CN105868004A - Cloud computing based business system scheduling method and apparatus - Google Patents
Cloud computing based business system scheduling method and apparatus Download PDFInfo
- Publication number
- CN105868004A CN105868004A CN201510035570.9A CN201510035570A CN105868004A CN 105868004 A CN105868004 A CN 105868004A CN 201510035570 A CN201510035570 A CN 201510035570A CN 105868004 A CN105868004 A CN 105868004A
- Authority
- CN
- China
- Prior art keywords
- virtual machine
- mapping relations
- operation system
- scheduling
- physical server
- 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
Links
Classifications
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06F—ELECTRIC DIGITAL DATA PROCESSING
- G06F9/00—Arrangements for program control, e.g. control units
- G06F9/06—Arrangements for program control, e.g. control units using stored programs, i.e. using an internal store of processing equipment to receive or retain programs
- G06F9/46—Multiprogramming arrangements
- G06F9/48—Program initiating; Program switching, e.g. by interrupt
Landscapes
- Engineering & Computer Science (AREA)
- Software Systems (AREA)
- Theoretical Computer Science (AREA)
- Physics & Mathematics (AREA)
- General Engineering & Computer Science (AREA)
- General Physics & Mathematics (AREA)
- Power Sources (AREA)
Abstract
The invention provides a cloud computing based business system scheduling method and apparatus. By analyzing a historical load, a future business situation is predicted, so that virtual machine scheduling is carried out scientifically and efficiently. According to the method and apparatus, a scheduling algorithm of multiple business cloud systems can be simplified, the business cloud systems can be scheduled more directly and efficiently, the complexity of the scheduling algorithm is lowered, the problem of scheduling jitter caused by a conventional scheduling algorithm is reduced, and the problem of business system performance reduction caused by unnecessary virtual machine migration during business peak is avoided.
Description
Technical field
The present invention relates to cloud computing resource scheduling method, especially relate to a kind of operation system based on cloud computing
Dispatching method and dispatching device.
Background technology
The industry such as the increment service system of telecom operators, such as sms center, MMS center, WAP gateway
Business system, is the important sources of operator's income, and therefore the construction of increment service system enjoys operator always
Pay attention to.In tradition construction scheme, every suit increment service system, all use proprietary server resource.
Such as, the server that MMS center uses, MMS center exclusively enjoy.The server that sms center is used,
Exclusively enjoyed by sms center.The construction mode of this free-standing chimney formula, an obvious shortcoming is physics clothes
Business device serious waste of resources.
Along with virtualization, the development of cloud computing technology, this independent, funnel-shaped building of increment service system
If pattern, gradually replaced by cloud computing mode.Under cloud computing mode, first by resource pool, i.e. pass through
Intel Virtualization Technology, is virtualized into multiple virtual machine by physical server, and operation system carries on a virtual machine.
So by after physical server resource pool, with the addition of one layer of virtualization layer, decoupled operation system and physics
Coupled relation between server so that one group of physical server of multiple operation system multiplexings is possibly realized.This
Plant the multi-service cloud system using cloud computing technology to build, not only increase the utilization rate of resource, be also energy-conservation
Consumption reduction, reduces O&M cost and provides possibility.
The operation system of telecom operators, also exists obvious tidal regime, such as daytime, due to people's
The vigorous load making operation system of communication requirement presents peak value, and night reduces due to the communication requirement of people,
The load of operation system will decline very significantly.So, use the multi-service cloud system of cloud computing technology,
Often by virtualized migration characteristic, carry out intelligentized scheduling virtual machine, to reach to reduce power consumption,
Reduce the effect of cost.A kind of typical scenario of prior art is, by day, by the void of all carrying business
Plan machine is evenly distributed in each station server of resource pool, and system is high with optimal performance state carrying business
The arrival at peak.When low-load at night, by the virtual machine of carrying business, compression moves to minimum server
Middle operation, will not run the free physical server of virtual machine simultaneously, carry out lower electricity or sleep operation, with
Reduce the consumption to electric energy of these servers, thus reduce the purpose of cost.
But, the problem of existing dispatching algorithm is, lacks the overall situation and considers, only considers the money at scheduling instance
Source loading condition, does not accounts for the diversity between the load of operation system different business, can bring in system
Between again there is scheduling in running thus produce scheduling jitter problem.
Summary of the invention
The embodiment of the present invention to solve the technical problem that the tune being to provide a kind of operation system based on cloud computing
Degree method and dispatching device, in order to reduce the complexity of dispatching algorithm, and reduce what conventional scheduling algorithms caused
Scheduling jitter problem.
