is less than or equal to the second threshold. Can be reduced.
  • the second partition may be a partition having the slowest input / output speed among partitions having an input / output speed equal to or less than a second threshold.
  • FIG. 4 illustrates a process in which the maximum input / output speed is adjusted in a state where the maximum input / output speed for the first three partitions is all limited to 100 MB / s.
  • the hypervisor divides the input / output speed from among the partitions a and c whose input / output speed is less than or equal to the second threshold value. Decide on one. Since the input / output speed 40 of the partition a is slower than the input / output speed 60 of the partition c, the hypervisor changes the maximum input / output speed of the partition a from 100 to 60, as shown in FIGS. 4 (b) and (c). Increase the maximum I / O speed of partition b from 100 to 140. Subsequently, as shown in FIG. 4C, when the input / output speed of the partition b is lowered below the first threshold value, the hypervisor lowers the input / output speed of the partition b to 100 and increases the input / output speed of the partition a to 100.
  • the hypervisor can adjust the maximum input / output speed when a partition having an average input / output speed of 90% or more of the maximum input / output speed exists among the partitions.
  • the hypervisor may calculate the average input / output speed by averaging the input / output speeds of the partitions stored in the log data.
  • the hypervisor may adjust the maximum input / output speed according to the maximum input / output speed adjustment request from the resource monitoring module for monitoring the resource utilization rate for the partition.
  • the resource monitoring module monitors resource utilization of the virtual machine, and may request to adjust the maximum input / output speed when a partition having an average input / output speed of more than a threshold exists.
  • the hypervisor according to the present invention adjusts the maximum input / output speed for each partition by using a queue set for each partition as described above to reduce bottlenecks between partitions.
  • the queue set for each partition is a queue which is different from the queue allocated to the virtual disk, and is a queue newly added in the present invention.
  • the hypervisor according to the present invention may adjust a time for which an input / output processing instruction waits in a queue.
  • the wait time may be determined according to the maximum input / output speed for each partition.
  • FIG. 5 is a hypervisor block diagram when two partitions are set. According to the present invention, as shown in FIG. 5, two blocks for storing input / output processing instructions for each partition are provided. Queues 510 and 520 are created.
  • Partition-specific I / O commands generated by the main process 530 are stored in queues allocated to each partition. That is, the input / output command for the first partition is stored in the first queue 510, and the input / output command for the second partition is stored in the second queue 520.
  • the I / O commands stored in the queues are processed by the I / O thread to perform a write or read system-call.
  • a write system-call is performed, a data block is written to the virtual disk 540, and when a read system-call is performed, the data block is read from the virtual disk 540.
  • the maximum I / O speed for each partition can be different, and the amount of I / O commands stored in the queue can be different. If the I / O commands are processed sequentially, bottlenecks occur between partitions. Accordingly, the hypervisor according to the present invention determines the waiting time according to the maximum input / output speed for each partition, and adjusts the waiting time of the input / output processing command in the queue.
  • the hypervisor queues the current I / O processing instruction by the difference between the maximum waiting time determined according to the maximum input / output speed and the time that data is processed according to the previous input / output processing instruction.
  • the time that the data is processed corresponds to the time when the above-described system-call is performed, and the maximum waiting time may be determined to be inversely proportional to the maximum input / output speed. Therefore, if the maximum input / output speed of the first partition is higher than the maximum input / output speed of the second partition, the maximum wait time for the first queue is shorter than the maximum wait time for the second queue.
  • the hypervisor waits for the current I / O command in the first queue for 0.7 seconds. Let's do it. That is, the previous input / output command is processed and the current input / output command is processed after 0.7 seconds.
  • the hypervisor stores the time difference between the maximum wait time and the data processing time, and reflects the time the next I / O command waits in the queue. For example, if the time taken to process the data according to the current I / O command is 0.3 seconds, the maximum wait time is 1 second, and the stored time difference is 0.2 second, the hypervisor will perform the next I / O command for 0.5 seconds minus 0.2 seconds. Queues.
  • the hypervisor according to the present invention may adjust a maximum input / output speed by adjusting a length of a queue set for each partition.
  • the hypervisor can increase the length of the queue for partitions that need to increase the maximum I / O speed, and reduce the length of the queue for partitions that need to reduce the maximum I / O speed.
  • the hypervisor according to the present invention adjusts the maximum I / O speed for each partition set in the virtual machine, the maximum I / O speed control in the situation where the input and output of data continuously may be overhead. Therefore, the hypervisor according to the present invention adjusts the maximum input / output speed adjustment timing according to the resource usage rate of each partition.
