CN101706743B - Dispatching method of virtual machine under multi-core environment - Google Patents
Dispatching method of virtual machine under multi-core environment Download PDFInfo
- Publication number
- CN101706743B CN101706743B CN200910241371A CN200910241371A CN101706743B CN 101706743 B CN101706743 B CN 101706743B CN 200910241371 A CN200910241371 A CN 200910241371A CN 200910241371 A CN200910241371 A CN 200910241371A CN 101706743 B CN101706743 B CN 101706743B
- Authority
- CN
- China
- Prior art keywords
- subregion
- virtual machine
- scheduling
- cpu
- module
- 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.)
- Expired - Fee Related
Links
Images
Landscapes
- Stored Programmes (AREA)
Abstract
The invention provides a dispatching method of a virtual machine under a multi-core environment. The dispatching method divides zones for CUP resources according to dispatching strategy types when a system is started, monitors the CPU loading condition of all divided zones in real time when the system operates, and dynamically adjusts the size of the CPU resources in the divided zones. The dispatching method carries out dispatching in the same divided zone by using the same dispatching strategy, thus improving dispatching efficiency, achieving the purpose of resource loading balance by dynamically adjusting the size of the CPU resources of the divided zone, realizing the full utilization of resources and reducing waste of resources.
Description
Technical field
The present invention relates to a kind of dispatching method of virtual machine, relate in particular to the dispatching method of virtual machine under a kind of multi-core environment, belong to computer operating system and technical field of virtualization.
Background technology
Along with the polycaryon processor development of technology; Number of processor cores integrated on the single processor increases year by year; The computing power of server is more and more stronger, and at present general server at least also is a 4-16 nuclear, and the increase of processor core makes can converge more applications on the individual server.While is based on the unprecedented concern that obtained of server virtualization technology; The basic thought of Intel Virtualization Technology is; Through virtual machine monitor (VMM) software the hardware resource of bottom is managed; And the virtual hardware execution environment (VM) of a plurality of mutual isolation is provided, make operation a plurality of different application (OS) on same physical host like this.
Intel Virtualization Technology originates from large scale computer; As far back as the sixties in last century; A kind of operating system virtual machine technique has just been invented by IBM Corporation, and upper layer software (applications) calling physical layer interface intercepted and captured in newly-increased virtual machine middle layer on existing computing machine level; And this is called make explanations again and handle, thereby realize the new function that existing software or hardware provide that is different from.Adopt Intel Virtualization Technology can shield the dynamic and the isomerism of hardware platform, the sharing and service of support hardware resource, and independent, the isolation Calculation environment that belongs to individual is provided for each user; Can realize the integrated service device, the saving fund improves resource utilization; Can realize fault isolation, improve the security of system.
An important use of Intel Virtualization Technology is exactly a Server Consolidation, improves the utilization factor of hardware resource to greatest extent, reduces energy consumption, reduces operating cost.Server Consolidation typically refers on a physical server a plurality of application services of operation and realizes.Therefore when the processor quantity of server increased, the application service value volume and range of product on the server was also increasing, application service is corresponding virtual machine, and this moment, the problem of scheduling virtual machine became increasingly conspicuous, and showed as:
How being every kind uses the needed computational resource of distribution and makes the computational resource of system obtain best utilization.As: application need periodically obtains certain computational resource in real time, and batch application then hopes to obtain long-time computational resource;
How to use the association scheduling that provides suitable, guarantee the overall performance of application associated for being mutually related.As: classical 3-tier architecture enterprise application need be worked in coordination with between Web logical layer, data Layer.
