CN109656678B - Dynamic resource management method based on virtualization - Google Patents
Dynamic resource management method based on virtualization Download PDFInfo
- Publication number
- CN109656678B CN109656678B CN201811299382.7A CN201811299382A CN109656678B CN 109656678 B CN109656678 B CN 109656678B CN 201811299382 A CN201811299382 A CN 201811299382A CN 109656678 B CN109656678 B CN 109656678B
- Authority
- CN
- China
- Prior art keywords
- virtual machine
- unit
- tmp
- physical machine
- machine
- 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.)
- Active
Links
Images
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
- G06F9/45533—Hypervisors; Virtual machine monitors
- G06F9/45558—Hypervisor-specific management and integration aspects
- G06F2009/45562—Creating, deleting, cloning virtual machine instances
-
- 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
- G06F2009/4557—Distribution of virtual machine instances; Migration and load balancing
Abstract
The invention relates to a dynamic resource management method based on virtualization, wherein the dynamic resource management method based on virtualization completes deployment of a virtual machine through a random variable according to resource requirements of the virtual machine. Under the condition that the resource demand of the virtual machine dynamically changes and needs to be predicted, the resource demand of the virtual machine is deployed to the physical machine in a random variable mode and removed from the physical machine, so that the virtual machine with the dynamically changed resource demand is deployed to the physical machine, and meanwhile, the physical machine has higher resource utilization efficiency.
Description
Technical Field
The invention relates to the field of deployment of virtual machines, in particular to a dynamic management method based on virtualization.
Background
The virtualization technology plays an important role in the cloud platform depending on the characteristics of flexibility in use, service isolation, convenience in management and the like. Due to the high creation/migration/termination cost of the virtual machine, the deployment strategy of the virtual machine has a large influence on the working efficiency of the cloud platform.
The deployment problem of virtual machines is similar to the binning problem, where several virtual machines with resource requirements are allocated to a cluster of physical machines with the largest capacity. The existing virtual machine deployment strategies are mainly divided into two types: the first type of deployment strategy treats the resource requirement of the virtual machine as a fixed value, thereby modeling the deployment problem as a boxing problem, and deploying the virtual machine through heuristic algorithms such as FIFO (first in first out), Fair and the like; the second type of deployment policy handles virtual machines whose resource requirements are specifically distributed, such as: gaussian distribution, exponential distribution, Poisson distribution, Bernoulli distribution and the like, and an approximate algorithm is designed for a certain type of virtual machine. Both of these solutions have disadvantages: the first type of scheme often has poor resource utilization efficiency because the resource demand of most virtual machines changes dynamically rather than being constant; the second type of scheme is only suitable for the situation that the resource demand is specifically distributed, and the universality is poor.
Therefore, a deployment scheme with high resource utilization efficiency and wide application range needs to be designed for the deployment problem of the virtual machine.
Disclosure of Invention
The invention aims to provide a dynamic resource management method based on virtualization.
In order to solve the technical problem, the invention provides a dynamic resource management method based on virtualization, which completes the deployment of a virtual machine through a random variable according to the resource requirement of the virtual machine.
Further, the method for completing deployment of the virtual machine by the random variable according to the resource demand of the virtual machine includes:
acquiring a random variable of the resource demand of the virtual machine;
deploying the virtual machine on the physical machine according to the random variable;
and when the deployed virtual machine stops running, removing the virtual machine from the physical machine according to the random variable.
Further, a random variable of the resource demand of the virtual machine, i.e.,
when a virtual machine with resource demand random variable X arrives, sorting the d values of X from small to large into (a)1，a2，...，ad) Define a two-dimensional array X*[1，2][1，...，d]The value of each item of the array is set as X*[1][k]＝([ak/X0]+1)*X0，X*[2][k]＝Pr(X＝ak) And k is 1 to d.
