CN105867998A - Virtual machine cluster deployment algorithm - Google Patents

Virtual machine cluster deployment algorithm Download PDF

Info

Publication number
CN105867998A
CN105867998A CN201610173478.3A CN201610173478A CN105867998A CN 105867998 A CN105867998 A CN 105867998A CN 201610173478 A CN201610173478 A CN 201610173478A CN 105867998 A CN105867998 A CN 105867998A
Authority
CN
China
Prior art keywords
cluster
load
virtual machine
physical machine
node
Prior art date
Legal status (The legal status 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 status listed.)
Withdrawn
Application number
CN201610173478.3A
Other languages
Chinese (zh)
Inventor
莫展鹏
杨松
季统凯
Current Assignee (The listed assignees may be inaccurate. Google has not performed a legal analysis and makes no representation or warranty as to the accuracy of the list.)
STATE INFORMATION CENTER
Cloud Computing Industry Technology Innovation and Incubation Center of CAS
G Cloud Technology Co Ltd
Original Assignee
STATE INFORMATION CENTER
Cloud Computing Industry Technology Innovation and Incubation Center of CAS
G Cloud Technology Co Ltd
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 STATE INFORMATION CENTER, Cloud Computing Industry Technology Innovation and Incubation Center of CAS, G Cloud Technology Co Ltd filed Critical STATE INFORMATION CENTER
Priority to CN201610173478.3A priority Critical patent/CN105867998A/en
Publication of CN105867998A publication Critical patent/CN105867998A/en
Withdrawn legal-status Critical Current

Links

Classifications

    • 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
    • G06F9/45504—Abstract machines for programme code execution, e.g. Java virtual machine [JVM], interpreters, emulators
    • 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
    • G06F9/46—Multiprogramming arrangements
    • G06F9/50—Allocation of resources, e.g. of the central processing unit [CPU]
    • G06F9/5061—Partitioning or combining of resources
    • G06F9/5077—Logical partitioning of resources; Management or configuration of virtualized resources
    • 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
    • G06F9/46—Multiprogramming arrangements
    • G06F9/50—Allocation of resources, e.g. of the central processing unit [CPU]
    • G06F9/5083—Techniques for rebalancing the load in a distributed system

Landscapes

  • Engineering & Computer Science (AREA)
  • Software Systems (AREA)
  • Theoretical Computer Science (AREA)
  • Physics & Mathematics (AREA)
  • General Engineering & Computer Science (AREA)
  • General Physics & Mathematics (AREA)
  • Management, Administration, Business Operations System, And Electronic Commerce (AREA)

Abstract

The present invention relates to the technical field of cloud computing, and in particular to a virtual machine cluster deployment algorithm. The virtual machine cluster deployment algorithm comprises the steps of: analyzing a resource type of a cluster; filtering out a request that cannot be met; creating a priority queue; cyclically traversing all physical machines, and calculating a load coefficient L of each physical machine, and determining whether the physical machine can be selected as a deployment node; calculating a load value of each physical machine according to the resource type of the cluster, inserting the load value into the priority queue, and sorting elements of the priority queue in descending order of the load values; performing virtual machine mirror image transmission one by one in an order of the queue; and starting all virtual machines of the cluster, so as to complete cluster deployment. The present invention realizes a cluster deployment algorithm based on load value calculation; and the algorithm can be used for virtual machine cluster deployment.