What traditional dispatching method of virtual machine was at first considered is the resource utilization and the distributional equity property of system, does not consider the demand difference of the resources of virtual machine of different application.Existing have three kinds of more famous dispatching algorithms, BVT (BorrowedVirtual Time) dispatching algorithm, SEDF (Simple Earliest Deadline First) dispatching algorithm and a Credit dispatching algorithm.They respectively have relative merits, but all fail to consider the scheduling to the different application type.As: BVT (BorrowedVirtual Time) dispatching algorithm is the preferential dispatching algorithm of a kind of fairness; Let big VCPU of load (virtual processor) and the little VCPU of load obtain the almost CPU time of as much; The true CPU time that the big VCPU of load just may occur distributing to is not enough, and the true CPU time of distributing to the little VCPU of load residue is arranged, caused the waste of cpu resource; Farthest do not utilize cpu resource, do not consider the different demands of different application yet; Though and SEDF has solved the problem of BVT, can only carry out the SEDF scheduling respectively to each CPU, under the multinuclear situation, this algorithm can not be accomplished load balance preferably, has also ignored the different demands of different application to resource simultaneously; Last Credit dispatching algorithm can be distributed to each virtual cpu efficiently with the CPU time justice; Also can each physical cpu be distributed to each virtual cpu, realize load balance with the mode of SMP; But same Credit dispatching algorithm is not also with the consideration aspect of the type of using as scheduling.
Therefore application service quantity is gathered under the situation of increasing the demand that the single virtual machine dispatching algorithm can't satisfy the resources effective utilization and use efficient operation on the development of current server multi-core technology and server.A kind of dispatching algorithm maybe to one type application effectively but will produce the unreasonable of resources allocation when polytype being applied in when dispatching together; Relatively be fit to use in real time like: BVT algorithm, but will produce the irrational situation of resources allocation together the time when application and batch application in real time.
Summary of the invention
The present invention proposes the dispatching method of virtual machine under a kind of multi-core environment; This method is in existing virtual machine monitor (VMM), to have added four modules, promptly dispatches initialization module, CPU monitoring module, scheduling decision module and CPU dynamic partition module.When system power-up started, control terminal and guest virtual machine communicated through control interface.At first all CPU are enumerated out by the control interface triggering scheduling initialization module of control terminal notice VMM; Through CPU dynamic partition module CPU is carried out subregion according to the scheduling strategy type of systemic presupposition support then, and each subregion is specified corresponding scheduling strategy by the scheduling decision module; During the initialization subregion, CPU dynamic partition module is that the division of every kind of preset scheduling strategy type gets minimum cpu resource (minimum value of resource is generally set by the keeper, is defaulted as 1), and remaining cpu resource will be placed in the Free Partition.When new virtual machine was created, CPU dynamic partition module joined corresponding subregion with it and dispatches according to the scheduling parameter of virtual machine.When system moves, the loading condition of each subregion of CPU monitoring module monitoring, excessive when certain subregion cpu load, or when too small, all can trigger CPU dynamic partition module, the size of subregion is adjusted.
Dispatching method of virtual machine under a kind of multi-core environment provided by the invention comprises following execution in step:
Step 1: power up the startup initialization; The scheduling initialization module loads during system start-up, carries out initial work, and this step comprises following two operations:
Step 1.1 is enumerated CPU and is read preset scheduling strategy type; When system power-up was accomplished startup, the scheduling initialization module loaded, and all CPU in the system are enumerated out, simultaneously preset scheduling strategy type in the reading system;
Step 1.2 is carried out primary partition for system; The scheduling initialization module is that every kind of scheduling strategy type is got the subregion of resource minimum value by CPU dynamic partition module after obtaining the scheduling strategy type, sets up Free Partition simultaneously in order to place remaining cpu resource, accomplishes system's primary partition;
Step 2: create virtual machine; When the system creation virtual machine, the scheduling decision module can read the scheduling parameter of virtual machine, and virtual machine is joined corresponding subregion, and this step comprises following two operations:
Step 2.1 scheduling decision module reads scheduling parameter; When new establishment virtual machine, system can formulate scheduling parameter according to the type that virtual machine is used in configuration file, and the configuration file that the scheduling decision module can initiatively read virtual machine obtains scheduling parameter;
Step 2.2 joins corresponding subregion with virtual machine; After scheduling parameter obtains; The scheduling decision module is searched corresponding subregion through scheduling parameter; When corresponding subregion existed, CPU dynamic partition module directly joined corresponding subregion scheduling with this virtual machine, when the corresponding subregion of scheduling parameter does not exist; CPU dynamic partition module is told a new subregion as this scheduling strategy categories subarea from Free Partition, and this virtual machine is added scheduling;
Step 3: scheduling virtual machine; When system moved, this step specifically comprised following operation:
Step 3.1 virtual machine is being transferred in the subregion separately; The virtual machine of operation can only be dispatched in the CPU of appointment subregion, and the scheduling decision module is dispatched the virtual machine in each subregion according to the dispatching algorithm of each subregion;
Step 3.2 each subregion loading condition of monitoring is also dynamically adjusted partition size; The cpu resource that the CPU monitoring module is being kept watch on each subregion utilizes situation; When the cpu load of certain subregion is excessive; The CPU monitoring module triggers CPU dynamic partition module, and CPU dynamic partition module is told cpu resource to the excessive subregion of load from Free Partition, when Free Partition is sky; From the less subregion of load, tell cpu resource in the excessive subregion of load; When the cpu load of subregion was too small, CPU dynamic partition module can the partitioned resources that load is little be recovered to Free Partition, dwindles the little partitioned resources of load.
