CN104951351A - Virtual machine dispatcher, dispatching demand manager and method and virtual machine dispatching system - Google Patents

Virtual machine dispatcher, dispatching demand manager and method and virtual machine dispatching system Download PDF

Info

Publication number
CN104951351A
CN104951351A CN201410122663.0A CN201410122663A CN104951351A CN 104951351 A CN104951351 A CN 104951351A CN 201410122663 A CN201410122663 A CN 201410122663A CN 104951351 A CN104951351 A CN 104951351A
Authority
CN
China
Prior art keywords
virtual machine
scheduling
dispatching requirement
requirement manager
dispatching
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.)
Granted
Application number
CN201410122663.0A
Other languages
Chinese (zh)
Other versions
CN104951351B (en
Inventor
陶一然
杨鹏
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.)
Hitachi Ltd
Original Assignee
Hitachi 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 Hitachi Ltd filed Critical Hitachi Ltd
Priority to CN201410122663.0A priority Critical patent/CN104951351B/en
Publication of CN104951351A publication Critical patent/CN104951351A/en
Application granted granted Critical
Publication of CN104951351B publication Critical patent/CN104951351B/en
Expired - Fee Related legal-status Critical Current
Anticipated expiration legal-status Critical

Links

Abstract

The invention relates to a virtual machine dispatcher, a dispatching demand manager and method and a virtual machine dispatching system. A virtual machine dispatching method is provided. The method is characterized by comprising the following steps that a dispatching requirement from a virtual machine of the dispatching demand manager is monitored, it is judged whether the dispatching requirement comprises dispatching resolution information which is reached by the dispatching demand manager and an adjacent dispatching demand manager or not, under the condition that the dispatching requirement comprises dispatching resolution information which is reached by the dispatching demand manager and the adjacent dispatching demand manager, a dispatching command is sent to the dispatching demand manager and the adjacent dispatching demand manager, and under the condition that a dispatching resolution is reached, the dispatching requirement further comprises information relevant to the adjacent dispatching demand manager.