Description

A kind of cluster virtual machine Deployment Algorithm
Technical field
The present invention relates to field of cloud computer technology, particularly a kind of cluster virtual machine Deployment Algorithm.
Background technology
Convenience due to cloud computing technology so that the application in cloud platform gets more and more, and the most much looks forward to Industry user also needs to the deployment of large-scale application cluster and creates, and traditional virtual cluster deployment algorithm is based only on thing The utilization rate of the most a certain resource of reason machine carries out the planning of clustered deploy(ment), there is problem below:
1, the resource characteristic of cluster is the most fully analyzed, it is impossible to make full use of the various resources of computer；
2, not accounting for the resource loss that mirror-image copies is brought, copy procedure easily affects the property of existing virtual machine Energy.
Summary of the invention
Present invention solves the technical problem that and be to provide a kind of cluster virtual machine Deployment Algorithm；Consider virtual The resource usage characteristic of cluster and the loss of mirror image transmission, solving conventional virtual machine clustered deploy(ment) algorithm can not Make full use of resource, the problem that mirror image transport overhead is excessive.
The present invention solves the technical scheme of above-mentioned technical problem:
Described algorithm comprises the following steps:
Step 1: analyze the resource type of cluster；
Step 2: compare the demand of various resources with surplus resources in cloud platform according to cluster, if cloud is put down The resource of platform is unsatisfactory for clustered deploy(ment) request, then filter out this request；
Step 3: assuming that the quantity of virtual machine is vnum, initializes the current required deployment of every physical machine of operation Virtual machine number node (i)=0, creating a capacity is the priority query of vnum priority_res_queue；
Step 4: all of physical machine of searching loop, and calculate the load factor L of each physical machine, work as load When coefficient L is less than optimum load interval limit, show that current relatively low can be used to of duty factor of physical machine is disposed Virtual machine；When load factor L is less than the optimum load interval upper limit more than optimum load interval limit, table Bright physical machine present load is in optimum state, can dispose virtual machine, but can not surpass after newly-increased virtual machine Cross the optimum load interval upper limit；When load factor L is more than the optimum load interval upper limit, show that physical machine is worked as Front overload, it should reselect node；
Step 5: calculate the load value of each physical machine according to the resource type of cluster, inserts one load value Individual priority query priority_res_queue, to virtual machine number node (i) needed for every physical machine Do and add 1 process；
Step 6: if also the load value of physical machine does not calculate, then continue executing with step 4；
Step 7: the element of priority query priority_res_queue is entered from high to low according to load value Row sequence；
Step 8: carry out virtual machine image transmission one by one according to the order of queue；
Step 9: start all virtual machines of cluster, complete clustered deploy(ment).
The resource type of described cluster refers to according to virtual machine on cluster CPU, internal memory, memory space etc. The feature of different resource usage degree and the application type distinguished；Cpu resource is used more group type Being defined as computation-intensive cluster, it is the intensive cluster of storage that memory space uses more cluster definition, Network I/O is used more cluster definition is traffic-intensive type cluster.
Described optimum load interval refers to the load factor scope being between two performance threshold, works as load When coefficient is between the two performance threshold, show that system is in optimum load state.
Described load factor refers to a quantizating index representing present physical machine load height, load factor Computing formula be:
L=α1·c(i)+α2·m(i)+α3·n(i)+node(i)*C；
Wherein L represents load system, and c (i), m (i), n (i) represent the utilization of resources of i-th physical machine respectively Rate, memory usage, broadband utilization rate；Node (i) represents that i-th physical machine needs the virtual board disposed Number；α1、α2、α3, C represent the proportionality coefficient of each utilization rate, 0 ＜ α1＜ 1,0 ＜ α2＜ 1,0 ＜ α3＜ 1,0 ＜ C ＜ 1, α1+α2+α3=1.
The load value of described physical machine refers to the adjustment to load factor of the resource type according to cluster；For Computation-intensive cluster, the intensive cluster of storage and traffic-intensive type cluster corresponding different computing formula respectively Fc (i), fm (i), fn (i) are as follows:
Fc (i)=α (1-c (i))+β (m (i)+n (i))+node (i) * C；
Fm (i)=α (1-m (i))+β (c (i)+n (i))+node (i) * M；
Fn (i)=α (1-n (i))+β (c (i)+m (i))+node (i) * N；
Wherein: 0 ＜ α ＜ 1,0 ＜ β ＜ 1, alpha+beta=1, α, β, C, M, N represent the ratio of each utilization rate Example coefficient.
The method of the present invention can produce following beneficial effect:
1, the inventive method is the algorithm of a kind of targeted equilibrium, has considered CPU, internal memory, storage The factors such as space, and according to the resource type feature calculation load value of cluster, the utilization rate of resource can be improved.
2, the inventive method is a kind of comprehensively algorithm, also allows for mirror image while considering resource utilization The expense of transmission, is avoided that the physical machine hydraulic performance decline that cluster virtual machine creating causes.
Accompanying drawing explanation
The present invention is further described below in conjunction with the accompanying drawings:
Fig. 1 is the flow chart of the present invention.
Detailed description of the invention
Below in conjunction with the accompanying drawing in the embodiment of the present invention, the technical scheme in the embodiment of the present invention is carried out clearly Chu, complete description, it is clear that described embodiment be only a part of embodiment of the present invention rather than Whole embodiments.Based on the embodiment in the present invention, those of ordinary skill in the art are not making creation The every other embodiment obtained under property work premise, broadly falls into the scope of protection of the invention.
False code based on algorithm is as follows:

Claims (7)

1. a cluster virtual machine Deployment Algorithm, it is characterised in that: described algorithm comprises the following steps:
Step 1: analyze the resource type of cluster；
Step 2: compare the demand of various resources with surplus resources in cloud platform according to cluster, if cloud is put down The resource of platform is unsatisfactory for clustered deploy(ment) request, then filter out this request；
Step 3: assuming that the quantity of virtual machine is vnum, initializes the current required deployment of every physical machine of operation Virtual machine number node (i)=0, creating a capacity is the priority query of vnum priority_res_queue；
Step 4: all of physical machine of searching loop, and calculate the load factor L of each physical machine, work as load When coefficient L is less than optimum load interval limit, show that current relatively low can be used to of duty factor of physical machine is disposed Virtual machine；When load factor L is less than the optimum load interval upper limit more than optimum load interval limit, table Bright physical machine present load is in optimum state, can dispose virtual machine, but can not surpass after newly-increased virtual machine Cross the optimum load interval upper limit；When load factor L is more than the optimum load interval upper limit, show that physical machine is worked as Front overload, it should reselect node；
Step 5: calculate the load value of each physical machine according to the resource type of cluster, inserts one load value Individual priority query priority_res_queue, to virtual machine number node (i) needed for every physical machine Do and add 1 process；
Step 6: if also the load value of physical machine does not calculate, then continue executing with step 4；
Step 7: the element of priority query priority_res_queue is entered from high to low according to load value Row sequence；
Step 8: carry out virtual machine image transmission one by one according to the order of queue；
Step 9: start all virtual machines of cluster, complete clustered deploy(ment).
Cluster virtual machine Deployment Algorithm the most according to claim 1, it is characterised in that described cluster Resource type refers to, according to virtual machine on cluster, the different resources such as CPU, internal memory, memory space are used journey The feature of degree and the application type distinguished；More group type is used to be defined as computation-intensive cpu resource Type cluster, uses memory space more cluster definition for the intensive cluster of storage, uses network I/O relatively Many cluster definitions are traffic-intensive type cluster.
Cluster virtual machine Deployment Algorithm the most according to claim 1, it is characterised in that described optimal negative Carry the load factor scope that interval refers to be between two performance threshold, when load factor is in the two Time between performance threshold, show that system is in optimum load state.
Cluster virtual machine Deployment Algorithm the most according to claim 2, it is characterised in that described optimal negative Carry the load factor scope that interval refers to be between two performance threshold, when load factor is in the two Time between performance threshold, show that system is in optimum load state.
5. according to the cluster virtual machine Deployment Algorithm described in any one of Claims 1-4, it is characterised in that Described load factor refers to a quantizating index representing present physical machine load height, the meter of load factor Calculation formula is:
L=α1·c(i)+α2·m(i)+α3·n(i)+node(i)*C；
Wherein L represents load system, and c (i), m (i), n (i) represent the utilization of resources of i-th physical machine respectively Rate, memory usage, broadband utilization rate；Node (i) represents that i-th physical machine needs the virtual board disposed Number；α1、α2、α3, C represent the proportionality coefficient of each utilization rate, 0 ＜ α1＜ 1,0 ＜ α2＜ 1,0 ＜ α3＜ 1,0 ＜ C ＜ 1, α1+α2+α3=1.
6. according to the cluster virtual machine Deployment Algorithm described in any one of Claims 1-4, it is characterised in that The load value of described physical machine refers to the adjustment to load factor of the resource type according to cluster；For calculating Intensive cluster, the intensive cluster of storage and traffic-intensive type cluster corresponding different computing formula respectively Fc (i), fm (i), fn (i) are as follows:
Fc (i)=α (1-c (i))+β (m (i)+n (i))+node (i) * C；
Fm (i)=α (1-m (i))+β (c (i)+n (i))+node (i) * M；
Fn (i)=α (1-n (i))+β (c (i)+m (i))+node (i) * N；
Wherein: 0 ＜ α ＜ 1,0 ＜ β ＜ 1, alpha+beta=1, α, β, C, M, N represent the ratio of each utilization rate Example coefficient.
Cluster virtual machine Deployment Algorithm the most according to claim 5, it is characterised in that described physical machine Load value refer to the adjustment to load factor of the resource type according to cluster；For computation-intensive cluster, The intensive cluster of storage and traffic-intensive type cluster corresponding different computing formula fc (i), fm (i), fn (i) respectively As follows:
Fc (i)=α (1-c (i))+β (m (i)+n (i))+node (i) * C；
Fm (i)=α (1-m (i))+β (c (i)+n (i))+node (i) * M；
Fn (i)=α (1-n (i))+β (c (i)+m (i))+node (i) * N；
Wherein: 0 ＜ α ＜ 1,0 ＜ β ＜ 1, alpha+beta=1, α, β, C, M, N represent the ratio of each utilization rate Example coefficient.
CN201610173478.3A 2016-03-24 2016-03-24 Virtual machine cluster deployment algorithm Withdrawn CN105867998A (en)