Describedly a kind ofly it is characterized in that based on the dispatching method of virtual machine under the multi-core environment scheduling strategy type described in the step 1 comprises BVT dispatching algorithm and SEDF dispatching algorithm.
Described a kind ofly it is characterized in that based on the dispatching method of virtual machine under the multi-core environment, the resource minimum value described in the step 1, this value is set by the keeper, is defaulted as 1 CPU.
Described a kind ofly it is characterized in that based on the dispatching method of virtual machine under the multi-core environment, the scheduling parameter described in the step 2 is meant that virtual machine uses the scheduling strategy that is fit to.
Describedly a kind ofly it is characterized in that the size of the adjustment subregion described in the step 3.2 based on the dispatching method of virtual machine under the multi-core environment; The size of the each adjustment of acquiescence subregion is a cpu resource; In particular cases, when the load of a certain subregion is 0%, show that then this subregion does not have scheduling virtual machine; Then CPU dynamic partition module is deleted this subregion, and resource is recovered in the Free Partition.
Comparing prior art the present invention has the following advantages and good effect:
(1) improves the efficient of dispatching
Traditional scheduling virtual machine mostly adopts overall uniform dispatching method, and each CPU has adopted identical scheduling strategy and parameter setting.In practical application; Different application is different to the requirement of resource; The different resources demand needs corresponding scheduling strategy just can reach the purpose that resource makes full use of; Periodically distribute timeslice like real-time system requirements, batch processing system then requires continuous distribution timeslice, and these two kinds of application just need different scheduling strategies.The present invention carries out subregion with cpu resource according to the scheduling strategy type, uses the virtual machine of identical scheduling strategy in same subregion, to dispatch, and has improved the efficient of scheduling like this.
(2) better utilization resource
The CPU subregion that the present invention proposes is dynamic, and the CPU monitoring module is the cpu load situation of each subregion of monitoring in real time, for the big subregion of load, can trigger CPU dynamic partition module the CPU in the Free Partition is distributed to the big subregion of load; For the little subregion of load, the resource that can trigger the CPU dynamic partition module subregion that load is little reclaims a part in Free Partition.Reach the purpose of balancing resource load through the size of dynamic adjustment subregion, realized making full use of of resource, reduced the waste of resource.
Description of drawings
Fig. 1 is a dispatching method of virtual machine model of module structural representation under the multi-core environment of the present invention;
Fig. 2 is the dispatching method of virtual machine total system execution in step process flow diagram under a kind of multi-core environment of proposing of the present invention;
Fig. 3 is that the dispatching method of virtual machine under a kind of multi-core environment of proposing of the present invention powers up and starts initialization execution in step process flow diagram.
Embodiment
Below in conjunction with accompanying drawing the present invention is further specified.
As shown in Figure 1, the dispatching method of virtual machine under a kind of multi-core environment that the present invention proposes adds four new modules in existing virtual machine monitor VMM1: CPU dynamic partition module 2, scheduling initialization module 3, scheduling decision module 4, CPU monitoring module 5.