  • the hypervisor may adjust the maximum input / output speed adjustment time point by comparing a current maximum input / output speed with a size of a data processing command stored in a co-routine queue. Depending on the current maximum input / output speed and the size of the coroutine queue, it may be determined whether the data input / output processing operations are large or small.
  • the hypervisor is requested to adjust the maximum input / output speed of the resource monitoring module and adjusts the maximum input / output speed after the threshold time.
  • the maximum input / output speed is adjusted after the threshold time.
  • the hypervisor waits for a threshold time and before the work on the new data block starts after the work on the data block currently in use ends. Adjust the maximum input / output speed. For example, even if the threshold time expires while work on file a is in progress, the hypervisor adjusts the maximum I / O speed before work on file a ends and work on file b begins.
  • FIG. 6 is a diagram illustrating an I / O performance limiting platform according to a specific embodiment of the present invention, and FIG. 6 illustrates a platform operating on a Linux basis.
  • FIG. 6 two virtual machines are driven on the KVM hypervisor 630, the first virtual machine 610 uses three partitions, and the second virtual machine 620 uses two partitions.
  • the KVM hypervisor 630 includes a throttle controller 631 and 632 and a partitioning pool 633 and 634 allocated to each virtual machine.
  • the throttle controllers 631 and 632 adjust the maximum I / O speed for each partition by adjusting the time waiting for the I / O processing instruction in the partitioning pools 633 and 634.
  • the throttle controllers 731 and 732 monitor the resource utilization rate of the partition to adjust the maximum input / output speed or the maximum input / output speed according to the maximum input / output speed adjustment request of the resource monitoring module 740.
  • Partitioning pools 633 and 634 are composed of queues allocated according to the number of partitions used in each virtual machine.
  • ThreadPools for storing I / O processing instructions in a Linux system are allocated to each partition.
  • the first partitioning pool 633 is composed of three queues, and the waiting time in each queue is adjusted according to 30, 20, and 50 MB / sec, which are maximum input / output speeds for each partition.
  • the second partitioning pool 634 is composed of two queues, and the waiting time in each queue is adjusted according to the maximum input / output speed of 30 and 120 MB / sec for each partition.
  • the resource monitoring module 640 monitors the resource utilization rates of the hypervisor 630 and the virtual machines 610 and 620 in real time and stores them in the database 650.
  • the resource monitoring module 640 requests the throttle controllers 631 and 632 to adjust the maximum input / output speed when the resource usage rate of each virtual machine is greater than or equal to a threshold.
  • the information and control information collected by the resource monitoring module 640 may be used as reference data for learning to be described later.
  • the resource monitoring module 640 periodically transmits a ping signal to the throttle controllers 631 and 632 to check whether the throttle controllers 631 and 632 operate normally.
  • FIG. 7 is a flowchart illustrating a method of limiting input / output performance for a virtual machine according to another embodiment of the present invention.
  • the hypervisor according to the present invention uses a learning result of resource utilization for a partition unlike the I / O performance limiting method described in FIG. 3. Learning can be performed through various machine learning algorithms.
  • the hypervisor stores an input / output processing command in a queue set for each partition of a virtual disk for a virtual machine (S710), and uses a learning result of resource utilization for a partition to calculate a maximum input / output speed for each partition. Adjust (S720).
  • the learning result may be a learning result of the input / output speed of the partition according to the data block used.
  • the hypervisor according to the present invention learns how the input / output speed of a partition varies according to the data block read or written for each partition. Then, the maximum I / O speed for each partition is adjusted according to which data block the data block read or write data block is.
  • the hypervisor increases the maximum I / O speed of the partition. Can be set.
  • the learning result may be a learning result of an input / output speed of a partition according to a process.
  • the hypervisor learns how the input / output speed of the partition changes for each process and adjusts the maximum input / output speed for each partition according to the currently executed process.
  • the hypervisor analyzes the pattern of resource utilization for the partition in a specific situation, determines whether the current situation corresponds to the learned situation, and determines the maximum input / output speed for each partition according to the analysis result. I can regulate it.
  • the technical contents described above may be embodied in the form of program instructions that may be executed by various computer means and may be recorded in a computer readable medium.
  • the computer readable medium may include program instructions, data files, data structures, etc. alone or in combination.
  • Program instructions recorded on the media may be those specially designed and constructed for the purposes of the embodiments, or they may be of the kind well-known and available to those having skill in the computer software arts.
  • Examples of computer-readable recording media include magnetic media such as hard disks, floppy disks, and magnetic tape, optical media such as CD-ROMs, DVDs, and magnetic disks, such as floppy disks.
  • Examples of program instructions include not only machine code generated by a compiler, but also high-level language code that can be executed by a computer using an interpreter or the like.
  • the hardware device may be configured to operate as one or more software modules to perform the operations of the embodiments, and vice versa.