Description

Scheduling virtual machine device, dispatching requirement manager and method and scheduling virtual machine system
Technical field
The present invention relates to distributed system and computer network field, more particularly, the present invention relates to a kind of scheduling virtual machine device, dispatching requirement manager and method thereof and scheduling virtual machine system.
Background technology
Cloud computing technology is flourish in recent years, and cloud computing technology improves the utilization factor of resource (such as hardware resource), reduces cost simultaneously, and support mass data analysis, storage, search and calamity in have fine performance.Meanwhile, the platform (cloud platform) based on cloud computing technology has good extensibility and balance load-bearing capacity.But, set up cloud platform and also technology had higher requirements simultaneously, still have a lot of problem to complete.Such as, how schedules system resources to improve resource utilization and balanced load is exactly a stubborn problem.
In order to change this situation, propose the scheduling virtual machine technology based on cloud platform, such as publication number is the patented claim (hereinafter referred to as patent documentation 1) of CN101719081A.Mention a kind of dispatching method of virtual machine in this patent documentation 1, utilize a scheduling virtual machine device, adopt bin packing algorithm to solve the problem (with reference to Fig. 1, Fig. 2 of the present invention) of data center resource calculating and Real-Time Scheduling virtual machine.
Scheme in patent documentation 1 has accomplished computational resource and scheduling, but, have employed single scheduling virtual machine device in the program, and need to use bin packing algorithm calculate system-wide resource after United Dispatching, be difficult to tackle the scheduling virtual machine problem of large scale system.In large scale system, video data formation speed is fast, change is large, and the pattern of unified calculation, United Dispatching is difficult to respond demand in time.
Summary of the invention
Based on above description, desirable to provide a kind of can better schedules system resources to improve resource utilization and the scheduling virtual machine device of balanced load, dispatching requirement manager and method and scheduling virtual machine system.
(dispatching method of virtual machine of underlying virtual machine scheduler)
According to a first aspect of the invention, provide a kind of dispatching method of virtual machine, it is characterized in that, comprising:
Monitor the dispatch request from the virtual machine of dispatching requirement manager,
Judge whether to comprise in this dispatch request to represent that described dispatching requirement manager has been reached with contiguous dispatching requirement manager and dispatch the information of resolving,
When this dispatch request comprise represent described dispatching requirement manager and contiguous dispatching requirement manager reached dispatch the information of resolving, dispatch command is sent respectively to described dispatching requirement manager and described contiguous dispatching requirement manager, wherein when having reached scheduling resolution, this dispatch request also comprises the information about described contiguous dispatching requirement manager.
By adopting the program, underlying virtual machine scheduler is without the need to dispatching according to concrete physical server information on load, and only whether comprise according to dispatch request and represent that described dispatching requirement manager and contiguous dispatching requirement manager have been reached and dispatch the information of resolving and carry out scheduling virtual machine, simplify virtual machine and be taken after mixing with liquid method.
According to a second aspect of the invention, in the dispatching method of virtual machine in first,
When this dispatch request do not comprise represent described dispatching requirement manager and described contiguous dispatching requirement manager reached dispatch the information of resolving, monitor the dispatch request of the virtual machine from other dispatching requirement manager,
According to the demand of the dispatch request of the virtual machine of described dispatching requirement manager with the dispatch request of the virtual machine of other dispatching requirement manager, judge whether the dispatch request of the dispatch request of the virtual machine of described dispatching requirement manager and the virtual machine of other dispatching requirement manager described can mate
If can mate, then send the information about the other side respectively to described dispatching requirement manager and other dispatching requirement manager described, and send dispatch command,
Wherein, whether the dispatch request of the virtual machine of described dispatching requirement manager and the dispatch request of other dispatching requirement manager described can mate expression: whether the request that the virtual machine of the request that the virtual machine of described dispatching requirement manager is moved out and other dispatching requirement manager described is moved into can mate; Or whether the request that the virtual machine of the request that the virtual machine of described dispatching requirement manager is moved into and other dispatching requirement manager described is moved out can mate.
By adopting the program, underlying virtual machine scheduler when this dispatch request do not comprise represent described dispatching requirement manager and described contiguous dispatching requirement manager reached dispatch the information of resolving, not only report an error or reporting scheduling request, but the dispatch request monitored from the virtual machine of other dispatching requirement manager, and when mating, send the information about the other side respectively to described dispatching requirement manager and other dispatching requirement manager described, and send dispatch command.Thus, the success ratio of scheduling can be improved.
(dispatching method of virtual machine of the underlying virtual machine scheduler of Multi-Level Virtual Machine scheduler)
According to a third aspect of the present invention, in the dispatching method of virtual machine in first,
When this dispatch request do not comprise represent described dispatching requirement manager and contiguous dispatching requirement manager reached dispatch the information of resolving, judge whether to there is last layer scheduling virtual machine device,
When there is last layer scheduling virtual machine device, above-mentioned dispatch request is reported last layer scheduling virtual machine device, and pass the information about other dispatching requirement manager received from last layer scheduling virtual machine device and dispatch command back described dispatching requirement manager.
By adopt the program, utilize multi-stage scheduling scheme, can better schedules system resources to improve resource utilization and balanced load.
According to a fourth aspect of the present invention, in the dispatching method of virtual machine in second,
When this dispatch request do not comprise represent described dispatching requirement manager be close to dispatching requirement manager reached the information of dispatching and resolving and the dispatch request of described dispatching requirement manager fails to mate with the dispatch request of other dispatching requirement manager, judge whether to there is last layer scheduling virtual machine device
When there is last layer scheduling virtual machine device, above-mentioned dispatch request is reported last layer scheduling virtual machine device, and pass the information about other dispatching requirement manager received from last layer scheduling virtual machine device and dispatch command back described dispatching requirement manager.
By adopt the program, utilize multi-stage scheduling scheme, can better schedules system resources to improve resource utilization and balanced load.
According to a fifth aspect of the present invention, in the dispatching method of virtual machine in first,
Described dispatch request comprises the request of being moved out by virtual machine or the request of being moved into by virtual machine.
(dispatching method of virtual machine of non-underlying virtual machine scheduler)
According to a sixth aspect of the invention, a kind of dispatching method of virtual machine is provided, it is characterized in that,
Monitor the dispatch request from two or more lower one deck scheduling virtual machine devices, described dispatch request comprises the request of being moved out by virtual machine or the request of being moved into by virtual machine,
The demand of the request of moving out according to the virtual machine from lower one deck scheduling virtual machine device and the request that the virtual machine of one deck scheduling virtual machine device is moved under another, judge whether the request that the virtual machine of the request that the virtual machine of described lower one deck scheduling virtual machine device is moved out and another lower one deck scheduling virtual machine device described is moved into can mate
If the request that the virtual machine of the request that the virtual machine of described lower one deck scheduling virtual machine device is moved out and another lower one deck scheduling virtual machine device described is moved into can be mated, then send the information about the relevant dispatching requirement manager in the dispatch request of the other side respectively to described lower one deck scheduling virtual machine device and another lower one deck scheduling virtual machine device described, and send dispatch command.
By adopt the program, utilize multi-stage scheduling scheme, can better schedules system resources to improve resource utilization and balanced load.
According to a seventh aspect of the present invention, in the 6th dispatching method of virtual machine in,
If the request that the virtual machine of the request that the virtual machine of described lower one deck scheduling virtual machine device is moved out and another lower one deck scheduling virtual machine device described is moved into can not be mated, judge whether to there is last layer scheduling virtual machine device,
When there is last layer scheduling virtual machine device, above-mentioned dispatch request is reported last layer scheduling virtual machine device, and pass the information about other dispatching requirement manager received from last layer scheduling virtual machine device and dispatch command back lower one deck scheduling virtual machine device.
By adopt the program, utilize multi-stage scheduling scheme, can better schedules system resources to improve resource utilization and balanced load.
(the dispatching requirement management method of dispatching requirement manager)
According to an eighth aspect of the present invention, a kind of dispatching requirement management method is provided, it is characterized in that,
Physical server information on load is obtained from physical machine watch-dog,
According to this physical server information on load computational physics server load state, and judge whether the scheduling needing virtual machine,
When being judged as the scheduling needing virtual machine, communicate with contiguous dispatching requirement manager, whether inquiry accepts the scheduling of virtual machine,
When receive from described contiguous dispatching requirement manager comprise represent described contiguous dispatching requirement manager accept the information of the scheduling of virtual machine, reach to dispatch with described contiguous dispatching requirement manager and resolve, and send to comprise to scheduling virtual machine device and represent described dispatching requirement manager and the dispatch request of being close to dispatching requirement manager and having reached the information of dispatching and resolving and the virtual machine about the information of described contiguous dispatching requirement manager
When receive from described contiguous dispatching requirement manager do not comprise represent described contiguous dispatching requirement manager accept the information of the scheduling of virtual machine, send not comprise to described scheduling virtual machine device and represent that described dispatching requirement manager has reached the dispatch request of dispatching the virtual machine of the information of resolving with contiguous dispatching requirement manager
According to the dispatch command from described scheduling virtual machine device or the information about other dispatching requirement manager and dispatch command, instruction virtual machine manager carries out scheduling virtual machine.
By adopting the program, dispatching requirement manager is not simply physical server information on load is sent to scheduling virtual machine device, but according to this physical server information on load computational physics server load state, and judge whether the scheduling needing virtual machine, when being judged as the scheduling needing virtual machine, communicate with contiguous dispatching requirement manager, whether inquiry accepts the scheduling of virtual machine, when receive from described contiguous dispatching requirement manager comprise represent described contiguous dispatching requirement manager accept the information of the scheduling of virtual machine, reach to dispatch with described contiguous dispatching requirement manager and resolve, and send to scheduling virtual machine device.The work load of scheduling virtual machine device can be reduced thus.
According to a ninth aspect of the present invention, in the dispatching requirement management method in the 8th,
The scheduling of virtual machine is moving into of moving out of virtual machine or virtual machine.
According to a tenth aspect of the present invention, in the dispatching requirement management method in the 8th,
When receiving the inquiry of the scheduling whether accepting virtual machine from described contiguous dispatching requirement manager, according to the physical server information on load obtained from physical machine watch-dog, judge whether the scheduling accepting virtual machine, and judged result is sent to described contiguous dispatching requirement manager
When being judged as the scheduling accepting virtual machine, reaching to dispatch with described contiguous dispatching requirement manager and resolving.
(underlying virtual machine scheduler)
According to an eleventh aspect of the present invention, a kind of scheduling virtual machine device is provided, it is characterized in that having:
Monitoring portion, monitors the dispatch request of the virtual machine from dispatching requirement manager;
Whether scheduling controlling portion, judge to comprise in this dispatch request to represent that described dispatching requirement manager has been reached with contiguous dispatching requirement manager and dispatch the information of resolving; And
Dispatch command portion, when this dispatch request comprise represent described dispatching requirement manager and contiguous dispatching requirement manager reached dispatch the information of resolving, dispatch command is sent respectively to described dispatching requirement manager and described contiguous dispatching requirement manager, wherein when having reached scheduling resolution, this dispatch request also comprises the information about described contiguous dispatching requirement manager.
According to a twelfth aspect of the present invention, in the scheduling virtual machine device in the 11,
When this dispatch request do not comprise represent described dispatching requirement manager and contiguous dispatching requirement manager reached dispatch the information of resolving, the monitoring of described monitoring portion from the dispatch request of the virtual machine of other dispatching requirement manager,
Described scheduling controlling portion is according to the demand of the dispatch request of the virtual machine of described dispatching requirement manager with the dispatch request of the virtual machine of other dispatching requirement manager, judge whether the dispatch request of the dispatch request of the virtual machine of described dispatching requirement manager and the virtual machine of other dispatching requirement manager described can mate
If can mate, described scheduling controlling portion sends the information about the other side respectively to described dispatching requirement manager and other dispatching requirement manager described, and sends dispatch command,