CPU dynamic partition module 2 is responsible for cpu resource is carried out subregion according to the scheduling strategy type, and is dynamically adjusted the size of each subregion, and in each subregion, adds newly-increased virtual machine.The CPU information and the preset scheduling strategy type information that at first pass over according to scheduling initialization module 3; System CPU is carried out primary partition, and the number of subregion is that the number of scheduling strategy type adds 1, because need a Free Partition; When initial is that busy division gets minimum cpu resource; The minimum value of resource generally has the keeper to set, and is defaulted as 1, and remaining CPU is placed on Free Partition.When system in when operation, CPU dynamic partition module 2 is adjusted the CPU partitioned resources according to the information that scheduling decision module 4 and CPU monitoring module 5 pass over dynamically, makes each subregion load reach load balancing.
Scheduling initialization module 3; Major function is when system power-up starts, to be loaded by control interface 6; Be responsible for enumerating all CPU; And then pass through the preset scheduling strategy type of supporting of control interface 6 reading systems, and give CPU dynamic partition module 2 with all CPU information and preset scheduling strategy type information delivery.
Scheduling decision module 4 is when virtual machine creating, reads the scheduling parameter of virtual machine, offers CPU dynamic partition module 2 subsequently, by CPU dynamic partition module 2 virtual machine is joined corresponding subregion; When moving, be responsible in each subregion, virtual machine being dispatched in system.At first when creating virtual machine, scheduling decision module 4 reads the scheduling parameter of guest virtual machine 8 through control interface 6, and scheduling parameter is offered CPU dynamic partition module 2; Whether CPU dynamic partition module 2 subregion of searching this scheduling strategy type according to scheduling parameter exists then; If exist, then this guest virtual machine 8 is joined this subregion, if do not exist; Then CPU dynamic partition module 2 is set up the corresponding subregion of this scheduling strategy type earlier, again guest virtual machine 8 is joined this subregion.When system in when operation, scheduling decision module 4 is responsible in each subregion, guest virtual machine being dispatched, and guarantees that the guest virtual machine in each subregion adopts corresponding scheduling strategy to dispatch.
CPU monitoring module 5, each subregion cpu load situation of monitoring in system's operational process, and 2 pairs of partitioned resources of notice CPU dynamic partition module are dynamically adjusted.System default is provided with bound for the load size of each subregion; As above be limited to 90%; Rolling off the production line is 40%, when a certain subregion load greater than 90% the time, notice CPU dynamic partition module 2; Resources allocation is to the big subregion of load from Free Partition for CPU dynamic partition module 2, and acquiescence is distributed a cpu resource at every turn; When a certain subregion load less than 40% the time; Notice CPU dynamic partition module 2; CPU dynamic partition module 2 reclaims resource in Free Partition from the little subregion of load, acquiescence is each to reclaim a cpu resource, but the rarest cpu resource in must the proof load little subregion.Under special circumstances, when the load of a certain subregion is 0%, show that then this subregion does not have scheduling virtual machine, then CPU dynamic partition module 2 is deleted these subregions, and resource is recovered in the Free Partition.
The present invention proposes the dispatching method of virtual machine under a kind of multi-core environment, and is as shown in Figure 2, comprises following steps:
Step 1: power up the startup initialization, its process flow diagram is as shown in Figure 3.Scheduling initialization module 3 loads during system start-up, carries out initial work.This step comprises following two operations: step 1.1 is enumerated CPU and is read preset scheduling strategy type; Step 1.2 is carried out primary partition for system.
Step 1.1 is enumerated CPU and is read preset scheduling strategy type; When system power-up is accomplished startup; Control terminal 7 loads scheduling initialization module 3; Scheduling initialization module 3 is enumerated out with all CPU in the system, the scheduling strategy type of presetting in the reading system simultaneously, and for example the scheduling strategy type of systemic presupposition has Credit, BVT and SEDF.