For solving above-mentioned technical problem, the scheduling of the operation system based on cloud computing that the embodiment of the present invention provides
Method, including:
When described operation system switches to a scene mode, obtain institute under the described scene mode pre-saved
State the mapping relations table between the virtual machine of operation system and physical server, record in described mapping relations table
There is described operation system the last time according to predetermined scheduling virtual machine algorithmic dispatching under described scene mode
The virtual machine obtained and the mapping relations of physical server, described scene mode is corresponding to described operation system
In a periodic time period;
According to the mapping relations in described mapping relations table, by the scheduling virtual machine of described operation system to corresponding
Physical server.
Wherein, in said method,
When described operation system switches to described scene mode, if not yet preserving under described scene mode
Described mapping relations table, then periodically gather the resource occupation of each physical server in described operation system
Situation, and according to predetermined scheduling virtual machine algorithm, perform the migration scheduling of virtual machine;
And, according to the mapping relations between the virtual machine after migration scheduling and physical server, record and protect
Deposit the described mapping relations table under described scene mode.
Wherein, in said method, described according to the mapping relations in described mapping relations table, by described business
The physical server that the scheduling virtual machine of system is the most corresponding, including:
The physical server of record in described mapping relations table is waken up up, and according in described mapping relations table
Mapping relations, by the scheduling virtual machine of record in described mapping relations table to corresponding physical server;And,
By in described operation system except in described mapping relations table record physical server in addition to residue
Reason server, carries out lower electricity or dormancy processes.
Wherein, in said method, described according to the mapping relations in described mapping relations table, by described industry
After the step of the physical server that the scheduling virtual machine of business system is the most corresponding, described method also includes:
Periodically gather the occupation condition of each physical server in described operation system, and according to institute
State occupation condition, according to predetermined scheduling virtual machine algorithm, perform the migration scheduling of virtual machine;
And, after the migration scheduling having performed described virtual machine, according to current virtual of described operation system
Mapping relations between machine and physical server, update described mapping relations table.
Wherein, in said method,
The working time section of described operation system include the busy period that is set in advance and that periodically present and
The idle period;
Described according to described occupation condition, according to predetermined scheduling virtual machine algorithm, perform virtual machine
Migration scheduling, including:
When the busy scene that described scene mode is during corresponding to the described busy period, calculate according to load balancing
Method, by the partial virtual machine on the resources occupation rate physical server beyond predetermined threshold, migrates to described industry
On the physical server that in business system, resources occupation rate is minimum;
When the idle scene that described scene mode is during corresponding to the described idle period, according to knapsack algorithm,
By on the virtual machine (vm) migration of described operation system to minimal number of physical server.
The embodiment of the present invention additionally provides the dispatching device of a kind of operation system based on cloud computing, including:
Acquiring unit, for when described operation system switches to a scene mode, obtains the institute pre-saved
State the mapping relations table between virtual machine and the physical server of described operation system under scene mode, described in reflect
Penetrate in relation table record have described operation system the last under described scene mode according to predetermined virtual
Machine dispatching algorithm dispatches the mapping relations of virtual machine and the physical server obtained, and described scene mode corresponds to
Described operation system in a periodic time period;
First scheduling unit, for according to the mapping relations in described mapping relations table, by described operation system
Scheduling virtual machine to corresponding physical server.
Wherein, in above-mentioned dispatching device, also include:
Second scheduling unit, for when described operation system switches to described scene mode, if not yet preserving
There is the described mapping relations table under described scene mode, then periodically gather each thing in described operation system
The occupation condition of reason server, and according to predetermined scheduling virtual machine algorithm, perform the migration of virtual machine
Scheduling;
First updating block, for closing according to the mapping between the virtual machine after migration scheduling and physical server
System, records and preserves the described mapping relations table under described scene mode.
Wherein, in above-mentioned dispatching device, described first scheduling unit includes:
Migration units, for waking up up the physical server of record in described mapping relations table, and according to described
Mapping relations in mapping relations table, by the scheduling virtual machine of record in described mapping relations table to corresponding thing
Reason server;And,
Power saving unit, for removing the physical server of record in described mapping relations table in described operation system
Outside residue physical server, carry out lower electricity or dormancy and process.
Wherein, in above-mentioned dispatching device, also include:
3rd scheduling unit, for periodically gathering the resource of each physical server in described operation system
Take situation, and according to described occupation condition, according to predetermined scheduling virtual machine algorithm, perform virtual
The migration scheduling of machine;And, after the migration scheduling having performed described virtual machine, according to described operation system
Mapping relations between current virtual machine and physical server, update described mapping relations table.