Further, the method for deploying the virtual machine on the physical machine according to the random variable includes:
step S110, find all
And will be
According to the physical machine
Is arranged in a sequence (P) from small to large1，P2，...，Pm) Step S120 is performed when i is 1;
step S120, the physical machine PiPT table of (1) is copied to array XtmpStep S130 is performed;
step S130, calculating Xtmp＝add(Xtmp，X*) Step S140 is performed;
step S140, if XtmpIs not greater than the system specified maximum collision rate δ, step S150 is performed, otherwise, step S160 is performed;
step S150, adding XtmpCopy to physical machine PiIn the PT table of (a) of (b),
is updated to
Adding item X in linked list Q, deploying the virtual machine to physical machine PiGo to step S180;
step S160, let i equal to i +1, if i > m, go to step S107, otherwise, go to step S120;
step S170, finding an idle physical machine, allocating the virtual machine to the physical machine, and updating the PT table to PT ═ add (PT, X)*)，
Is updated to
Adding an item X in the linked list Q, and performing step S180;
in step S180, the process ends.
Further, the method for removing the virtual machine from the physical machine according to the random variable after the deployed virtual machine stops running includes:
step S210, searching a resource demand random variable X corresponding to the virtual machine from a linked list Q of a physical machine in which the virtual machine is positioned, deleting X from the linked list Q, and performing step S220;
step S220, setting the first item of the PT table of the physical machine to be 1, and setting other items to be 0, and then performing step S230;
step S230, the remaining random variables in the linked list Q are sorted into (X)1，X2，...，Xr) If i is 1, go to step S240;
step S240, PT table updateNew is PT ═ add (PT, X)i) Step S250 is performed;
step S250, let i be i +1, if i > r, go to step S260, otherwise, go to step S240;
in the step S260, the process is executed,
is updated to
Recovering the resources occupied by the virtual machine, and performing step S270;
in step S270, the process ends.
Further, the add function is calculated by the method that if the number of PT terms is n, X*D is the number of items in (1), PT ═ add (PT, X)*) The calculating method comprises the following steps:
in step S310, the PT table is copied to a copy P of the PT tabletmpIn step S320, let i equal to 1;
step S320, if j is 1, go to step S330;
step S330, let Ptmp[i]＝Ptmp[i]-PT[i]*X*[2][j]Step S340 is performed;
in step S340, set variable num ═ i + X*[1][j]/X0Step S350 is performed;
step S350, if num is less than n, performing step S360, otherwise, performing step S370;
step S360, let ptmp[num]＝Ptmp[num]+PT[i]*X*[2][j]Go to step 380;
step S370, let ptmp[n]＝Ptmp[n]+PT[i]*X*[2][j]Go to step 380;
step S380, let j equal j +1, if j > d, go to step 390, otherwise, go to step 330;
step S390, if i is equal to i +1, if i is less than n, go to step 320, otherwise, go to step 410;
step S410, adding PtmpThe values of the items in the list are copied into a PT table correspondingly;
in step S420, the process ends.
The invention has the beneficial effects that the invention provides a dynamic resource management method based on virtualization, wherein the dynamic resource management method based on virtualization completes the deployment of the virtual machine through the random variable according to the resource requirement of the virtual machine. Under the condition that the resource demand of the virtual machine dynamically changes and needs to be predicted, the resource demand of the virtual machine is deployed to the physical machine in a random variable mode and removed from the physical machine, so that the virtual machine with the dynamically changed resource demand is deployed to the physical machine, and meanwhile, the physical machine has higher resource utilization efficiency.
Drawings
The invention is further illustrated with reference to the following figures and examples.
FIG. 1 is a partial flow diagram of a virtualization-based dynamic resource management method provided by the present invention.
FIG. 2 is another partial flow diagram of a virtualization-based dynamic resource management method provided by the present invention.
Detailed Description
The present invention will now be described in further detail with reference to the accompanying drawings. These drawings are simplified schematic views illustrating only the basic structure of the present invention in a schematic manner, and thus show only the constitution related to the present invention.
Examples
The embodiment of the invention provides a dynamic resource management method based on virtualization. The dynamic resource management method based on virtualization completes deployment of the virtual machine through random variables according to resource requirements of the virtual machine. Under the condition that the resource demand of the virtual machine dynamically changes and needs to be predicted, the resource demand of the virtual machine is deployed to the physical machine in a random variable mode and removed from the physical machine, so that the virtual machine with the dynamically changed resource demand is deployed to the physical machine, and meanwhile, the physical machine has higher resource utilization efficiency.