Step 1.2 is carried out primary partition for system; Scheduling initialization module 3 is got the subregion of resource minimum value by CPU dynamic partition module 2 for every kind of scheduling strategy type after obtaining the scheduling strategy type, the minimum value of resource generally by keeper's setting, is defaulted as 1 CPU.Set up Free Partition simultaneously and place remaining cpu resource, accomplish system's primary partition.For example when preset scheduling strategy type has BVT and SEDF; System can be divided into 3 subregions with cpu resource; Be respectively the first scheduling strategy subregion (the scheduling strategy type is BVT), the second scheduling strategy subregion (the scheduling strategy type is SEDF) and Free Partition; Wherein the first scheduling strategy subregion and the second scheduling strategy subregion initial resource quantity are 1 cpu resource, and remaining CPU is placed in the Free Partition.
Step 2: create virtual machine, as shown in Figure 2; When system creation guest virtual machine 8, scheduling decision module 4 can read the scheduling parameter of guest virtual machine 8, and guest virtual machine 8 is joined corresponding subregion.This step comprises following operation: step 2.1 scheduling decision module reads scheduling parameter; Step 2.2 joins corresponding subregion with virtual machine.
Step 2.1 scheduling decision module reads scheduling parameter:
As shown in Figure 2; When guest virtual machine 8 is created; System can formulate scheduling parameter according to the type that guest virtual machine 8 is used in configuration file; So-called scheduling parameter just is meant the scheduling strategy that guest virtual machine 8 application are fit to, and for example uses in real time to be fit to use the BVT scheduling strategy, and scheduling parameter just will be configured to " BVT " so; The configuration file that scheduling decision module 4 can initiatively read guest virtual machine 8 obtains scheduling parameter.For example; As shown in Figure 1, create a guest virtual machine 8, this virtual machine is will be as the application of a Voip; In the configuration file of guest virtual machine 8, scheduling parameter is arranged to " BVT ", the configuration file that scheduling decision module 4 can initiatively read this virtual machine obtains scheduling parameter " BVT ".
Step 2.2 joins corresponding subregion with virtual machine:
After scheduling parameter obtains; Scheduling decision module 4 is searched corresponding subregion through scheduling parameter; When corresponding subregion exists, be about to guest virtual machine 8 and join this subregion scheduling, for example when the corresponding subregion of BVT scheduling strategy be that the first scheduling strategy subregion is when existing; As shown in Figure 1, guest virtual machine 8 will be added in the first scheduling strategy subregion; When corresponding subregion does not exist; CPU dynamic partition module 2 is the newly-built subregion of this scheduling strategy type, and this virtual machine is joined this subregion, for example; When the corresponding subregion of BVT scheduling strategy does not exist; Will set up the corresponding subregion of BVT scheduling strategy, and, then guest virtual machine 8 joined in the corresponding subregion of BVT scheduling strategy for it distributes i.e. 1 cpu resource of least resource.
Step 3 scheduling virtual machine, as shown in Figure 2; When system moves, CPU monitoring module 5, scheduling decision module 4,2 each the Xingqi duty of CPU dynamic partition module are dispatched guest virtual machine 8, and this step comprises following operation: step 3.1 virtual machine is being transferred in the subregion separately; Step 3.2 each subregion loading condition of monitoring is also dynamically adjusted partition size.
Step 3.1 virtual machine is being transferred in the subregion separately;
When system moves; Scheduling decision module 4 is dispatched each virtual machine in each subregion, and each subregion adopts different scheduling strategies, for example; Guest virtual machine in the corresponding subregion of BVT scheduling strategy can only adopt the BVT scheduling strategy to dispatch in this subregion.
Step 3.2 each subregion loading condition of monitoring is also dynamically adjusted partition size;
The load of 5 pairs of each subregions of CPU monitoring module is monitored; When the load of subregion is excessive; Can activate CPU dynamic partition module 2 inspection Free Partitions, be not under the situation of sky at Free Partition, can resource in the Free Partition (giving tacit consent to each 1 cpu resource that distributes) be distributed to load and cross bigdos; Enlarge load and cross the resource of bigdos; When if Free Partition is empty, can obtain the relatively little partitioned resources of load (acquiescence is told a cpu resource at every turn) and distribute to the excessive subregion of load, enlarge the resource that bigdos is crossed in load; When the load of subregion is too small; CPU dynamic partition module 2 can the partitioned resources that load is little be recovered to Free Partition (cpu resource of the each recovery of acquiescence), dwindles the little partitioned resources of load, under special circumstances; When the load of a certain subregion is 0%; Show that then this subregion does not have scheduling virtual machine, then CPU dynamic partition module 2 is deleted these subregions, and resource is recovered in the Free Partition.