Wherein, whether the dispatch request of the virtual machine of described dispatching requirement manager and the dispatch request of other dispatching requirement manager described can mate expression: whether the request that the virtual machine of the request that the virtual machine of described dispatching requirement manager is moved out and other dispatching requirement manager described is moved into can mate; Or whether the request that the virtual machine of the request that the virtual machine of described dispatching requirement manager is moved into and other dispatching requirement manager described is moved out can mate.
(the underlying virtual machine scheduler in Multi-Level Virtual Machine scheduler)
According to the 13 aspect of the present invention, in the scheduling virtual machine device in the 11, also have:
First judging part, when this dispatch request do not comprise represent described dispatching requirement manager and contiguous dispatching requirement manager reached dispatch the information of resolving, judge whether to there is last layer scheduling virtual machine device; And
Dispatch request portion, when there is last layer scheduling virtual machine device, above-mentioned dispatch request is reported last layer scheduling virtual machine device, and pass the information about other dispatching requirement manager received from last layer scheduling virtual machine device and dispatch command back described dispatching requirement manager.
According to the 14 aspect of the present invention, in the scheduling virtual machine device in the 12, also have:
Second judging part, when this dispatch request do not comprise represent described dispatching requirement manager be close to dispatching requirement manager reached the information of dispatching and resolving and the dispatch request of described dispatching requirement manager fails to mate with the dispatch request of other dispatching requirement manager, judge whether to there is last layer scheduling virtual machine device
Dispatch request portion, when there is last layer scheduling virtual machine device, above-mentioned dispatch request is reported last layer scheduling virtual machine device, and pass the information about other dispatching requirement manager received from last layer scheduling virtual machine device and dispatch command back described dispatching requirement manager.
According to the 15 aspect of the present invention, in the scheduling virtual machine device in the 11,
Described dispatch request is the request of being moved out by virtual machine or the request of being moved into by virtual machine.
(non-underlying virtual machine scheduler)
According to the 16 aspect of the present invention, a kind of scheduling virtual machine device is provided, it is characterized in that having:
Monitoring portion, monitor the dispatch request from two or more lower one deck scheduling virtual machine devices, described dispatch request comprises the request of being moved out by virtual machine or the request of being moved into by virtual machine; And
Scheduling controlling portion, the demand of the request of moving out according to the virtual machine from lower one deck scheduling virtual machine device and the request that the virtual machine of one deck scheduling virtual machine device is moved under another, judge whether the request that the virtual machine of the request that the virtual machine of described lower one deck scheduling virtual machine device is moved out and another lower one deck scheduling virtual machine device described is moved into can mate
Dispatch command portion, if the request that the virtual machine of the request that the virtual machine of described lower one deck scheduling virtual machine device is moved out and another lower one deck scheduling virtual machine device described is moved into can be mated, then send the information about the relevant dispatching requirement manager in the other side's dispatch request respectively to described lower one deck scheduling virtual machine device and another lower one deck scheduling virtual machine device described, and send dispatch command.
According to the 17 aspect of the present invention, in the scheduling virtual machine device in the 16, also possess:
3rd judging part, if the request that the virtual machine of the request of moving out of the virtual machine of described lower one deck scheduling virtual machine device and another lower one deck scheduling virtual machine device described is moved into can not be mated, judges whether to there is last layer scheduling virtual machine device,
Dispatch request portion, when there is last layer scheduling virtual machine device, above-mentioned dispatch request is reported last layer scheduling virtual machine device, and pass the information about other dispatching requirement manager received from last layer scheduling virtual machine device and dispatch command back lower one deck scheduling virtual machine device.
(dispatching requirement manager)
According to the 18 aspect of the present invention, a kind of dispatching requirement manager is provided, it is characterized in that possessing:
Physical server information on load acquisition unit, obtains physical server information on load from physical machine watch-dog;
Scheduling judging part, according to this physical server information on load computational physics server load state, and judges whether the scheduling needing virtual machine;
Dispatching communication portion, when being judged as the scheduling needing virtual machine, communicate with contiguous dispatching requirement manager, whether inquiry accepts the scheduling of virtual machine;
Scheduling resolution portion, when receive from described contiguous dispatching requirement manager comprise represent described contiguous dispatching requirement manager accept the information of the scheduling of virtual machine, reach to dispatch with described contiguous dispatching requirement manager and resolve, and send to comprise to scheduling virtual machine device and represent described dispatching requirement manager and the dispatch request of being close to dispatching requirement manager and having reached the information of dispatching and resolving and the virtual machine about the information of described contiguous dispatching requirement manager
Described scheduling resolution portion when receive from described contiguous dispatching requirement manager do not comprise represent that described contiguous dispatching requirement manager accepts the information of the scheduling of virtual machine, send not comprise to described scheduling virtual machine device and represent that described dispatching requirement manager has reached the dispatch request of dispatching the virtual machine of the information of resolving with contiguous dispatching requirement manager; And
Dispatch command portion, according to the dispatch command from described scheduling virtual machine device or the information about other dispatching requirement manager and dispatch command, instruction virtual machine manager carries out scheduling virtual machine.
According to nineteen aspect of the present invention, in the dispatching requirement manager in the 18,
The scheduling of virtual machine is moving into of moving out of virtual machine or virtual machine.
According to the 20 aspect of the present invention, in the dispatching requirement manager in the 18, it is characterized in that,
When receiving the inquiry of the scheduling whether accepting virtual machine from described contiguous dispatching requirement manager, according to the physical server information on load obtained from physical machine watch-dog, judge whether the scheduling accepting virtual machine, and judged result is sent to described contiguous dispatching requirement manager
When dispatching judging part and being judged as the scheduling accepting virtual machine, described scheduling resolution portion and described contiguous dispatching requirement pipe
(only having the scheduling virtual machine system of a scheduling virtual machine device)
According to the 21 aspect of the present invention, a kind of scheduling virtual machine system is provided, it is characterized in that, comprising:
The scheduling virtual machine device of the 11 aspect or the 12 aspect; And
Multiple physical server, each physical server comprises the dispatching requirement manager of one of 20 aspects in one or more virtual machine, physical machine watch-dog, virtual machine manager and the 18 aspect-the,
Wherein, described virtual machine is distributed on described physical server, provides service to user, can run a virtual machine or run multiple virtual machine simultaneously in each physical server,
Wherein, every platform physical server runs physical machine watch-dog, the physical server information on load of the real-time monitors physical server of physical machine watch-dog, and send physical server information on load to dispatching requirement manager.
(there is the scheduling virtual machine system of Multi-Level Virtual Machine scheduler)
According to the 22 aspect of the present invention, a kind of scheduling virtual machine system is provided, it is characterized in that, comprising:
The scheduling virtual machine device of multiple 13 aspects or the 14 aspect is as underlying virtual machine scheduler;
The scheduling virtual machine device of one or more 16 or the 17 aspect is as non-underlying virtual machine scheduler;
Multiple physical server, each physical server comprises the dispatching requirement manager of one of 20 aspects in one or more virtual machine, physical machine watch-dog, virtual machine manager and the 18 aspect-the,
Wherein, described virtual machine is distributed on described physical server, provides service to user, can run a virtual machine or run multiple virtual machine simultaneously in each physical server,
Wherein, every platform physical server runs physical machine watch-dog, the physical server information on load of the real-time monitors physical server of physical machine watch-dog, and send physical server information on load to dispatching requirement manager.
According to the 23 aspect of the present invention, a kind of video monitoring system is provided, it is characterized in that, there is the scheduling virtual machine device described in any one in the 11 aspect-15 aspects, the scheduling virtual machine system of the scheduling virtual machine device described in any one in aspect, 20, the 16 aspect-the or the 21 aspect or the 22 aspect.
The present invention by adopt technique scheme, can better schedules system resources to improve resource utilization and balanced load.
Accompanying drawing explanation
In conjunction with the drawings disclosure illustrative embodiments is described in more detail, above-mentioned and other object of the present disclosure, Characteristics and advantages will become more obvious, wherein, in disclosure illustrative embodiments, identical reference number represents same parts usually.
Fig. 1 illustrates the structural drawing of the existing data center in above-mentioned patent documentation 1.
Fig. 2 illustrates the structural drawing of a kind of existing scheduling virtual machine system in above-mentioned patent documentation 1.
Fig. 3 illustrates the structural drawing of the scheduling virtual machine system of one embodiment of the present of invention.
Fig. 4 illustrates the stratal diagram of the scheduling virtual machine system of one embodiment of the present of invention.
Fig. 5 is the process flow diagram of the dispatching method of virtual machine in the scheduling virtual machine system of one embodiment of the present of invention.Fig. 5 a illustrates the process flow diagram of the dispatching method of virtual machine of an embodiment of underlying virtual machine scheduler.Fig. 5 b illustrates the process flow diagram of the dispatching method of virtual machine of an embodiment of underlying virtual machine scheduler.Fig. 5 c illustrates the process flow diagram of the dispatching method of virtual machine of underlying virtual machine scheduler when to have two-stage or more level scheduling virtual machine device in scheduling virtual machine system.
Fig. 6 is the process flow diagram of the dispatching method of virtual machine of non-underlying virtual machine scheduler in the scheduling virtual machine system of one embodiment of the present of invention.Fig. 6 a illustrates the process flow diagram of the dispatching method of virtual machine of an embodiment of the non-underlying virtual machine scheduler in scheduling virtual machine system.Fig. 6 b illustrates the process flow diagram of the dispatching method of virtual machine of an embodiment of middle layer scheduling virtual machine device in Multi-Level Virtual Machine dispatching system or top layer scheduling virtual machine device.
Fig. 7 illustrates the process flow diagram of the dispatching requirement management method of one embodiment of the present of invention.
Fig. 8 illustrates the message time sequence figure in the scheduling virtual machine system of one embodiment of the present of invention.
Fig. 9 illustrates the structured flowchart of underlying virtual machine scheduler of the present invention.
Figure 10 illustrates the structured flowchart of non-underlying virtual machine scheduler of the present invention.
Figure 11 illustrates the structured flowchart of dispatching requirement manager of the present invention.
Embodiment
Below with reference to accompanying drawings preferred implementation of the present disclosure is described in more detail.Although show preferred implementation of the present disclosure in accompanying drawing, but should be appreciated that, the disclosure can be realized in a variety of manners and not should limit by the embodiment of setting forth here.On the contrary, provide these embodiments to be to make the disclosure more thorough and complete, and the scope of the present disclosure intactly can be conveyed to those skilled in the art.
Person of ordinary skill in the field knows, the present invention can be implemented as device, system, method.Therefore, the disclosure can be implemented as following form, that is: can be completely hardware, also can be software (comprising firmware, resident software, microcode etc.) completely, can also be the form that hardware and software combines, be commonly referred to as " portion ", " device " or " system " herein.
The computer program code operated for performing the present invention can be write with one or more programming languages or its combination, described programming language comprises object oriented program language-such as Java, Smalltalk, C++, also comprises conventional process type programming language-such as " C " language or similar programming language.Program code can fully perform on the user computer, partly perform on the user computer, as one, independently software package performs, partly part performs on the remote computer or performs on remote computer or server completely on the user computer.In the situation relating to remote computer, remote computer can by the network of any kind---comprise LAN (Local Area Network) (LAN) or wide area network (WAN)-be connected to subscriber computer, or, outer computer (such as utilizing ISP to pass through Internet connection) can be connected to.
Below with reference to the method for the embodiment of the present invention, the process flow diagram of device (system) and/or block diagram, the present invention is described.Should be appreciated that the combination of each square frame in each square frame of process flow diagram and/or block diagram and process flow diagram and/or block diagram, can be realized by computer program instructions.These computer program instructions can be supplied to the processor of multi-purpose computer, special purpose computer or other programmable data treating apparatus, thus produce a kind of machine, these computer program instructions are performed by computing machine or other programmable data treating apparatus, create the device of the function/operation specified in the square frame in realization flow figure and/or block diagram.
Also can these computer program instructions be stored in the computer-readable medium that computing machine or other programmable data treating apparatus can be made to work in a specific way, like this, the instruction be stored in computer-readable medium just produces the manufacture (manufacture) of the command device (instruction means) of the function/operation specified in a square frame comprising in realization flow figure and/or block diagram.
Also can computer program instructions be loaded on computing machine, other programmable data treating apparatus or miscellaneous equipment, make to perform sequence of operations step on computing machine, other programmable data treating apparatus or miscellaneous equipment, to produce computer implemented process, thus make the instruction performed on computing machine or other programmable device can provide the process of the function/operation specified in the square frame in realization flow figure and/or block diagram.