In this embodiment, the method for completing deployment of the virtual machine by using the random variable according to the resource requirement of the virtual machine includes:
step S1: acquiring a random variable of resource demand of the virtual machine, namely, when a virtual machine with a resource demand random variable of X arrives, storing the distribution of X into a two-dimensional array X*In the method, d values of X are ordered from small to large as (a)1，a2，...，ad) Define a two-dimensional array X*[1，2][1，...，d]The value of each item of the array is set as X*[1][k]＝([ak/X0]+1)*X0，X*[2][k]＝Pr(X＝ak) And k is 1 to d. Wherein, the probability of X ═ a is the probability of the virtual machine resource requirement value being a.
Step S2: deploying the virtual machine on the physical machine according to the random variable;
in this embodiment, the virtual machines are deployed on the physical machine according to the random variable, and a probability (denoted as an overload probability) that a sum of resource requirements of all the virtual machines on the same physical machine exceeds a maximum available resource amount of the physical machine does not exceed a maximum collision rate δ specified by a system, thereby achieving effective utilization of resources of the physical machine.
Step S3: and when the deployed virtual machine stops running, removing the virtual machine from the physical machine according to the random variable.
In this embodiment, the dynamic resource management method based on virtualization according to the embodiment of the present invention needs to record some values for each physical machine to optimize the calculation process, where the values include ① maintaining a PT table for each physical machine, and n is the number of entries in the PT table (the maximum available resource amount of the physical machine/X)0+2)，X0The value is set according to the demand, the value of the ith item (0 < i < n) indicates that the load of the physical machine (the sum of the resource demands of the virtual machines on the physical machine) is (i-1) ×0② maintaining a variable for each physical machine
The expected value of the physical machine load is represented by the following calculation method
③ maintain a linked list Q for each physical machine that records the resource demand random variables of the virtual machine on it.
Specifically, before virtual machine deployment is required to be performed on a physical machine, the physical machine needs to be initialized, that is, the value of the first entry of each PT table is initialized to 1, and the other entries are initialized to 0, which means that the probability that the load of the physical machine is 0 is 100%, that is, all the physical machines are in an idle state; of all physical machines
Is initialized to 0; the linked lists Q for all physical machines are initialized to an empty linked list.
Referring to fig. 1, in the embodiment, the method for deploying a virtual machine on a physical machine according to a random variable includes
Step S110, find all
And will be
According to the physical machine
Is arranged in a sequence (P) from small to large1，P2，...，Pm) Where m is the number of physical machines that are not idle. Step S120 is performed when i is 1;
step S120, the physical machine PiPT table of (1) is copied to array XtmpStep S130 is performed;
step S130, calculating Xtmp＝add(Xtmp，X*) Step S140 is performed;
step S140, if XtmpIs not greater than the system specified maximum collision rate δ, step S150 is performed, otherwise, step S160 is performed;
step S150, adding XtmpCopy to physical machine PiIn the PT table of (a) of (b),
is updated to
Adding item X in linked list Q, deploying the virtual machine to physical machine PiGo to step S180;
step S160, let i equal to i +1, if i > m, go to step S107, otherwise, go to step S120;
step S170, finding an idle physical machine, allocating the virtual machine to the physical machine, and updating the PT table to PT ═ add (PT, X)*)，
Is updated to
Adding an item X in the linked list Q, and performing step S180;
in step S180, the process ends.
Referring to fig. 2, in this embodiment, the method for removing the virtual machine from the physical machine according to the random variable after the deployed virtual machine stops running includes:
the resource demand random variable of the stopped virtual machine is marked as X.
Step S210, searching a resource demand random variable X corresponding to the virtual machine from a linked list Q of a physical machine in which the virtual machine is positioned, deleting X from the linked list Q, and performing step S220;
step S220, setting the first item of the PT table of the physical machine to be 1, and setting other items to be 0, and then performing step S230;
step S230, the remaining random variables in the linked list Q are sorted into (X)1，X2，...，Xr) If i is 1, go to step S240;
in step S240, the PT table is updated to PT ═ add (PT, X)i) Step S250 is performed;
step S250, let i be i +1, if i > r, go to step S260, otherwise, go to step S240;
in the step S260, the process is executed,
is updated to
Recovering the resources occupied by the virtual machine, and performing step S270;
in step S270, the process ends.