Claims (5)
1. one kind based on the dispatching method of virtual machine under the multi-core environment; It is characterized in that; This method has added scheduling initialization module, CPU monitoring module, scheduling decision module and CPU dynamic partition module in existing virtual machine monitor VMM, specifically comprise following execution in step:
Step 1: power up the startup initialization; The scheduling initialization module loads during system start-up, carries out initial work, and this step comprises following two operations:
Step 1.1 is enumerated CPU and is read preset scheduling strategy type; When system power-up was accomplished startup, the scheduling initialization module loaded, and all CPU in the system are enumerated out, simultaneously preset scheduling strategy type in the reading system;
Step 1.2 is carried out primary partition for system; The scheduling initialization module is that every kind of scheduling strategy type is got the subregion of resource minimum value by CPU dynamic partition module after obtaining the scheduling strategy type, sets up Free Partition simultaneously in order to place remaining cpu resource, accomplishes system's primary partition;
Step 2: create virtual machine; When the system creation virtual machine, the scheduling decision module can read the scheduling parameter of virtual machine, and virtual machine is joined corresponding subregion, and this step comprises following two operations:
Step 2.1 scheduling decision module reads scheduling parameter; When new establishment virtual machine, system can formulate scheduling parameter according to the type that virtual machine is used in configuration file, and the configuration file that the scheduling decision module can initiatively read virtual machine obtains scheduling parameter;
Step 2.2 joins corresponding subregion with virtual machine; After scheduling parameter obtains; The scheduling decision module is searched corresponding subregion through scheduling parameter; When corresponding subregion existed, CPU dynamic partition module directly joined corresponding subregion scheduling with this virtual machine, when the corresponding subregion of scheduling parameter does not exist; CPU dynamic partition module is told a new subregion as this scheduling strategy categories subarea from Free Partition, and this virtual machine is added scheduling;
Step 3: scheduling virtual machine; When system moved, this step specifically comprised following operation:
Step 3.1 virtual machine is being dispatched in the subregion separately; The virtual machine of operation can only be dispatched in the CPU of appointment subregion, and the scheduling decision module is dispatched the virtual machine in each subregion according to the dispatching algorithm of each subregion;
Step 3.2 each subregion loading condition of monitoring is also dynamically adjusted partition size; The cpu resource that the CPU monitoring module is being kept watch on each subregion utilizes situation; When the cpu load of certain subregion is excessive; The CPU monitoring module triggers CPU dynamic partition module, and CPU dynamic partition module is told cpu resource to the excessive subregion of load from Free Partition, when Free Partition is sky; From the less subregion of load, tell cpu resource in the excessive subregion of load; When the cpu load of subregion was too small, CPU dynamic partition module can the partitioned resources that load is little be recovered to Free Partition, dwindles the little partitioned resources of load.
2. according to claim 1ly a kind ofly it is characterized in that based on the dispatching method of virtual machine under the multi-core environment scheduling strategy type described in the step 1 comprises Credit dispatching algorithm, BVT dispatching algorithm and SEDF dispatching algorithm.
3. according to claim 1 a kind ofly it is characterized in that based on the dispatching method of virtual machine under the multi-core environment, the resource minimum value described in the step 1, this value is set by the keeper, is defaulted as 1 CPU.
4. according to claim 1 a kind ofly it is characterized in that based on the dispatching method of virtual machine under the multi-core environment, the scheduling parameter described in the step 2 is meant that virtual machine uses the scheduling strategy that is fit to.
5. according to claim 1ly a kind ofly it is characterized in that the size of the adjustment subregion described in the step 3.2 based on the dispatching method of virtual machine under the multi-core environment; The size of the each adjustment of acquiescence subregion is a cpu resource; In particular cases, when the load of a certain subregion is 0%, show that then this subregion does not have scheduling virtual machine; Then CPU dynamic partition module is deleted this subregion, and resource is recovered in the Free Partition.