Fig. 1 illustrates the structural drawing of the existing data center 1 in above-mentioned patent documentation 1.In the framework of current data center 1, have: data lan hub 2, physical server 3, specific schedule net 4, monitoring host computer 5 and centralized storage 6.One end of physical server 3 is connected to data lan hub 2, and the other end is connected to specific schedule net 4.The centralized storage 6 centralized stores data of this data center 1.
Fig. 2 illustrates the structural drawing of existing scheduling virtual machine system in above-mentioned patent documentation 1.This scheduling virtual machine system comprises scheduling virtual machine device 7, physical machine watch-dog 8, virtual machine 9, virtual machine manager 10.The physical machine status information that the real-time monitors physical monitor unit 8 of scheduling virtual machine device 7 transmits, passes judgment on resource utilization according to the physical machine status information received, uses dispatching algorithm to calculate optimal scheduling scheme, implement to dispatch to virtual machine 9.Virtual machine manager 10 real-time management is controlled oneself the virtual machine 9 in the physical machine of place.
Fig. 3 illustrates the structural drawing of the scheduling virtual machine system of one embodiment of the present of invention.The scheduling virtual machine system of one embodiment of the present of invention comprises top layer scheduling virtual machine device 11, scheduling virtual machine device 12 is (for different scheduling virtual machine devices 12, with 12a, 12b (not shown), 12m ... represent), physical machine watch-dog 13 is (for different physical machine watch-dogs 13, with 13a, 13b, 13m ... represent), virtual machine 14, dispatching requirement manager 15 is (for different dispatching requirement managers 15, with 15a, 15b, 15m ... represent) and virtual machine manager 16 (for different virtual machine managers 16, with 16a, 16b, 16m ... represent).Wherein, physical machine watch-dog 13, virtual machine 14, dispatching requirement manager 15 and virtual machine manager 16 are contained in physical server.
1) virtual machine 14: be distributed on physical server, provides service to user or miscellaneous equipment, and can run multiple stage virtual machine 14 in a physical server simultaneously.Each physical server in multiple physical server can be isomorphism, may not be isomorphism.If physical server is isomery physical server, then according to isomorphic relations, physical server is divided into multiple management domain, physical server included in each management domain is isomorphism physical server.
A physical host is invented multiple stage virtual machine by using virtual management program (or virtualization software) by physical server, and separate between each virtual machine, the various hardware resources on share host comprise CPU, storage, network etc.Physical server is by the virtual level of resources utilization that not only can improve whole data center IT environment, the operating load of whole cloud computing system can also be managed by technology such as scheduling virtual machines more efficiently, strengthen elasticity and the dirigibility of cloud computing system.
2) virtual machine manager 16: run a virtual machine manager 16 on every platform physical server, for virtual machine provides operation platform.
3) physical machine watch-dog 13: run a physical machine watch-dog 13 on every platform physical server, various loads on real-time monitors physical server 3, such as CPU usage, internal memory use number percent etc., and send information on load to dispatching requirement manager 15.
4) dispatching requirement manager 15: run a dispatching requirement manager 15 on every platform physical server.
Dispatching requirement manager 15 obtains the information on load of physical server from physical machine watch-dog 13, according to the information on load computational physics server load state of this physical server, and judge whether the scheduling needing virtual machine, when being judged as the scheduling needing virtual machine, communicate with contiguous dispatching requirement manager 15, whether inquiry accepts the scheduling of virtual machine, when the other side accepts the scheduling of virtual machine 14, reach to dispatch with described contiguous dispatching requirement manager 15 and resolve, and send to comprise to scheduling virtual machine device 12 and represent described dispatching requirement manager and the dispatch request of being close to dispatching requirement manager and having reached the information of dispatching and resolving and the virtual machine about the information of described contiguous dispatching requirement manager, when the other side does not accept the scheduling of virtual machine 14, send not comprise to described scheduling virtual machine device 12 and represent that described dispatching requirement manager 15 has reached the dispatch request of the virtual machine 14 dispatching the information of resolving with contiguous dispatching requirement manager 15, according to the dispatch command from described scheduling virtual machine device 12, or about the information of other dispatching requirement manager 15 and dispatch command, instruction virtual machine manager 16 carries out scheduling virtual machine.
Represent that described dispatching requirement manager 15 and contiguous dispatching requirement manager 15 have been reached and dispatch the information of resolve as long as comprise and can inform that the scheduling that dispatching requirement manager 15 described in scheduling virtual machine device 12 instruction and vicinity dispatching requirement manager 15 carry out virtual machine is just passable.
On the other hand, do not comprise and represent that described dispatching requirement manager 15 and contiguous dispatching requirement manager 15 have been reached and dispatch the information of resolving and can inform that scheduling virtual machine device 12 needs find other dispatching requirement manager 15 matched with the dispatching requirement of dispatching requirement manager 15.
5) scheduling virtual machine device 12: have one or more scheduling virtual machine device 12 in this scheduling virtual machine system, operates on special equipment, can think that the equipment at scheduling virtual machine device 12 place and physical server are isomeries.The no more than physical server of quantity of scheduling virtual machine device 12, and according to the network delay size between physical server, postponing to set up scheduling virtual machine device 12 between less physical server as far as possible.
Such as, there is physical server 3A, 3B, 3C, 3D (all not shown), and there is scheduling virtual machine device 12a and 12b (not shown), and the delay of physical server 3A and 3B is less, the delay of physical server 3C and 3D is less, but physical server 3A, 3B and physical server 3C, the delay of 3D is comparatively large (namely, physical server 3A and physical server 3C, 3D postpones larger, physical server 3B and physical server 3C, 3D postpones larger), then between physical server 3A and 3B, set up scheduling virtual machine device 12a, scheduling virtual machine device 12b is set up between physical server 3C and 3D.
6) top layer scheduling virtual machine device 11: top layer scheduling virtual machine device 11 and scheduling virtual machine device 12 can be the equipment of isomorphism also can be the equipment of tripe systems.Top layer scheduling virtual machine device 11 needs to connect all scheduling virtual machine devices 12, and there is not upper-level virtual machine scheduler 12.Top layer scheduling virtual machine device 11 only can receive the dispatch request from lower floor's scheduling virtual machine device 12, and finds the dispatching requirement that can mate.
Fig. 4 is the stratal diagram of the scheduling virtual machine system of one embodiment of the present of invention, wherein comprises three layers: scheduling virtual machine layer 401, Physical layer 407 and virtual machine layer 405.Top layer scheduling virtual machine device 402 (the top layer scheduling virtual machine device 11 corresponding to Fig. 3) and scheduling virtual machine device 403 (the scheduling virtual machine device 12 corresponding to Fig. 3) is comprised in scheduling virtual machine layer 401.Comprise physical server 404 (corresponding to about the physical server in the explanation of Fig. 3) in Physical layer 407, this physical server 404 comprises virtual machine 406 (virtual machine 14 corresponding to Fig. 3), physical machine watch-dog 408 (the physical machine watch-dog 13 corresponding to Fig. 3), dispatching requirement manager 409 (the dispatching requirement manager 15 corresponding to Fig. 3) and virtual machine manager 410 (virtual machine manager 16 corresponding to Fig. 3).Also have virtual machine layer 405 within the physical layer, this virtual machine layer 405 comprises virtual machine 406 (virtual machine 14 corresponding to Fig. 3).
Although in the diagram of Fig. 3 and Fig. 4, the scheduling virtual machine system of one embodiment of the present of invention comprises two-stage scheduling virtual machine device, but in scheduling virtual machine system of the present invention, namely only can comprise a scheduling virtual machine device (being equivalent to primary virtual machine scheduler), also can comprise the scheduling virtual machine device of more than three grades (comprising three grades).
When scheduling virtual machine system of the present invention only comprises a scheduling virtual machine device, be equivalent to scheduling virtual machine system and comprise scheduling virtual machine device 12a in Fig. 3, physical machine watch-dog 13 is (for different physical machine watch-dogs 13, with 13a, 13b, represent), virtual machine 14, dispatching requirement manager 15 is (for different dispatching requirement managers 15, with 15a, 15b, represent) and virtual machine manager 16 (for different virtual machine managers 16, with 16a, 16b, represent), and do not comprise top layer scheduling virtual machine device 11, scheduling virtual machine device 12m and physical machine watch-dog 13m, virtual machine 14, dispatching requirement manager 15m and virtual machine manager 16m.
Fig. 5 is the process flow diagram of the dispatching method of virtual machine in the scheduling virtual machine system of one embodiment of the present of invention.
Fig. 5 a illustrates the process flow diagram of the dispatching method of virtual machine of an embodiment of underlying virtual machine scheduler 12.
The dispatching method of virtual machine of the underlying virtual machine scheduler 12 in Fig. 5 a had both been applicable to the situation only having a scheduling virtual machine device 12a in scheduling virtual machine system, was also applicable to the situation of two-stage or more level scheduling virtual machine device 12.
Scheduling virtual machine device 12 is responsible for monitoring the message (ST501) from dispatching requirement manager 15, judge whether to comprise in this dispatch request to represent that described dispatching requirement manager 15a has reached the information (ST502) of dispatching and resolving with contiguous dispatching requirement manager 15b, when this dispatch request comprise represent described dispatching requirement manager 15a and contiguous dispatching requirement manager 15b reached dispatch the information of resolving (ST502: "Yes"), dispatch command is sent respectively to described dispatching requirement manager 15a and described contiguous dispatching requirement manager 15b, wherein when having reached scheduling resolution, this dispatch request also comprises the information (ST503) about described contiguous dispatching requirement manager 15b.
As an example, scheduling virtual machine device 12 according to first receiving the order first returned, can send dispatch command.
Wherein, described dispatch command instruction dispatching requirement manager 15 starts to carry out scheduling virtual machine.Described dispatch command only can comprise the information that instruction scheduling demand management device 15 starts to carry out scheduling virtual machine, still can comprise the information when instruction scheduling demand management device 15 starts to carry out scheduling virtual machine.As long as can start to carry out scheduling virtual machine by instruction scheduling demand management device 15, be then not particularly limited in form.
In this example embodiment, resolve because dispatching requirement manager 15a and contiguous dispatching requirement manager 15b has reached to dispatch, as long as therefore scheduling virtual machine device 12 sends dispatch command respectively to dispatching requirement manager 15a and contiguous dispatching requirement manager 15b, then described dispatching requirement manager 15a and described contiguous dispatching requirement manager 15b just can perform the scheduling of virtual machine.
When this dispatch request do not comprise represent described dispatching requirement manager 15a and contiguous dispatching requirement manager 15b reached dispatch the information of resolving (ST502: "No"), can generation error information, notify keeper, also can send error message to the dispatching requirement manager 15 sending dispatch request, also can carry out the operation (ST5001) as Fig. 5 b.
Fig. 5 b illustrates the process flow diagram of the dispatching method of virtual machine of an embodiment of underlying virtual machine scheduler.More particularly, the process flow diagram of the dispatching method of virtual machine of the embodiment that Fig. 5 b is the ST502 of Fig. 5 a when being "No".Especially, it should be noted that, as mentioned above when the ST502 of Fig. 5 a is "No", be not limited to the dispatching method of virtual machine of Fig. 5 b, also can generation error information, notify keeper, also can send error message to the dispatching requirement manager 15 sending dispatch request.
The dispatching method of virtual machine of the underlying virtual machine scheduler 12 in Fig. 5 b had both been applicable to the situation only having a scheduling virtual machine device 12a in scheduling virtual machine system, was also applicable to the situation of two-stage or more level scheduling virtual machine device 12.
ST501-ST503 in ST501-ST503 and Fig. 5 a of Fig. 5 b is identical, therefore does not carry out repeat specification.
When this dispatch request do not comprise represent described dispatching requirement manager 15a and contiguous dispatching requirement manager 15b reached dispatch the information of resolving (ST502: "No"), monitor the dispatch request (ST504) from other dispatching requirement manager, according to the dispatch request of virtual machine of described dispatching requirement manager 15a and the demand of the dispatch request of other dispatching requirement manager 15, judge whether the dispatch request of the virtual machine of described dispatching requirement manager 15a and the dispatch request of other dispatching requirement manager 15 described can mate (ST505), if can mate, then send the information about the other side respectively to described dispatching requirement manager 15a and other dispatching requirement manager 15 described, and send dispatch command (ST506).
In such scheme, when this dispatch request do not comprise represent described dispatching requirement manager 15a and contiguous dispatching requirement manager 15b reached dispatch the information of resolving, the reason monitored from the dispatch request of other dispatching requirement manager instead of the dispatch request of non-adjacent dispatching requirement manager is, can simple flow.In this scenario, when monitoring, without the need to judge the dispatch request of dispatching requirement manager from non-adjacent dispatching requirement manager still from being close to dispatching requirement manager.Certainly, also can be set to and monitor from the dispatch request of non-adjacent dispatching requirement manager, if now judge the dispatch request of dispatching requirement manager from non-adjacent dispatching requirement manager still from being close to dispatching requirement manager.
The reason sent about the information of the other side respectively to described dispatching requirement manager 15a and other dispatching requirement manager 15 described when mating is, makes described dispatching requirement manager 15a and other dispatching requirement manager 15 described can know that next step carries out scheduling virtual machine to whom.The described information about the other side can comprise the address of the other side.As long as can identify the other side, then the described information about the other side is not limited to address information, also can be such as number information etc.