Wherein, the add function is calculated by the following method, that is, if the number of the PT terms is n, X*The number of terms of (d) and (P)tmpA copy of the PT table is used as a temporary array in the add function for updating the value of the PT table;
then PT is addd (PT, X)*) The calculating method comprises the following steps:
in step S310, the PT table is copied to a copy P of the PT tabletmpIn step S320, let i equal to 1;
step S320, j is a loop variable in the loop statement, which indicates that what is being processed in the current loop statement is that X takes a value of ajIn the case of time, step S330 is performed with j being 1;
step S330, let Ptmp[i]＝Ptmp[i]-PT[i]*X*[2][j]Step S340 is performed;
in step S340, set variable num ═ i + X*[1][j]/X0Num represents the new resource usage under the current value of i, j, and the step S350 is performed;
step S350, if num is less than n, performing step S360, otherwise, performing step S370;
step S360, let Ptmp[num]＝Ptmp[num]+PT[i]*X*[2][j]Go to step S380;
step S370, let Ptmp[n]＝Ptmp[n]+PT[i]*X*[2][j]Go to step S380;
step S380, let j equal j +1, if j > d, go to step 390, otherwise, go to step 330;
step S390, if i is equal to i +1, if i is less than n, go to step 320, otherwise, go to step 410;
step S410, adding PtmpThe values of the items in the list are copied into a PT table correspondingly;
in step S420, the process ends.
In summary, the present invention provides a dynamic resource management method based on virtualization, wherein the dynamic resource management method based on virtualization completes deployment of a virtual machine through a random variable according to resource requirements of the virtual machine. Under the condition that the resource demand of the virtual machine dynamically changes and needs to be predicted, the resource demand of the virtual machine is deployed to the physical machine in a random variable mode and removed from the physical machine, so that the virtual machine with the dynamically changed resource demand is deployed to the physical machine, and meanwhile, the physical machine has higher resource utilization efficiency.
In light of the foregoing description of the preferred embodiment of the present invention, many modifications and variations will be apparent to those skilled in the art without departing from the spirit and scope of the invention. The technical scope of the present invention is not limited to the content of the specification, and must be determined according to the scope of the claims.
Claims (3)
1. A virtualized dynamic resource management system is characterized in that the virtualized dynamic resource management system completes deployment of a virtual machine through a random variable according to resource requirements of the virtual machine;
the virtualized dynamic resource management system comprises:
the virtual machine resource demand acquisition module is used for acquiring a random variable of the resource demand of the virtual machine;
the virtual machine deployment module is used for deploying the virtual machine on the physical machine according to the random variable;
the virtual machine deleting module is used for removing the virtual machine from the physical machine according to the random variable after the deployed virtual machine stops running;
the virtual machine resource demand obtaining module obtains the random variable of the resource demand of the virtual machine, namely, when a virtual machine with a resource demand random variable of X arrives, d values of X are ordered from small to large as (a)1,a2,...,ad) Define a two-dimensional array X*[1,2][1,...,d]The value of each item of the array is set as X*[1][k]＝([ak/X0]+1)*X0，X*[2][k]＝Pr(X＝ak) K is 1 to d;
the virtual machine deployment module comprises:
non-idle physical machine sequencing unit to find out all
And will be
According to the physical machine
Is arranged in a sequence (P) from small to large1,P2,...,Pm) Changing i to 1, and transferring to a PT copying unit;
PT copy unit, copying physical machine PiPT table of (1) is copied to array XtmpTurning to a computing unit;
a calculation unit calculating Xtmp＝add(Xtmp,X*) Go to a collision rate determination unit;
a collision rate determination unit, if XtmpIf the value of the last item is not more than the maximum conflict rate delta specified by the system, the virtual machine deployment unit is carried out, otherwise, the deployed physical machine item number judgment unit is switched to;
a virtual machine deployment unit to deploy XtmpCopy to physical machine PiIn the PT table of (a) of (b),
is updated to
Adding item X in linked list Q, deploying the virtual machine to physical machine PiGo to the distribution completion unit;
the deployed physical machine item number judging unit makes i equal to i +1, if i is greater than m, the step S107 is carried out, and otherwise, the PT copying unit is switched to;
the virtual machine allocation unit is used for searching an idle physical machine, allocating the virtual machine to the physical machine, and updating the PT table to PT ═ add (PT, X)*)，
Is updated to
Adding an item X in the linked list Q, and transferring to an allocation completion unit;
the unit is completed and the process ends.