Priority Applications (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN200910241371A CN101706743B (en) | 2009-12-07 | 2009-12-07 | Dispatching method of virtual machine under multi-core environment |
Applications Claiming Priority (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN200910241371A CN101706743B (en) | 2009-12-07 | 2009-12-07 | Dispatching method of virtual machine under multi-core environment |
Publications (2)
| Publication Number | Publication Date |
|---|---|
| CN101706743A CN101706743A (en) | 2010-05-12 |
| CN101706743B true CN101706743B (en) | 2012-09-05 |
Family
ID=42376969
Family Applications (1)
| Application Number | Title | Priority Date | Filing Date |
|---|---|---|---|
| CN200910241371A Expired - Fee Related CN101706743B (en) | 2009-12-07 | 2009-12-07 | Dispatching method of virtual machine under multi-core environment |
Country Status (1)
| Country | Link |
|---|---|
| CN (1) | CN101706743B (en) |
Families Citing this family (18)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN101984414B (en) * | 2010-11-05 | 2013-08-07 | 中兴通讯股份有限公司 | Method and device for scheduling central processing unit (CPU) resource |
| CN102651729A (en) * | 2011-02-23 | 2012-08-29 | 中国移动通信集团公司 | Resource configuration method and device |
| CN102270159B (en) * | 2011-08-11 | 2013-02-27 | 浙江大学 | Access controlling and load balancing method for virtualized environment |
| CN103186410B (en) * | 2011-12-28 | 2016-08-17 | 英业达股份有限公司 | The collocation method of virtual machine and the server system using the method |
| CN103294554A (en) * | 2012-03-05 | 2013-09-11 | 中兴通讯股份有限公司 | SOC multiprocessor dispatching method and apparatus |
| CN102779075B (en) * | 2012-06-28 | 2014-12-24 | 华为技术有限公司 | Method, device and system for scheduling in multiprocessor nuclear system |
| CN103180830B (en) * | 2012-08-31 | 2016-10-12 | 华为技术有限公司 | Central processing unit resource allocation methods and calculating node |
| CN106126319B (en) * | 2012-08-31 | 2019-11-01 | 华为技术有限公司 | Central processing unit resource allocation methods and calculate node |
| CN102929689B (en) * | 2012-11-01 | 2016-05-18 | 曙光云计算技术有限公司 | In cloud computing system, create the method for virtual machine |
| CN103051564B (en) * | 2013-01-07 | 2016-03-30 | 杭州华三通信技术有限公司 | The method and apparatus of dynamic resource allocation |
| CN103428099B (en) * | 2013-08-21 | 2016-02-24 | 中国人民解放军国防科学技术大学 | A kind of method of universal multi-core network processor flow control |
| CN103645935B (en) * | 2013-12-17 | 2017-01-18 | 山东中创软件工程股份有限公司 | Device and method for adjusting JVM (java virtual machine) parameters in domestic environment |
| CN104714845B (en) * | 2013-12-17 | 2018-06-26 | 华为技术有限公司 | Resource dynamic regulation method, device and more kernel operating systems |
| US9813335B2 (en) * | 2014-08-05 | 2017-11-07 | Amdocs Software Systems Limited | System, method, and computer program for augmenting a physical system utilizing a network function virtualization orchestrator (NFV-O) |
| CN103970587B (en) * | 2014-04-21 | 2018-02-13 | 华为技术有限公司 | A kind of method, apparatus and system of scheduling of resource |
| CN107122233B (en) * | 2017-03-27 | 2020-08-28 | 西安电子科技大学 | TSN service-oriented multi-VCPU self-adaptive real-time scheduling method |
| CN113656141A (en) * | 2020-05-12 | 2021-11-16 | 华为技术有限公司 | Input/output I/O request processing method and host equipment |
| CN111580937B (en) * | 2020-06-17 | 2022-11-04 | 中国人民解放军国防科技大学 | Automatic virtual machine scheduling method for Feiteng multi-core/many-core hybrid cluster |
Citations (3)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN1391671A (en) * | 1999-09-28 | 2003-01-15 | 国际商业机器公司 | Workload management in computing environment |
| CN1619498A (en) * | 2003-11-21 | 2005-05-25 | 国际商业机器公司 | Method and system for dynamically adjusting the workload of an active resource |
| CN101499021A (en) * | 2008-01-31 | 2009-08-05 | 国际商业机器公司 | Method and apparatus for dynamically distributing resources on a plurality of virtual machines |