Wherein, whether the dispatch request of the virtual machine of described dispatching requirement manager 15a and the dispatch request of other dispatching requirement manager 15 described can mate expression: whether the request that the virtual machine of the request that the virtual machine of described dispatching requirement manager 15a is moved out and other dispatching requirement manager 15 described is moved into can mate; Or whether the request that the virtual machine of the request that the virtual machine of described dispatching requirement manager 15a is moved into and other dispatching requirement manager 15 described is moved out can mate.
When this dispatch request do not comprise represent described dispatching requirement manager 15a and contiguous dispatching requirement manager 15b reached dispatch the information of resolving, wait for the dispatch request certain hour (such as 10 minutes, can preset) from the coupling of other dispatching requirement manager.
As a specific embodiment, if the dispatch request transmitted by dispatching requirement manager 15a does not comprise and represents that described dispatching requirement manager 15a has reached with contiguous dispatching requirement manager 15b and dispatch the information of resolving, and this dispatch request represents to be needed to move out virtual machine 14 to alleviate cpu load 30%, then scheduling virtual machine device 12a starts to monitor needs to move into from the expression of other dispatching requirement manager 15c the dispatch request of virtual machine.If the expression listened to from other dispatching requirement manager 15c needs to move into, the dispatch request of virtual machine represents that virtual machine 14 of will moving into is to increase cpu load 30% or more, then be judged as mating, then send the information about the other side respectively to described dispatching requirement manager 15a and other dispatching requirement manager 15 described, and send dispatch command.
After sending the information about the other side respectively to described dispatching requirement manager 15a and other dispatching requirement manager 15 described, after also can waiting for described dispatching requirement manager 15a and other dispatching requirement manager 15 described response, send dispatch command.
As another specific embodiment, if the dispatch request transmitted by dispatching requirement manager 15a does not comprise and represents that described dispatching requirement manager 15a has reached with contiguous dispatching requirement manager 15b and dispatch the information of resolving, and this dispatch request represents to be needed to move into virtual machine 14 to increase cpu load 30%, then scheduling virtual machine device 12a starts to monitor needs to move out from the expression of other dispatching requirement manager 15c the dispatch request of virtual machine.If the expression listened to from other dispatching requirement manager 15c needs to move out, the dispatch request of virtual machine represents that virtual machine 14 of will moving out is to reduce cpu load 30% or less, then be judged as mating, then send the information about the other side respectively to described dispatching requirement manager 15a and other dispatching requirement manager 15 described, and send dispatch command.
When the dispatch request of virtual machine and the dispatch request of other dispatching requirement manager 15 described that are judged as described dispatching requirement manager 15a can not be mated (ST505: no), can generation error information, notify keeper, also can send error message to the dispatching requirement manager 15 sending dispatch request, also can carry out the operation (ST5002) as Fig. 5 c.
Fig. 5 c illustrates the process flow diagram of the dispatching method of virtual machine of underlying virtual machine scheduler when to have two-stage or more level scheduling virtual machine device in scheduling virtual machine system.More particularly, the process flow diagram of the dispatching method of virtual machine of the embodiment that Fig. 5 c is the ST502 of Fig. 5 a when being "No".Especially, it should be noted that, as mentioned above when the ST502 of Fig. 5 a is "No", be not limited to the dispatching method of virtual machine of Fig. 5 c, also can generation error information, notify keeper, also can send error message to the dispatching requirement manager 15 sending dispatch request.
ST501-ST503 in ST501-ST503 and Fig. 5 a of Fig. 5 c is identical, therefore does not carry out repeat specification.
When dispatch request do not comprise represent described dispatching requirement manager 15a and contiguous dispatching requirement manager 15b reached dispatch the information of resolving (in ST502, being judged as "No"), judge whether to there is last layer scheduling virtual machine device 12 (ST507), when having last layer scheduling virtual machine device 12, above-mentioned dispatch request is reported last layer scheduling virtual machine device 12, and pass the information about other dispatching requirement manager 15a received from last layer scheduling virtual machine device 12 and dispatch command back described dispatching requirement manager 15 (ST508).
In addition, when also can be judged as "No" in the ST505 of Fig. 5 b, above-mentioned ST507, ST508 is carried out.Specifically, if (such as 10 minutes) do not find the dispatch request of coupling within a certain period of time, the dispatch request of the virtual machine 14 of moving out transmitted from dispatching requirement manager 15a is continued to report upper level scheduling virtual machine device 12, this higher level's scheduling virtual machine device 12 also may be top layer scheduling virtual machine device 11 (situation of two-stage scheduling virtual machine device), and passes the information about other dispatching requirement manager 15a received from last layer scheduling virtual machine device 12 and dispatch command back described dispatching requirement manager 15.
Fig. 6 is the process flow diagram of the dispatching method of virtual machine of the non-underlying virtual machine scheduler of one embodiment of the present of invention.
Fig. 6 a illustrates the process flow diagram of the dispatching method of virtual machine of an embodiment of the non-underlying virtual machine scheduler in scheduling virtual machine system.
Non-underlying virtual machine scheduler 11 (or top layer scheduling virtual machine device 11) monitors the dispatch request (ST601) from two or more lower one deck scheduling virtual machine devices 12, described dispatch request comprises the request of being moved out by virtual machine and the request of being moved into by virtual machine, the demand of the request of moving out according to the virtual machine from lower one deck scheduling virtual machine device 12a and the request that the virtual machine of one deck scheduling virtual machine device 12m is moved under another, judge whether the request that the virtual machine of the request that the virtual machine of described lower one deck scheduling virtual machine device 12a is moved out and another lower one deck scheduling virtual machine device 12m described is moved into can mate (ST602), if the request that the virtual machine of the request that the virtual machine of described lower one deck scheduling virtual machine device 12a is moved out and another lower one deck scheduling virtual machine device 12m described is moved into can mate (ST602: "Yes"), then send the information about the relevant dispatching requirement manager in the other side's dispatch request respectively to described lower one deck scheduling virtual machine device 12a and another lower one deck scheduling virtual machine device 12m described, and send dispatch command (ST603).
An embodiment, the expression received from lower one deck scheduling virtual machine device 12a of non-underlying virtual machine scheduler 11 is moved out the dispatch request of demand of virtual machine 14, this dispatch request represents that dispatching requirement manager 15a needs to move out virtual machine 14 to alleviate cpu load 30%, then non-underlying virtual machine scheduler 11 starts to find the expression transmitted from other next stage scheduling virtual machine device 12 and to move into the dispatch request of demand of virtual machine 14, the dispatch request demand fulfillment of this virtual machine 14 of moving into moves into virtual machine 14 to increase the requirement of cpu load 30% or more, the message notice sending the dispatching requirement manager 15m of this demand found first is subsequently to scheduling virtual machine device 12a, and by the message notice of dispatching requirement manager 15a to scheduling virtual machine device 12m, and send dispatch command.
By the message notice of dispatching requirement manager 15m to scheduling virtual machine device 12a, and by the message notice of dispatching requirement manager 15a to after scheduling virtual machine device 12m, also, after can waiting for the response of scheduling virtual machine device 12a and scheduling virtual machine device 12m, dispatch command is sent.
If (such as 10 minutes) do not find the demand of coupling within a certain period of time, also pairing failure result can be informed to scheduling virtual machine device 12a or/and scheduling virtual machine device 12m.
As another embodiment, the expression received from lower one deck scheduling virtual machine device 12a of non-underlying virtual machine scheduler 11 is moved into the dispatch request of demand of virtual machine 14, this dispatch request represents that dispatching requirement manager 15a needs to move into virtual machine 14 to increase cpu load 30%, then non-underlying virtual machine scheduler 11 starts to find the expression transmitted from other next stage scheduling virtual machine device 12 and to move out the dispatch request of demand of virtual machine 14, the dispatch request demand fulfillment of this virtual machine 14 of moving out moves into virtual machine 14 to reduce cpu load 30% or less requirement, the message notice sending the dispatching requirement manager 15m of this demand found first is subsequently to scheduling virtual machine device 12a, and by the message notice of dispatching requirement manager 15a to scheduling virtual machine device 12m, and send dispatch command.
If the request that the virtual machine of the request that the virtual machine of described lower one deck scheduling virtual machine device 12a is moved out and another lower one deck scheduling virtual machine device 12m described is moved into can not mate (ST602: "No"), can generation error information, notify keeper, also can send error message to the dispatching requirement manager 15 sending dispatch request, also can carry out the operation (ST6001) as Fig. 6 b.
Fig. 6 b illustrates the process flow diagram of the dispatching method of virtual machine of an embodiment of middle layer scheduling virtual machine device in Multi-Level Virtual Machine dispatching system or top layer scheduling virtual machine device.More particularly, the process flow diagram of the dispatching method of virtual machine of the embodiment that Fig. 6 b is the ST602 of Fig. 6 a when being "No".Especially, it should be noted that, as mentioned above when the ST602 of Fig. 6 a is "No", be not limited to the dispatching method of virtual machine of Fig. 6 b, also can generation error information, notify keeper, also can send error message to the dispatching requirement manager 15 sending dispatch request.
ST601-ST603 in ST601-ST603 and Fig. 6 a of Fig. 6 b is identical, therefore does not carry out repeat specification.
If the request that the virtual machine of the request that the virtual machine of described lower one deck scheduling virtual machine device 12a is moved out and another lower one deck scheduling virtual machine device 12m described is moved into can not mate (ST602: "No"), scheduling virtual machine device 11 judges whether to there is last layer scheduling virtual machine device (ST604).
When there is last layer scheduling virtual machine device (ST604: "Yes"), above-mentioned dispatch request is reported last layer scheduling virtual machine device, and pass the information about other dispatching requirement manager 15 received from last layer scheduling virtual machine device and dispatch command back lower one deck scheduling virtual machine device 12a (ST605).
Fig. 7 illustrates the process flow diagram of the dispatching requirement management method of one embodiment of the present of invention.
Below, with dispatching requirement manager 15a for example be described dispatching requirement manager 15 perform four actions:
(1) information on load (ST701) of physical server is obtained from physical machine watch-dog 13a.
(2) information on load of physical server is obtained from physical machine watch-dog 13, according to the load condition (ST702) of information on load (such as about the information of CPU, internal memory) the computational physics server of this physical server, and judge whether the scheduling (ST703) needing virtual machine.Load condition represents that such as cpu load is excessive, cpu load is too small or cpu load normal, or Memory Load is excessive, Memory Load is too small or Memory Load is normal.The information on load of physical server also can be the information of other the Physics server except CPU, internal memory.
(3) when being judged as needing the scheduling of virtual machine 14 (moving out or moving into of virtual machine 14) (ST703: "Yes"), communicate with contiguous dispatching requirement manager 15b, whether inquiry accepts the scheduling (ST704) of virtual machine 14.
When being judged as not needing the scheduling of virtual machine 14 (moving out or moving into of virtual machine 14) (ST703: "No"), such as, step ST701 can be turned back to.
When receive from described contiguous dispatching requirement manager 15b comprise represent that described contiguous dispatching requirement manager 15b accepts the information of the scheduling of virtual machine (ST705: "Yes"), reach to dispatch with described contiguous dispatching requirement manager 15b and resolve, and send to comprise to scheduling virtual machine device 12a and represent described dispatching requirement manager 15a and the dispatch request (ST706) of being close to dispatching requirement manager 15b and having reached the information of dispatching and resolving and the virtual machine about the information of described contiguous dispatching requirement manager 15b.
When receive from described contiguous dispatching requirement manager 15b do not comprise represent that described contiguous dispatching requirement manager 15b accepts the information of the scheduling of virtual machine (ST705: "No"), send not comprise to described scheduling virtual machine device 12a and represent that described dispatching requirement manager 15a has reached the dispatch request (ST707) of dispatching the virtual machine of the information of resolving with contiguous dispatching requirement manager 15b.
Described dispatching requirement manager 15a can send response time test request to other dispatching requirement manager 15, obtain the response time of other dispatching requirement manager 15, according to the response time from being short to long order, inquire whether other dispatching requirement manager 15 accepts the scheduling of virtual machine successively, until a dispatching requirement manager 15 in other dispatching requirement manager 15 accepts the scheduling of virtual machine, described dispatching requirement manager 15a reaches to dispatch with other scheduling virtual machine device 15 of the scheduling accepting virtual machine and resolves, and be considered as described dispatching requirement manager 15a and reached to dispatch with contiguous dispatching requirement manager 15 and resolve, if in official hour, described dispatching requirement manager 15a does not receive the information of the scheduling accepting virtual machine, then be considered as described dispatching requirement manager 15a not reach to dispatch with contiguous dispatching requirement manager 15 and resolve.
In such scheme, described dispatching requirement manager 15a also can certain several (such as 3 of simultaneously query-response shortest time, the delay situation concrete modification according to network can be followed to be 1, 2, 4 ... platform etc.) whether other dispatching requirement manager 15 accept the scheduling of virtual machine, and reach to dispatch with the dispatching requirement manager 15 of the information sending the scheduling accepting virtual machine the earliest and resolve, if in official hour, described dispatching requirement manager 15a does not receive the information of the scheduling accepting virtual machine, then be considered as described dispatching requirement manager 15a not reach to dispatch with contiguous dispatching requirement manager 15 and resolve.