Landscapes

  • Engineering & Computer Science (AREA)
  • Software Systems (AREA)
  • Theoretical Computer Science (AREA)
  • Physics & Mathematics (AREA)
  • General Engineering & Computer Science (AREA)
  • General Physics & Mathematics (AREA)
  • Artificial Intelligence (AREA)
  • Computer Vision & Pattern Recognition (AREA)
  • Data Mining & Analysis (AREA)
  • Evolutionary Computation (AREA)
  • Medical Informatics (AREA)
  • Computing Systems (AREA)
  • Mathematical Physics (AREA)
  • Debugging And Monitoring (AREA)

Abstract

L'invention concerne un procédé permettant de limiter les performances d'entrée et de sortie d'une machine virtuelle, de façon à distribuer efficacement des ressources à des machines virtuelles. Le procédé permettant de limiter les performances d'entrée et de sortie d'une machine virtuelle comprend les étapes suivantes : stockage des instructions de traitement d'entrée et de sortie dans une file d'attente établie pour chaque partition d'un disque virtuel de la machine virtuelle ; et ajustement d'une vitesse d'entrée et de sortie maximale pour chaque partition en fonction d'un taux d'utilisation des ressources pour les partitions.
PCT/KR2017/013473 2017-04-26 2017-11-24 Procédé permettant de limiter les performances d'entrée et de sortie d'une machine virtuelle WO2018199415A1 (fr)

Applications Claiming Priority (2)

Application Number Priority Date Filing Date Title
KR10-2017-0053459 2017-04-26
KR1020170053459A KR101899719B1 (ko) 2017-04-26 2017-04-26 가상 머신에 대한 입출력 성능 제한 방법

Publications (1)

Publication Number Publication Date
WO2018199415A1 true WO2018199415A1 (fr) 2018-11-01

Family

ID=63721870

Family Applications (1)

Application Number Title Priority Date Filing Date
PCT/KR2017/013473 WO2018199415A1 (fr) 2017-04-26 2017-11-24 Procédé permettant de limiter les performances d'entrée et de sortie d'une machine virtuelle

Country Status (2)

Country Link
KR (1) KR101899719B1 (fr)
WO (1) WO2018199415A1 (fr)

Cited By (1)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
CN112783441A (zh) * 2021-01-08 2021-05-11 中国银联股份有限公司 一种调整虚拟机磁盘读写限速的方法、装置及计算设备

Families Citing this family (3)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
WO2021020746A1 (fr) * 2019-07-31 2021-02-04 고려대학교 산학협력단 Appareil et procédé de gestion de machine virtuelle
KR102299040B1 (ko) * 2019-07-31 2021-09-08 고려대학교 산학협력단 가상 머신 관리 장치 및 방법
KR102283364B1 (ko) * 2019-11-29 2021-07-29 세종대학교산학협력단 가상화 환경에서, 대역폭 조절 방법

Citations (5)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
JPH11194972A (ja) * 1997-12-26 1999-07-21 Nec Corp リソースブロックによる資源分割管理方法及びプログラムを記録した機械読み取り可能な記録媒体
KR20090063852A (ko) * 2007-12-14 2009-06-18 한국전자통신연구원 서버 가상화 환경을 위한 디스크 입출력 스케쥴러 및 그의스케쥴링 방법
JP2011159132A (ja) * 2010-02-01 2011-08-18 Fujitsu Ltd スイッチ装置、スイッチ制御方法、及びストレージシステム
KR20150101763A (ko) * 2014-02-27 2015-09-04 단국대학교 산학협력단 가상화 환경에서의 대역폭 할당 방법 및 장치
JP2016521417A (ja) * 2013-04-18 2016-07-21 アリババ・グループ・ホールディング・リミテッドＡｌｉｂａｂａ Ｇｒｏｕｐ Ｈｏｌｄｉｎｇ Ｌｉｍｉｔｅｄ 仮想ディスクの入出力ポートをスケジューリングするための方法およびデバイス