-
2009
- 2009-12-07 CN CN200910241371A patent/CN101706743B/en not_active Expired - Fee Related
Patent Citations (3)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN1391671A (en) * | 1999-09-28 | 2003-01-15 | 国际商业机器公司 | Workload management in computing environment |
| CN1619498A (en) * | 2003-11-21 | 2005-05-25 | 国际商业机器公司 | Method and system for dynamically adjusting the workload of an active resource |
| CN101499021A (en) * | 2008-01-31 | 2009-08-05 | 国际商业机器公司 | Method and apparatus for dynamically distributing resources on a plurality of virtual machines |
Also Published As
| Publication number | Publication date |
|---|---|
| CN101706743A (en) | 2010-05-12 |
Similar Documents
| Publication | Publication Date | Title |
|---|---|---|
| CN101706743B (en) | Dispatching method of virtual machine under multi-core environment | |
| Zhong et al. | A cost-efficient container orchestration strategy in kubernetes-based cloud computing infrastructures with heterogeneous resources | |
| CN103597449B (en) | The Heterogeneous Computing of operating system decoupling | |
| Buyya et al. | Energy-efficient management of data center resources for cloud computing: a vision, architectural elements, and open challenges | |
| Beloglazov et al. | A taxonomy and survey of energy-efficient data centers and cloud computing systems | |
| US8250574B2 (en) | Virtual machine management via use of table in which virtual machine information is registered on a time basis | |
| CN103365700B (en) | A kind of facing cloud calculates monitoring resource and the adjustment System of virtualized environment | |
| CN100511151C (en) | Multiple-path multiple-core server and CPU virtualization processing method thereof | |
| CN102521055B (en) | Virtual machine resource allocating method and virtual machine resource allocating system | |
| CN102096596A (en) | Cloud computing service Cache system based on internal memory template of virtual machine | |
| CN102662750A (en) | Virtual machine resource optimal control method and control system based on elastic virtual machine pool | |
| CN103179048B (en) | Main frame qos policy transform method and the system of cloud data center | |
| CN102708000A (en) | System and method for realizing energy consumption control through virtual machine migration | |
| Lee et al. | A dynamic resource management with energy saving mechanism for supporting cloud computing | |
| CN102081554A (en) | Cloud computing operating system as well as kernel control system and method thereof | |
| CN105404549B (en) | Scheduling virtual machine system based on yarn framework | |
| CN105068873A (en) | Isomerous virtual resource and task scheduling method and system | |
| CN107203255A (en) | Power-economizing method and device are migrated in a kind of network function virtualized environment | |
| Sharma et al. | Managing risk in a derivative IaaS cloud | |
| Al Shayeji et al. | An energy-aware virtual machine migration algorithm | |
| CN112764909A (en) | Sharing method and system based on cloud architecture workstation | |
| CN105100180A (en) | Cluster node dynamic loading method, device and system | |
| CN104468710A (en) | Mixed big data processing system and method | |
| US10007559B1 (en) | Virtual tiering | |
| CN107203256A (en) | Energy-conservation distribution method and device under a kind of network function virtualization scene |
Legal Events
| Date | Code | Title | Description |
|---|---|---|---|
| C06 | Publication | ||
| PB01 | Publication | ||
| C10 | Entry into substantive examination | ||
| SE01 | Entry into force of request for substantive examination | ||
| C14 | Grant of patent or utility model | ||
| GR01 | Patent grant | ||
| CF01 | Termination of patent right due to non-payment of annual fee |
Granted publication date: 20120905 Termination date: 20141207 |
|
| EXPY | Termination of patent right or utility model |