In such scheme, described dispatching requirement manager 15a also can not send response time test request, and whether other dispatching requirement manager 15 that direct access inquiry is connected with same scheduling virtual machine device 12 accepts the scheduling of virtual machine, and reach to dispatch with the dispatching requirement manager 15 of the information sending the scheduling accepting virtual machine the earliest and resolve.
Communicate with contiguous dispatching requirement manager, the dispatching method whether inquiry accepts virtual machine is diversified, is not limited to above-mentioned scheme.
Example one, the cpu load on this physical server is excessive, and dispatching requirement manager 15a sends virtual machine 14 request of moving out to 3 dispatching requirement managers 15b, 15c, 15d that network delay is minimum, request move out 30% cpu load.Dispatching requirement manager 15b is when receiving the inquiry of the scheduling whether accepting virtual machine from described contiguous dispatching requirement manager 15a, according to the physical server information on load obtained from physical machine watch-dog, judge whether the scheduling accepting virtual machine, and judged result is sent to described contiguous dispatching requirement manager 15a.Such as, if dispatching requirement manager 15b is judged as the cpu load requiring that the amount of moving into is more than or equal to 30%, then the scheduling accepting virtual machine is judged as.For dispatching requirement manager 15c, 15d, its action is identical with dispatching requirement manager 15b.If all have sent the judged result of receiving scheduling request from dispatching requirement manager 15b, 15c, 15d, dispatching requirement manager 15a only reaches scheduling with the dispatching requirement manager 15c of the judged result of wherein sending receiving scheduling request the earliest and resolves.
Example two, the cpu load on this physical server is too small, and dispatching requirement manager 15a sends virtual machine 14 request of moving into 3 dispatching requirement managers 15b, 15c, 15d that network delay is minimum, request move into 30% cpu load.Dispatching requirement manager 15b is when receiving the inquiry of the scheduling whether accepting virtual machine from described contiguous dispatching requirement manager 15a, according to the physical server information on load obtained from physical machine watch-dog, judge whether the scheduling accepting virtual machine, and judged result is sent to described contiguous dispatching requirement manager 15a.Such as, if dispatching requirement manager 15b is judged as the cpu load requiring that the amount of moving out is less than or equal to 30%, then the scheduling accepting virtual machine is judged as.For dispatching requirement manager 15c, 15d, its action is identical with dispatching requirement manager 15b.
(3) according to the dispatch command from described scheduling virtual machine device 12 or the information about other dispatching requirement manager 15 and dispatch command, instruction virtual machine manager 16 carries out scheduling virtual machine (ST708).
When receiving the inquiry of the scheduling whether accepting virtual machine from contiguous dispatching requirement manager 15a, according to the physical server information on load obtained from physical machine watch-dog 13, judge whether the scheduling accepting virtual machine, and judged result is sent to described contiguous dispatching requirement manager 15, when being judged as the scheduling accepting virtual machine, reaching to dispatch with described contiguous dispatching requirement manager 15 and resolving.
Fig. 8 illustrates the message time sequence figure in the scheduling virtual machine system of one embodiment of the present of invention.Suppose that this scheduling virtual machine system has two-stage scheduling virtual machine device.Dispatching requirement manager 15 asks physical server information on load (ST801) to physical machine watch-dog 13, wait for that physical server information on load returns rear (ST802), according to this physical server information on load computational physics server load state, and judge whether the scheduling (ST803) needing virtual machine, if be judged as not needing scheduling virtual machine, then repeat the step of ST801-ST803, when being judged as the scheduling needing virtual machine, communicate with contiguous dispatching requirement manager 15, whether inquiry accepts the scheduling of virtual machine, if contiguous dispatching requirement manager 15 accepts dispatch request, then send to comprise to scheduling virtual machine device 12 and represent described dispatching requirement manager and the dispatch request (ST804) of being close to dispatching requirement manager and having reached the information of dispatching and resolving and the virtual machine about the information of described contiguous dispatching requirement manager, scheduling virtual machine device 12 returns dispatch command (ST805), if scheduling virtual machine device 12 needs to report further, scheduling virtual machine device 12 reports top layer scheduling virtual machine device 11 (ST806), etc. to be returned.If top layer scheduling virtual machine device 11 finds scheduling scheme, return dispatch command to scheduling virtual machine device 12 (ST807), scheduling virtual machine device 12 returns dispatch command to dispatching requirement manager 15 (ST808).
In the above example, also physical server information on load (ST801) can initiatively be sent by physical machine watch-dog 13 to dispatching requirement manager 15, dispatching requirement manager 15 according to this physical server information on load computational physics server load state, and judges whether the scheduling (ST803) needing virtual machine.
Fig. 9 illustrates the structured flowchart of underlying virtual machine scheduler of the present invention.
Underlying virtual machine scheduler of the present invention, has: monitoring portion 901, monitors the dispatch request of the virtual machine from dispatching requirement manager; Whether scheduling controlling portion 902, judge to comprise in this dispatch request to represent that described dispatching requirement manager has been reached with contiguous dispatching requirement manager and dispatch the information of resolving; And dispatch command portion 903, when this dispatch request comprise represent described dispatching requirement manager and contiguous dispatching requirement manager reached dispatch the information of resolving, dispatch command is sent respectively to described dispatching requirement manager and described contiguous dispatching requirement manager, wherein when having reached scheduling resolution, this dispatch request also comprises the information about described contiguous dispatching requirement manager.
Figure 10 illustrates the structured flowchart of non-underlying virtual machine scheduler of the present invention.
Non-underlying virtual machine scheduler 12 of the present invention has: monitoring portion 1001, scheduling controlling portion 1002 and dispatch command portion 1003.Monitoring portion 1001, monitor the dispatch request from two or more lower one deck scheduling virtual machine devices, described dispatch request comprises the request of being moved out by virtual machine and the request of being moved into by virtual machine.The demand of the request that the request moved out according to the virtual machine from lower one deck scheduling virtual machine device in scheduling controlling portion 1002 is moved into the virtual machine of another lower one deck scheduling virtual machine device, judges whether the request that the virtual machine of the request that the virtual machine of described lower one deck scheduling virtual machine device is moved out and another lower one deck scheduling virtual machine device described is moved into can mate.If the request that the virtual machine of the request that the virtual machine of the described lower one deck scheduling virtual machine device in dispatch command portion 1003 is moved out and another lower one deck scheduling virtual machine device described is moved into can be mated, then respectively to described lower one deck scheduling virtual machine device and another lower one deck scheduling virtual machine device described send about the relevant dispatching requirement manager in the other side's dispatch request information, and send dispatch command.
Figure 11 illustrates the structured flowchart of dispatching requirement manager of the present invention.
Dispatching requirement manager 15 of the present invention possesses: physical server information on load acquisition unit 1101, scheduling judging part 1102, dispatching communication portion 1103, scheduling resolution portion 1104 and dispatch command portion 1105.Physical server information on load acquisition unit 1101 obtains physical server information on load from physical machine watch-dog.Scheduling judging part 1102 according to this physical server information on load computational physics server load state, and judges whether the scheduling needing virtual machine.Dispatching communication portion 1103 is when being judged as the scheduling needing virtual machine, and communicate with contiguous dispatching requirement manager, whether inquiry accepts the scheduling of virtual machine.Scheduling resolution portion 1104 when receive from described contiguous dispatching requirement manager comprise represent described contiguous dispatching requirement manager accept the information of the scheduling of virtual machine, reach to dispatch with described contiguous dispatching requirement manager and resolve, and send to comprise to scheduling virtual machine device and represent described dispatching requirement manager and the dispatch request of being close to dispatching requirement manager and having reached the information of dispatching and resolving and the virtual machine about the information of described contiguous dispatching requirement manager, described scheduling resolution portion 1104 when receive from described contiguous dispatching requirement manager do not comprise represent described contiguous dispatching requirement manager accept the information of the scheduling of virtual machine, send not comprise to described scheduling virtual machine device and represent that described dispatching requirement manager has reached the dispatch request of dispatching the virtual machine of the information of resolving with contiguous dispatching requirement manager.Dispatch command portion 1105, according to the dispatch command from described scheduling virtual machine device or the information about other dispatching requirement manager and dispatch command, instruction virtual machine manager carries out scheduling virtual machine.
The present invention also provides a kind of scheduling virtual machine system, comprising: above-mentioned underlying virtual machine scheduler, and multiple physical server, each physical server comprises one or more virtual machine, physical machine watch-dog, virtual machine manager, and above-mentioned dispatching requirement manager, wherein, described virtual machine is distributed on described physical server, service is provided to user, a virtual machine can be run in each physical server or run multiple virtual machine simultaneously, wherein, every platform physical server runs physical machine watch-dog, the physical server information on load of the real-time monitors physical server of physical machine watch-dog, and send physical server information on load to dispatching requirement manager.
The present invention also provides another kind of scheduling virtual machine system, it is characterized in that, comprising: multiple above-mentioned underlying virtual machine scheduler, one or more above-mentioned non-underlying virtual machine scheduler, multiple physical server, each physical server comprises one or more virtual machine, physical machine watch-dog, virtual machine manager, and above-mentioned dispatching requirement manager, wherein, described virtual machine is distributed on described physical server, service is provided to user, a virtual machine can be run in each physical server or run multiple virtual machine simultaneously, wherein, every platform physical server runs physical machine watch-dog, the physical server information on load of the real-time monitors physical server of physical machine watch-dog, and send physical server information on load to dispatching requirement manager.
In modern city, the application of video monitoring system is more and more extensive, such as city security protection, traffic monitoring, building monitoring etc.And the scale of video monitoring is also more and more huger, the number of simple camera all can have surge every year.The expansion of city size simultaneously, increasing of monitored object, the requirement for video monitoring system is also more and more higher, and simple camera coordinates the pattern of video recorder cannot meet current demand.Modern video supervisory system requires that monitoring scale is increasing, sometimes even in units of city.For traditional video monitoring system and data center, be more and more difficult to the further expansion tackling monitoring scale.
In view of above situation, lot of challenges that extensive video monitoring system brings is all well and good selection to use cloud platform to go to solve, but simultaneously, cloud platform itself brings new problem to us again, that is exactly the video monitoring data of the dynamic change in the face of magnanimity, how dynamic scheduling system resource, raises the efficiency to greatest extent.Intrinsic typical data center is difficult to accomplish dynamic dispatching.The server thermophoresis technology based on server virtualization technology popular at present can realize dynamic dispatching, but thermophoresis technology needs accurately to hold the state of system resource and real-time responding system dispatching requirement.Further, in the face of extensive video monitoring system, whole state of whole system be unifiedly calculated and make corresponding in real time.Also provide a kind of video monitoring system according to the present invention, it is characterized in that, there is above-mentioned underlying virtual machine scheduler, non-underlying virtual machine scheduler or above-mentioned scheduling virtual machine system.
By adopting above-mentioned underlying virtual machine scheduler of the present invention, non-underlying virtual machine scheduler or above-mentioned scheduling virtual machine system in video monitoring system, can schedules system resources to improve resource utilization and balanced load.
Process flow diagram in accompanying drawing and block diagram show system according to multiple embodiment of the present invention, the architectural framework in the cards of method and computer program product, function and operation.In this, each square frame in process flow diagram or block diagram can represent a part for module, program segment or a code, and a part for described module, program segment or code comprises one or more executable instruction for realizing the logic function specified.Also it should be noted that at some as in the realization of replacing, the function marked in square frame also can be different from occurring in sequence of marking in accompanying drawing.Such as, in fact two continuous print square frames can perform substantially concurrently, and they also can perform by contrary order sometimes, and this determines according to involved function.Also it should be noted that, the combination of the square frame in each square frame in block diagram and/or process flow diagram and block diagram and/or process flow diagram, can realize by the special hardware based system of the function put rules into practice or operation, or can realize with the combination of specialized hardware and computer instruction.
Be described above various embodiments of the present invention, above-mentioned explanation is exemplary, and non-exclusive, and be also not limited to disclosed each embodiment.When not departing from the scope and spirit of illustrated each embodiment, many modifications and changes are all apparent for those skilled in the art.The selection of term used herein, is intended to explain best the principle of each embodiment, practical application or to the technological improvement in market, or makes other those of ordinary skill of the art can understand each embodiment disclosed herein.