Priority Applications (1)

Application Number Priority Date Filing Date Title
CN201610173478.3A CN105867998A (en) 2016-03-24 2016-03-24 Virtual machine cluster deployment algorithm

Applications Claiming Priority (1)

Application Number Priority Date Filing Date Title
CN201610173478.3A CN105867998A (en) 2016-03-24 2016-03-24 Virtual machine cluster deployment algorithm

Publications (1)

Publication Number Publication Date
CN105867998A true CN105867998A (en) 2016-08-17

Family

ID=56625873

Family Applications (1)

Application Number Title Priority Date Filing Date
CN201610173478.3A Withdrawn CN105867998A (en) 2016-03-24 2016-03-24 Virtual machine cluster deployment algorithm

Country Status (1)

Country Link
CN (1) CN105867998A (en)

Cited By (6)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
CN106775457A (en) * 2016-11-28 2017-05-31 国云科技股份有限公司 A kind of method of the general acquisition disk utilization based on isomery storage virtual machine
CN107908457A (en) * 2017-11-08 2018-04-13 河海大学 A kind of containerization cloud resource distribution method based on stable matching
CN107968719A (en) * 2016-10-20 2018-04-27 上海盛霄云计算技术有限公司 The method of physical machine resource rational utilization in cloud computing
CN108287747A (en) * 2017-01-09 2018-07-17 中国移动通信集团贵州有限公司 Method and apparatus for virtual machine backup
CN113568746A (en) * 2021-07-27 2021-10-29 北京达佳互联信息技术有限公司 Load balancing method and device, electronic equipment and storage medium
CN116582546A (en) * 2023-07-12 2023-08-11 深圳市智博通电子有限公司 Network intercommunication method of virtual machine cluster based on super fusion node