2. The virtualized dynamic resource management system of claim 1 wherein the virtual machine delete module comprises:
a resource demand deleting unit, which searches a resource demand random variable X corresponding to the virtual machine from a linked list Q of a physical machine in which the virtual machine is positioned, deletes the X from the linked list Q and transfers the X to an initializing unit;
the initialization unit sets a first item of a PT table of the physical machine to be 1, sets other items to be 0, and switches to the random variable sorting unit;
a random variable sorting unit for sorting the remaining random variables in the chain table Q into (X)1,X2,...,Xr) Changing i to 1, and turning to a PT table updating unit;
a PT table update unit for updating the PT table to PT ═ add (PT, X)i) Transferring to a deleted physical machine item number judging unit;
the deleted physical machine item number judging unit makes i equal to i +1, if i is greater than r, the virtual machine occupied resource recycling unit is switched to, and if not, the PT table updating unit is switched to;
the virtual machine occupies the resource recovery unit and the resource recovery unit,
is updated to
Recovering the resources occupied by the virtual machine, and transferring to a deletion completion unit;
the completion unit is deleted and the process ends.
3. The virtualized dynamic resource management system of claim 2 wherein the add function is such that if the number of PT terms is n, X*D is the number of items in (1), PT ═ add (PT, X)*) The method comprises the following steps:
step S1, copy PT table to a copy P of PT tabletmpStep S2 is performed with i equal to 1;
step S2, if j is 1, proceeds to step S3;
step S3, let Ptmp[i]＝Ptmp[i]-PT[i]*X*[2][j]Step S4 is performed;
in step S4, a variable num ═ i + X is set*[1][j]/X0Step S5 is performed;
step S5, if num is less than n, go to step S6, otherwise, go to step S7;
step S36, let Ptmp[num]＝Ptmp[num]+PT[i]*X*[2][j]Go to step S8;
step S7, let Ptmp[n]＝Ptmp[n]+PT[i]*X*[2][j]Go to step S8;
step S8, if j is j +1, if j > d, go to step S9, otherwise go to step S3;
step S9, if i is equal to i +1, if i is less than n, go to step S2, otherwise, go to step S10;
step S10, adding PtmpThe values of the items in the list are copied into a PT table correspondingly;
in step S11, the process ends.
Priority Applications (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN201811299382.7A CN109656678B (en) | 2018-11-01 | 2018-11-01 | Dynamic resource management method based on virtualization |
Applications Claiming Priority (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN201811299382.7A CN109656678B (en) | 2018-11-01 | 2018-11-01 | Dynamic resource management method based on virtualization |
Publications (2)
| Publication Number | Publication Date |
|---|---|
| CN109656678A CN109656678A (en) | 2019-04-19 |
| CN109656678B true CN109656678B (en) | 2020-05-08 |
Family
ID=66110639
Family Applications (1)
| Application Number | Title | Priority Date | Filing Date |
|---|---|---|---|
| CN201811299382.7A Active CN109656678B (en) | 2018-11-01 | 2018-11-01 | Dynamic resource management method based on virtualization |
Country Status (1)
| Country | Link |
|---|---|
| CN (1) | CN109656678B (en) |
Citations (4)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN102681899A (en) * | 2011-03-14 | 2012-09-19 | 金剑 | Virtual computing resource dynamic management system of cloud computing service platform |
| CN104601664A (en) * | 2014-12-22 | 2015-05-06 | 西安电子科技大学 | Cloud computing platform resource management and virtual machine dispatching control system |
| CN106293871A (en) * | 2016-07-22 | 2017-01-04 | 浪潮(北京)电子信息产业有限公司 | A kind of resource regulating method of cluster virtual machine |
| CN107423109A (en) * | 2017-05-24 | 2017-12-01 | 兰雨晴 | Virtual machine energy-saving scheduling method based on anonymous stochastic variable |