Claims (23)

1. a dispatching method of virtual machine, is characterized in that, comprising:
Monitor the dispatch request from the virtual machine of dispatching requirement manager,
Judge whether to comprise in this dispatch request to represent that described dispatching requirement manager has been reached with contiguous dispatching requirement manager and dispatch the information of resolving,
When this dispatch request comprise represent described dispatching requirement manager and contiguous dispatching requirement manager reached dispatch the information of resolving, dispatch command is sent respectively to described dispatching requirement manager and described contiguous dispatching requirement manager, wherein when having reached scheduling resolution, this dispatch request also comprises the information about described contiguous dispatching requirement manager.
2. dispatching method of virtual machine as claimed in claim 1, is characterized in that,
When this dispatch request do not comprise represent described dispatching requirement manager and described contiguous dispatching requirement manager reached dispatch the information of resolving, monitor the dispatch request of the virtual machine from other dispatching requirement manager,
According to the demand of the dispatch request of the virtual machine of described dispatching requirement manager with the dispatch request of the virtual machine of other dispatching requirement manager, judge whether the dispatch request of the dispatch request of the virtual machine of described dispatching requirement manager and the virtual machine of other dispatching requirement manager described can mate
If can mate, then send the information about the other side respectively to described dispatching requirement manager and other dispatching requirement manager described, and send dispatch command,
Wherein, whether the dispatch request of the virtual machine of described dispatching requirement manager and the dispatch request of other dispatching requirement manager described can mate expression: whether the request that the virtual machine of the request that the virtual machine of described dispatching requirement manager is moved out and other dispatching requirement manager described is moved into can mate; Or whether the request that the virtual machine of the request that the virtual machine of described dispatching requirement manager is moved into and other dispatching requirement manager described is moved out can mate.
3. dispatching method of virtual machine as claimed in claim 1, is characterized in that,
When this dispatch request do not comprise represent described dispatching requirement manager and contiguous dispatching requirement manager reached dispatch the information of resolving, judge whether to there is last layer scheduling virtual machine device,
When there is last layer scheduling virtual machine device, above-mentioned dispatch request is reported last layer scheduling virtual machine device, and pass the information about other dispatching requirement manager received from last layer scheduling virtual machine device and dispatch command back described dispatching requirement manager.
4. dispatching method of virtual machine as claimed in claim 2, is characterized in that,
When this dispatch request do not comprise represent described dispatching requirement manager be close to dispatching requirement manager reached the information of dispatching and resolving and the dispatch request of described dispatching requirement manager fails to mate with the dispatch request of other dispatching requirement manager, judge whether to there is last layer scheduling virtual machine device
When there is last layer scheduling virtual machine device, above-mentioned dispatch request is reported last layer scheduling virtual machine device, and pass the information about other dispatching requirement manager received from last layer scheduling virtual machine device and dispatch command back described dispatching requirement manager.
5. dispatching method of virtual machine as claimed in claim 1, is characterized in that,
Described dispatch request comprises the request of being moved out by virtual machine or the request of being moved into by virtual machine.
6. a dispatching method of virtual machine, is characterized in that,
Monitor the dispatch request from two or more lower one deck scheduling virtual machine devices, described dispatch request comprises the request of being moved out by virtual machine or the request of being moved into by virtual machine,
The demand of the request of moving out according to the virtual machine from lower one deck scheduling virtual machine device and the request that the virtual machine of one deck scheduling virtual machine device is moved under another, judge whether the request that the virtual machine of the request that the virtual machine of described lower one deck scheduling virtual machine device is moved out and another lower one deck scheduling virtual machine device described is moved into can mate
If the request that the virtual machine of the request that the virtual machine of described lower one deck scheduling virtual machine device is moved out and another lower one deck scheduling virtual machine device described is moved into can be mated, then send the information about the relevant dispatching requirement manager in the dispatch request of the other side respectively to described lower one deck scheduling virtual machine device and another lower one deck scheduling virtual machine device described, and send dispatch command.
7. dispatching method of virtual machine as claimed in claim 6, is characterized in that,
If the request that the virtual machine of the request that the virtual machine of described lower one deck scheduling virtual machine device is moved out and another lower one deck scheduling virtual machine device described is moved into can not be mated, judge whether to there is last layer scheduling virtual machine device,
When there is last layer scheduling virtual machine device, above-mentioned dispatch request is reported last layer scheduling virtual machine device, and pass the information about other dispatching requirement manager received from last layer scheduling virtual machine device and dispatch command back lower one deck scheduling virtual machine device.
8. a dispatching requirement management method, is characterized in that,
Physical server information on load is obtained from physical machine watch-dog,
According to this physical server information on load computational physics server load state, and judge whether the scheduling needing virtual machine,
When being judged as the scheduling needing virtual machine, communicate with contiguous dispatching requirement manager, whether inquiry accepts the scheduling of virtual machine,
When receive from described contiguous dispatching requirement manager comprise represent described contiguous dispatching requirement manager accept the information of the scheduling of virtual machine, reach to dispatch with described contiguous dispatching requirement manager and resolve, and send to comprise to scheduling virtual machine device and represent described dispatching requirement manager and the dispatch request of being close to dispatching requirement manager and having reached the information of dispatching and resolving and the virtual machine about the information of described contiguous dispatching requirement manager
When receive from described contiguous dispatching requirement manager do not comprise represent described contiguous dispatching requirement manager accept the information of the scheduling of virtual machine, send not comprise to described scheduling virtual machine device and represent that described dispatching requirement manager has reached the dispatch request of dispatching the virtual machine of the information of resolving with contiguous dispatching requirement manager
According to the dispatch command from described scheduling virtual machine device or the information about other dispatching requirement manager and dispatch command, instruction virtual machine manager carries out scheduling virtual machine.
9. dispatching requirement management method as claimed in claim 8, is characterized in that,
The scheduling of virtual machine is moving into of moving out of virtual machine or virtual machine.
10. dispatching requirement management method as claimed in claim 8, is characterized in that,
When receiving the inquiry of the scheduling whether accepting virtual machine from described contiguous dispatching requirement manager, according to the physical server information on load obtained from physical machine watch-dog, judge whether the scheduling accepting virtual machine, and judged result is sent to described contiguous dispatching requirement manager
When being judged as the scheduling accepting virtual machine, reaching to dispatch with described contiguous dispatching requirement manager and resolving.
11. 1 kinds of scheduling virtual machine devices, is characterized in that having:
Monitoring portion, monitors the dispatch request of the virtual machine from dispatching requirement manager;
Whether scheduling controlling portion, judge to comprise in this dispatch request to represent that described dispatching requirement manager has been reached with contiguous dispatching requirement manager and dispatch the information of resolving; And
Dispatch command portion, when this dispatch request comprise represent described dispatching requirement manager and contiguous dispatching requirement manager reached dispatch the information of resolving, dispatch command is sent respectively to described dispatching requirement manager and described contiguous dispatching requirement manager, wherein when having reached scheduling resolution, this dispatch request also comprises the information about described contiguous dispatching requirement manager.
12. scheduling virtual machine devices as claimed in claim 11, is characterized in that,
When this dispatch request do not comprise represent described dispatching requirement manager and described contiguous dispatching requirement manager reached dispatch the information of resolving, the monitoring of described monitoring portion from the dispatch request of the virtual machine of other dispatching requirement manager,
Described scheduling controlling portion is according to the demand of the dispatch request of the virtual machine of described dispatching requirement manager with the dispatch request of the virtual machine of other dispatching requirement manager, judge whether the dispatch request of the dispatch request of the virtual machine of described dispatching requirement manager and the virtual machine of other dispatching requirement manager described can mate
If can mate, described scheduling controlling portion sends the information about the other side respectively to described dispatching requirement manager and other dispatching requirement manager described, and sends dispatch command,
Wherein, whether the dispatch request of the virtual machine of described dispatching requirement manager and the dispatch request of other dispatching requirement manager described can mate expression: whether the request that the virtual machine of the request that the virtual machine of described dispatching requirement manager is moved out and other dispatching requirement manager described is moved into can mate; Or whether the request that the virtual machine of the request that the virtual machine of described dispatching requirement manager is moved into and other dispatching requirement manager described is moved out can mate.
13. scheduling virtual machine devices as claimed in claim 11, is characterized in that also having:
First judging part, when this dispatch request do not comprise represent described dispatching requirement manager and contiguous dispatching requirement manager reached dispatch the information of resolving, judge whether to there is last layer scheduling virtual machine device; And
Dispatch request portion, when there is last layer scheduling virtual machine device, above-mentioned dispatch request is reported last layer scheduling virtual machine device, and pass the information about other dispatching requirement manager received from last layer scheduling virtual machine device and dispatch command back described dispatching requirement manager.
14. scheduling virtual machine devices as claimed in claim 12, is characterized in that also having:
Second judging part, when this dispatch request do not comprise represent described dispatching requirement manager be close to dispatching requirement manager reached the information of dispatching and resolving and the dispatch request of described dispatching requirement manager fails to mate with the dispatch request of other dispatching requirement manager, judge whether to there is last layer scheduling virtual machine device
Dispatch request portion, when there is last layer scheduling virtual machine device, above-mentioned dispatch request is reported last layer scheduling virtual machine device, and pass the information about other dispatching requirement manager received from last layer scheduling virtual machine device and dispatch command back described dispatching requirement manager.
15. scheduling virtual machine devices as claimed in claim 11, is characterized in that,
Described dispatch request is the request of being moved out by virtual machine or the request of being moved into by virtual machine.
16. 1 kinds of scheduling virtual machine devices, is characterized in that having:
Monitoring portion, monitor the dispatch request from two or more lower one deck scheduling virtual machine devices, described dispatch request comprises the request of being moved out by virtual machine or the request of being moved into by virtual machine; And
Scheduling controlling portion, the demand of the request of moving out according to the virtual machine from lower one deck scheduling virtual machine device and the request that the virtual machine of one deck scheduling virtual machine device is moved under another, judge whether the request that the virtual machine of the request that the virtual machine of described lower one deck scheduling virtual machine device is moved out and another lower one deck scheduling virtual machine device described is moved into can mate
Dispatch command portion, if the request that the virtual machine of the request that the virtual machine of described lower one deck scheduling virtual machine device is moved out and another lower one deck scheduling virtual machine device described is moved into can be mated, then respectively to described lower one deck scheduling virtual machine device and another lower one deck scheduling virtual machine device described send about the relevant dispatching requirement manager in the other side's dispatch request information, and send dispatch command.
17. scheduling virtual machine devices as claimed in claim 16, is characterized in that also possessing:
3rd judging part, if the request that the virtual machine of the request of moving out of the virtual machine of described lower one deck scheduling virtual machine device and another lower one deck scheduling virtual machine device described is moved into can not be mated, judges whether to there is last layer scheduling virtual machine device,
Dispatch request portion, when there is last layer scheduling virtual machine device, above-mentioned dispatch request is reported last layer scheduling virtual machine device, and pass the information about other dispatching requirement manager received from last layer scheduling virtual machine device and dispatch command back lower one deck scheduling virtual machine device.
18. 1 kinds of dispatching requirement managers, is characterized in that possessing:
Physical server information on load acquisition unit, obtains physical server information on load from physical machine watch-dog;
Scheduling judging part, according to this physical server information on load computational physics server load state, and judges whether the scheduling needing virtual machine;
Dispatching communication portion, when being judged as the scheduling needing virtual machine, communicate with contiguous dispatching requirement manager, whether inquiry accepts the scheduling of virtual machine;
Scheduling resolution portion, when receive from described contiguous dispatching requirement manager comprise represent described contiguous dispatching requirement manager accept the information of the scheduling of virtual machine, reach to dispatch with described contiguous dispatching requirement manager and resolve, and send to comprise to scheduling virtual machine device and represent described dispatching requirement manager and the dispatch request of being close to dispatching requirement manager and having reached the information of dispatching and resolving and the virtual machine about the information of described contiguous dispatching requirement manager
Described scheduling resolution portion when receive from described contiguous dispatching requirement manager do not comprise represent that described contiguous dispatching requirement manager accepts the information of the scheduling of virtual machine, send not comprise to described scheduling virtual machine device and represent that described dispatching requirement manager has reached the dispatch request of dispatching the virtual machine of the information of resolving with contiguous dispatching requirement manager; And
Dispatch command portion, according to the dispatch command from described scheduling virtual machine device or the information about other dispatching requirement manager and dispatch command, instruction virtual machine manager carries out scheduling virtual machine.
19. dispatching requirement managers as claimed in claim 18, is characterized in that,
The scheduling of virtual machine is moving into of moving out of virtual machine or virtual machine.
20. dispatching requirement managers as claimed in claim 18, is characterized in that,
When receiving the inquiry of the scheduling whether accepting virtual machine from described contiguous dispatching requirement manager, according to the physical server information on load obtained from physical machine watch-dog, judge whether the scheduling accepting virtual machine, and judged result is sent to described contiguous dispatching requirement manager
When dispatching judging part and being judged as the scheduling accepting virtual machine, described scheduling resolution portion and described contiguous dispatching requirement manager are reached to dispatch and are resolved.
21. 1 kinds of scheduling virtual machine systems, is characterized in that, comprising:
Scheduling virtual machine device described in claim 11 or 12; And
Multiple physical server, each physical server comprises the dispatching requirement manager described in any one in one or more virtual machine, physical machine watch-dog, virtual machine manager and claim 18-20,
Wherein, described virtual machine is distributed on described physical server, provides service to user, can run a virtual machine or run multiple virtual machine simultaneously in each physical server,
Wherein, every platform physical server runs physical machine watch-dog, the physical server information on load of the real-time monitors physical server of physical machine watch-dog, and send physical server information on load to dispatching requirement manager.
22. 1 kinds of scheduling virtual machine systems, is characterized in that, comprising:
Scheduling virtual machine device described in multiple claim 13 or 14 is as underlying virtual machine scheduler;
Scheduling virtual machine device described in one or more claim 16 or 17 is as non-underlying virtual machine scheduler;
Multiple physical server, each physical server comprises the dispatching requirement manager described in any one in one or more virtual machine, physical machine watch-dog, virtual machine manager and claim 18-20,
Wherein, described virtual machine is distributed on described physical server, provides service to user, can run a virtual machine or run multiple virtual machine simultaneously in each physical server,
Wherein, every platform physical server runs physical machine watch-dog, the physical server information on load of the real-time monitors physical server of physical machine watch-dog, and send physical server information on load to dispatching requirement manager.
23. 1 kinds of video monitoring systems, it is characterized in that there is the scheduling virtual machine device described in any one in the scheduling virtual machine device described in any one in claim 11-15, claim 16-20 or claim 21 or scheduling virtual machine system according to claim 22.
CN201410122663.0A 2014-03-28 2014-03-28 Scheduling virtual machine device, dispatching requirement manager and method and scheduling virtual machine system Expired - Fee Related CN104951351B (en)