Citations (3)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
CN102567077A (en) * 2011-12-15 2012-07-11 杭州电子科技大学 Virtualized resource distribution method based on game theory
CN103176849A (en) * 2013-03-12 2013-06-26 浙江大学 Virtual machine clustering deployment method based on resource classification
US9047136B2 (en) * 2010-06-11 2015-06-02 Oracle International Corporation Method and system for migrating the state of a virtual cluster

Patent Citations (3)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
US9047136B2 (en) * 2010-06-11 2015-06-02 Oracle International Corporation Method and system for migrating the state of a virtual cluster
CN102567077A (en) * 2011-12-15 2012-07-11 杭州电子科技大学 Virtualized resource distribution method based on game theory
CN103176849A (en) * 2013-03-12 2013-06-26 浙江大学 Virtual machine clustering deployment method based on resource classification

Cited By (7)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
CN107968719A (en) * 2016-10-20 2018-04-27 上海盛霄云计算技术有限公司 The method of physical machine resource rational utilization in cloud computing
CN106775457A (en) * 2016-11-28 2017-05-31 国云科技股份有限公司 A kind of method of the general acquisition disk utilization based on isomery storage virtual machine
CN108287747A (en) * 2017-01-09 2018-07-17 中国移动通信集团贵州有限公司 Method and apparatus for virtual machine backup
CN107908457A (en) * 2017-11-08 2018-04-13 河海大学 A kind of containerization cloud resource distribution method based on stable matching
CN107908457B (en) * 2017-11-08 2020-03-17 河海大学 Containerized cloud resource allocation method based on stable matching
CN113568746A (en) * 2021-07-27 2021-10-29 北京达佳互联信息技术有限公司 Load balancing method and device, electronic equipment and storage medium
CN116582546A (en) * 2023-07-12 2023-08-11 深圳市智博通电子有限公司 Network intercommunication method of virtual machine cluster based on super fusion node

Similar Documents

Publication Publication Date Title
CN105867998A (en) Virtual machine cluster deployment algorithm
CN110287245B (en) Method and system for scheduling and executing distributed ETL (extract transform load) tasks
Chen et al. How does the workload look like in production cloud? analysis and clustering of workloads on alibaba cluster trace
CN103312825B (en) A kind of distributed data storage method and apparatus
US20050177833A1 (en) Method and apparatus for reassigning objects to processing units
CN101923491A (en) Thread group address space scheduling and thread switching method under multi-core environment
Wang et al. A new algorithm for grid independent task schedule: genetic simulated annealing
CN110187960A (en) A kind of distributed resource scheduling method and device
CN106250233B (en) MapReduce performance optimization system and optimization method
US7920282B2 (en) Job preempt set generation for resource management
CN108427602B (en) Distributed computing task cooperative scheduling method and device
Zhang et al. The real-time scheduling strategy based on traffic and load balancing in storm
Schlag et al. Scalable edge partitioning
CN106033442B (en) A kind of parallel breadth first search method based on shared drive architecture
CN105488134A (en) Big data processing method and big data processing device
CN103902384A (en) Method and device for allocating physical machines for virtual machines
CN107070645A (en) Compare the method and system of the data of tables of data
CN107360031B (en) Virtual network mapping method based on optimized overhead-to-revenue ratio
CN107172193A (en) A kind of load-balancing method and its device based on cluster
CN105740077A (en) Task assigning method applicable to cloud computing
CN116244050B (en) High-response-ratio-based pneumatic characteristic example scheduling method
CN112395052A (en) Container-based cluster resource management method and system for mixed load
CN106096117B (en) Uncertain graph key side appraisal procedure based on flow and reliability
KhudaBukhsh et al. Generalized cost-based job scheduling in very large heterogeneous cluster systems
CN110084507A (en) The scientific workflow method for optimizing scheduling of perception is classified under cloud computing environment

Legal Events

Date Code Title Description
C06 Publication
PB01 Publication
C10 Entry into substantive examination
SE01 Entry into force of request for substantive examination
WW01 Invention patent application withdrawn after publication

Application publication date: 20160817

WW01 Invention patent application withdrawn after publication