Family Cites Families (2)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN104123189B (en) * | 2014-06-30 | 2017-12-01 | 复旦大学 | A kind of Web multilayer application dynamic resource methods of adjustment perceived based on the application of IaaS layers |
| CN107197045A (en) * | 2017-07-19 | 2017-09-22 | 湖南大学 | The distribution method and distribution system of resources of virtual machine under cloud computing environment |
-
2018
- 2018-11-01 CN CN201811299382.7A patent/CN109656678B/en active Active
Patent Citations (4)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN102681899A (en) * | 2011-03-14 | 2012-09-19 | 金剑 | Virtual computing resource dynamic management system of cloud computing service platform |
| CN104601664A (en) * | 2014-12-22 | 2015-05-06 | 西安电子科技大学 | Cloud computing platform resource management and virtual machine dispatching control system |
| CN106293871A (en) * | 2016-07-22 | 2017-01-04 | 浪潮(北京)电子信息产业有限公司 | A kind of resource regulating method of cluster virtual machine |
| CN107423109A (en) * | 2017-05-24 | 2017-12-01 | 兰雨晴 | Virtual machine energy-saving scheduling method based on anonymous stochastic variable |
Non-Patent Citations (1)
| Title |
|---|
| 《资源调度等待开销感知的虚拟机整合》;李铭夫,毕经平,李忠诚;《软件学报》;20140715;第1388-1402页 * |
Also Published As
| Publication number | Publication date |
|---|---|
| CN109656678A (en) | 2019-04-19 |
Similar Documents
| Publication | Publication Date | Title |
|---|---|---|
| CN110618870B (en) | Working method and device for deep learning training task | |
| CN101764835B (en) | Task allocation method and device based on MapReduce programming framework | |
| WO2014210499A1 (en) | Computing connected components in large graphs | |
| CN108924187B (en) | Task processing method and device based on machine learning and terminal equipment | |
| CN108509280B (en) | Distributed computing cluster locality scheduling method based on push model | |
| CN110008013B (en) | Spark task allocation method for minimizing job completion time | |
| CN104199739A (en) | Speculation type Hadoop scheduling method based on load balancing | |
| CN106033442B (en) | A kind of parallel breadth first search method based on shared drive architecture | |
| CN104506669B (en) | The IP address distribution system and method for a kind of Based on Distributed network simulation platform | |
| CN105867998A (en) | Virtual machine cluster deployment algorithm | |
| CN111209106B (en) | Flow chart dividing method and system based on caching mechanism | |
| CN109669760B (en) | Virtualized dynamic resource management system | |
| CN111083189B (en) | System and method for processing data skew at runtime | |
| CN102082832A (en) | Data storage method and device in distributed file system | |
| CN106325995A (en) | GPU resource distribution method and system | |
| CN113805816B (en) | Disk space management method, device, equipment and storage medium | |
| CN109656678B (en) | Dynamic resource management method based on virtualization | |
| WO2017100987A1 (en) | Embedding implementation method for non-uniform bandwidth virtual data centre based on congestion avoidance | |
| CN108388471B (en) | Management method based on double-threshold constraint virtual machine migration | |
| CN112256441B (en) | Memory allocation method and device for neural network inference | |
| US9110823B2 (en) | Adaptive and prioritized replication scheduling in storage clusters | |
| CN112506813B (en) | Memory management method and system | |
| CN109040214B (en) | Service deployment method for enhancing reliability in cloud environment | |
| CN113326099A (en) | Resource management method, device, electronic equipment and storage medium | |
| JPH1139340A (en) | Data base retrieval system, multiprocessor system and data base retrieval method |
Legal Events
| Date | Code | Title | Description |
|---|---|---|---|
| PB01 | Publication | ||
| PB01 | Publication | ||
| SE01 | Entry into force of request for substantive examination | ||
| SE01 | Entry into force of request for substantive examination | ||
| GR01 | Patent grant | ||
| GR01 | Patent grant |