Priority Applications (1)

Application Number Priority Date Filing Date Title
CN201410122663.0A CN104951351B (en) 2014-03-28 2014-03-28 Scheduling virtual machine device, dispatching requirement manager and method and scheduling virtual machine system

Applications Claiming Priority (1)

Application Number Priority Date Filing Date Title
CN201410122663.0A CN104951351B (en) 2014-03-28 2014-03-28 Scheduling virtual machine device, dispatching requirement manager and method and scheduling virtual machine system

Publications (2)

Publication Number Publication Date
CN104951351A true CN104951351A (en) 2015-09-30
CN104951351B CN104951351B (en) 2018-12-18

Family

ID=54166020

Family Applications (1)

Application Number Title Priority Date Filing Date
CN201410122663.0A Expired - Fee Related CN104951351B (en) 2014-03-28 2014-03-28 Scheduling virtual machine device, dispatching requirement manager and method and scheduling virtual machine system

Country Status (1)

Country Link
CN (1) CN104951351B (en)

Cited By (2)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
CN106919445A (en) * 2015-12-28 2017-07-04 华为技术有限公司 A kind of method and apparatus of the container of Parallel Scheduling in the cluster
CN111104203A (en) * 2019-12-13 2020-05-05 广东省华南技术转移中心有限公司 Virtual machine decentralized scheduling method and device, electronic device and storage medium

Citations (9)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
CN101694633A (en) * 2009-09-30 2010-04-14 曙光信息产业(北京)有限公司 Equipment, method and system for dispatching of computer operation
CN101719081A (en) * 2009-12-01 2010-06-02 北京大学 Method for scheduling virtual machines
US20110088030A1 (en) * 2005-02-25 2011-04-14 Vmware, Inc. Efficient Virtualization of Input/Output Completions for a Virtual Device
CN102158535A (en) * 2011-02-10 2011-08-17 浪潮(北京)电子信息产业有限公司 Cloud computing operating system
US20120096240A1 (en) * 2010-10-15 2012-04-19 International Business Machines Corporation Application Performance with Support for Re-Initiating Unconfirmed Software-Initiated Threads in Hardware
CN102479099A (en) * 2010-11-22 2012-05-30 中兴通讯股份有限公司 Virtual machine management system and use method thereof
CN102609309A (en) * 2012-01-19 2012-07-25 中兴通讯股份有限公司 Strategy scheduling system for cloud computing and strategy scheduling method for cloud computing
CN102981910A (en) * 2012-11-02 2013-03-20 曙光云计算技术有限公司 Realization method and realization device for virtual machine scheduling
CN104199718A (en) * 2014-08-22 2014-12-10 上海交通大学 Dispatching method of virtual processor based on NUMA high-performance network cache resource affinity

Patent Citations (9)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
US20110088030A1 (en) * 2005-02-25 2011-04-14 Vmware, Inc. Efficient Virtualization of Input/Output Completions for a Virtual Device
CN101694633A (en) * 2009-09-30 2010-04-14 曙光信息产业(北京)有限公司 Equipment, method and system for dispatching of computer operation
CN101719081A (en) * 2009-12-01 2010-06-02 北京大学 Method for scheduling virtual machines
US20120096240A1 (en) * 2010-10-15 2012-04-19 International Business Machines Corporation Application Performance with Support for Re-Initiating Unconfirmed Software-Initiated Threads in Hardware
CN102479099A (en) * 2010-11-22 2012-05-30 中兴通讯股份有限公司 Virtual machine management system and use method thereof
CN102158535A (en) * 2011-02-10 2011-08-17 浪潮(北京)电子信息产业有限公司 Cloud computing operating system
CN102609309A (en) * 2012-01-19 2012-07-25 中兴通讯股份有限公司 Strategy scheduling system for cloud computing and strategy scheduling method for cloud computing
CN102981910A (en) * 2012-11-02 2013-03-20 曙光云计算技术有限公司 Realization method and realization device for virtual machine scheduling
CN104199718A (en) * 2014-08-22 2014-12-10 上海交通大学 Dispatching method of virtual processor based on NUMA high-performance network cache resource affinity

Cited By (4)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
CN106919445A (en) * 2015-12-28 2017-07-04 华为技术有限公司 A kind of method and apparatus of the container of Parallel Scheduling in the cluster
CN106919445B (en) * 2015-12-28 2021-08-20 华为技术有限公司 Method and device for scheduling containers in cluster in parallel
CN111104203A (en) * 2019-12-13 2020-05-05 广东省华南技术转移中心有限公司 Virtual machine decentralized scheduling method and device, electronic device and storage medium
CN111104203B (en) * 2019-12-13 2023-04-28 广东省华南技术转移中心有限公司 Virtual machine decentralized scheduling method and device, electronic equipment and storage medium

Also Published As

Publication number Publication date
CN104951351B (en) 2018-12-18

Similar Documents

Publication Publication Date Title
CN110752947B (en) K8s cluster deployment method and device, and deployment platform
US9223632B2 (en) Cross-cloud management and troubleshooting
CN106254471A (en) Resource United Dispatching method and system under a kind of isomery cloud environment
CN102571452A (en) Multi-node management method and system
CN104219316A (en) Method and device for processing call request in distributed system
AU2012259086A1 (en) Cross-cloud management and troubleshooting
CN105847088A (en) Virtual machine performance testing system based on cloud service
CN105553682B (en) Event notification method and the system notified for event
CN103188098B (en) A kind of disaster tolerance switching method, system and device
CN106844315A (en) Based on the method and device that tax return is automatically generated under an operation interface
CN103577165A (en) Form generation method and form generator
CN107005452A (en) A kind of network function virtual resources processing method and virtual network function manager
CN104935639A (en) Intelligent terminal, and applied distributed exploitation system and method
CN104199683A (en) ISO mirror image document loading method and device
CN103685535A (en) Interface design of cloud database system for large-scale distributed heterogeneous data treatment
CN110109371A (en) Intelligent household equipment management system and device management method
CN105404530B (en) It is a kind of to realize easy deployment and the system and method using private clound
US9132550B2 (en) Apparatus and method for managing robot components
CN104951351A (en) Virtual machine dispatcher, dispatching demand manager and method and virtual machine dispatching system
CN106547645A (en) Automatically the method and server system of image shelves are repaired
CN104270274A (en) Fault processing method, device and system based on form
CN108287619A (en) A kind of method, system and the readable storage medium storing program for executing of control server kvm system
CN103164776B (en) A kind of medicine traceability system and medicine retroactive method thereof and flow monitoring method
CN101465730A (en) Method and system for automatically uploading and downloading data
CN104063451A (en) Database connection management method and system

Legal Events

Date Code Title Description
C06 Publication
PB01 Publication
C10 Entry into substantive examination
SE01 Entry into force of request for substantive examination
GR01 Patent grant
GR01 Patent grant
CF01 Termination of patent right due to non-payment of annual fee
CF01 Termination of patent right due to non-payment of annual fee

Granted publication date: 20181218

Termination date: 20210328