Family Cites Families (2)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
KR100886214B1 (ko) 2008-04-21 2009-03-10 김대봉 튜브부양체를 이용한 공중 풍력발전시스템
US9201677B2 (en) 2011-05-23 2015-12-01 Intelligent Intellectual Property Holdings 2 Llc Managing data input/output operations

Patent Citations (5)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
JPH11194972A (ja) * 1997-12-26 1999-07-21 Nec Corp リソースブロックによる資源分割管理方法及びプログラムを記録した機械読み取り可能な記録媒体
KR20090063852A (ko) * 2007-12-14 2009-06-18 한국전자통신연구원 서버 가상화 환경을 위한 디스크 입출력 스케쥴러 및 그의스케쥴링 방법
JP2011159132A (ja) * 2010-02-01 2011-08-18 Fujitsu Ltd スイッチ装置、スイッチ制御方法、及びストレージシステム
JP2016521417A (ja) * 2013-04-18 2016-07-21 アリババ・グループ・ホールディング・リミテッドＡｌｉｂａｂａ Ｇｒｏｕｐ Ｈｏｌｄｉｎｇ Ｌｉｍｉｔｅｄ 仮想ディスクの入出力ポートをスケジューリングするための方法およびデバイス
KR20150101763A (ko) * 2014-02-27 2015-09-04 단국대학교 산학협력단 가상화 환경에서의 대역폭 할당 방법 및 장치

Cited By (2)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
CN112783441A (zh) * 2021-01-08 2021-05-11 中国银联股份有限公司 一种调整虚拟机磁盘读写限速的方法、装置及计算设备
CN112783441B (zh) * 2021-01-08 2023-08-01 中国银联股份有限公司 一种调整虚拟机磁盘读写限速的方法、装置及计算设备

Also Published As

Publication number Publication date
KR101899719B1 (ko) 2018-09-17

Similar Documents

Publication Publication Date Title
WO2018199415A1 (fr) Procédé permettant de limiter les performances d'entrée et de sortie d'une machine virtuelle
KR102028096B1 (ko) 하이퍼바이저 기반의 가상 머신 격리 장치 및 방법
US10387202B2 (en) Quality of service implementation in a networked storage system with hierarchical schedulers
US11237856B2 (en) Mobility operation resource allocation
US9183016B2 (en) Adaptive task scheduling of Hadoop in a virtualized environment
US10693806B2 (en) Network bandwidth reservations for system traffic and virtual computing instances
US9588789B2 (en) Management apparatus and workload distribution management method
CN112148421B (zh) 虚拟机迁移的方法以及装置
KR101680109B1 (ko) 복수 코어 장치 및 그의 로드 조정 방법
US20130339956A1 (en) Computer system and optimal arrangement method of virtual machine in computer system
US8645592B2 (en) Balancing usage of hardware devices among clients
US7774794B2 (en) Method and system for managing bandwidth in a virtualized system
CN102073462A (zh) 虚拟存储迁移方法、系统和虚拟机监控器
US20140298047A1 (en) Power budget allocation in a cluster infrastructure
JP6072084B2 (ja) 仮想計算機システムおよび仮想計算機システムのデータ転送制御方法
JP6679146B2 (ja) 電力管理のための論理的にパーティション化された環境のイベント駆動型再最適化
WO2017082505A1 (fr) Dispositif électronique à système d'exploitation multiple et son procédé de gestion de mémoire dynamique
KR101896497B1 (ko) 성능 제어 방법 및 이를 위한 장치
JPH012145A (ja) 仮想計算機システムの資源管理方式
US20220058044A1 (en) Computer system and management method
WO2022107932A1 (fr) Système de virtualisation à base de conteneurs et procédé d'expansion de ressources de conteneurs dans un système de virtualisation
CN107436795B (zh) 一种保障虚拟机在线迁移服务质量的方法
CN115629839A (zh) 虚拟机热迁移中基于脏页配额的限速方法、系统及介质
WO2018173300A1 (fr) Procédé de commande d'e/s et système de commande d'e/s
CN116974690A (zh) 一种资源管理方法、计算设备和计算机可读存储介质

Legal Events

Date Code Title Description
121 Ep: the epo has been informed by wipo that ep was designated in this application

Ref document number: 17907929

Country of ref document: EP

Kind code of ref document: A1

NENP Non-entry into the national phase

Ref country code: DE

122 Ep: pct application non-entry in european phase

Ref document number: 17907929

Country of ref document: EP

Kind code of ref document: A1