Wherein, in above-mentioned dispatching device, the working time section of described operation system include set in advance and
The busy period periodically presented and idle period;
Described second scheduling unit or the 3rd scheduling unit, be further used for described scene mode for corresponding to
During busy scene during the described busy period, according to load-balancing algorithm, by resources occupation rate beyond predetermined door
Partial virtual machine on the physical server of limit, migrates to the thing that in described operation system, resources occupation rate is minimum
On reason server;And, when the idle scene that described scene mode is during corresponding to the described idle period,
According to knapsack algorithm, by the virtual machine (vm) migration of described operation system to minimal number of physical server.
Compared with prior art, the dispatching method of the operation system based on cloud computing that the embodiment of the present invention provides
And dispatching device, the dispatching algorithm of business cloud system can be simplified, it is possible to more directly and efficiently to business system
System is scheduling, and reduces the complexity of dispatching algorithm, and decreases the scheduling that conventional scheduling algorithms causes
" shake " problem, business that when peak traffic unnecessary virtual machine (vm) migration caused has been reduced or avoided
Systematic function declines problem, improves the stability of operation system.
Accompanying drawing explanation
The schematic flow sheet of the dispatching method based on operation system that Fig. 1 provides for the embodiment of the present invention;
The structural representation of the dispatching device based on operation system that Fig. 2 provides for the embodiment of the present invention.
Detailed description of the invention
For making the technical problem to be solved in the present invention, technical scheme and advantage clearer, below in conjunction with attached
Figure and specific embodiment are described in detail.
The multi-service cloud system using cloud computing technology to build, not only increases the utilization rate of resource, also for joint
Can lower consumption, reduce O&M cost and provide possibility.The operation system of telecom operators, also exists obvious tide
Nighttide characteristic, a kind of scheduling scheme used for above-mentioned tidal regime is as follows, the most by day, along with load
Increase, Servers-all is powered up.Then by the virtual machine of all carrying business, the tune of load balancing is used
Degree principle, by each station server in virtual machine average mobility to system.At night, use energy-conservation tune
Degree algorithm arrangement, typically uses knapsack algorithm, virtual machine is compressed on minimum server operation.
Above-mentioned scheduling scheme lacks the overall situation and considers, only considers the resource load situation at scheduling instance, does not examine
Consider the diversity to operation system load, can bring and running occur scheduling again in the middle of system thus produce
Raw scheduling jitter problem.Illustrate, it is assumed that operation system is in the principle equalized according to CPU, by virtual
When machine spreads out on Servers-all, it is contemplated that cpu load be the cpu load of scheduling instance point.
As time goes on, the load summit of each operation system will gradually arrive.Due to each business system
The load of system is discrepant, such as, carry the virtual machine of MMS center, and CPU usage during peak reaches
To 45%;The virtual machine of carrying sms center, CPU usage during peak reaches 55%, then if
By when running on the scheduling virtual machine of two sms centers to physical server, arrive on business load peak
After Laiing, it is likely that trigger and again dispatch, wherein run scheduling virtual machine is walked, such as that multimedia message is empty
Plan machine moves to other physical servers, and scheduling jitter i.e. occurs.This scheduling jitter phenomenon is in prior art
Dispatching algorithm in be difficult to avoid that.
The scheduling of Incumbent multi-service cloud system, owing to not accounting for the difference of each operation system load
The opposite sex, the most the load to operation system is not predicted, and when therefore easily causing peak traffic, carries industry
The virtual machine of business system, due to physical server load too high, is performed scheduling by cloud management system and migrates.And
The virtual machine (vm) migration occurred during peak traffic, then can because of virtual machine when migrating hydraulic performance decline itself and make
Become the hydraulic performance decline of whole operation system, thus cause the instability of whole operation system.
For solving the problems referred to above, the embodiment of the present invention proposes a kind of new dispatching algorithm, is reduced or avoided above-mentioned
Scheduling jitter phenomenon occurs, and its concrete thought is by the analysis to historic load, predicts the business in future
Situation, thus scientific and efficient ground carries out scheduling virtual machine.Use dispatching method described in the embodiment of the present invention, can
To simplify the dispatching algorithm of multi-service cloud system, more directly and efficiently multi-service cloud system is scheduling,
Reduce the complexity of dispatching algorithm, and reduce the scheduling that conventional scheduling algorithms causes and " shake " problem,
The operation system degradation problem that virtual machine (vm) migration unnecessary when avoiding peak traffic causes.
Find after existing operation system is studied, when operation system is run, if used daytime
The dispatching algorithm of load balancing class is scheduling, may eventually reach a stable state, i.e. virtual machine will not again due to
The load situation of change of operation system and cause the migration of virtual machine, now, referred to as stable state.By now institute
There is the relation between physical server and all virtual machines run on it, map according to membership.
Form the mapping relations table of a stable state, corresponding to day mode.
In like manner, at night, use the Energy Saving Algorithm including knapsack algorithm, virtual machine is compressed to as far as possible
Run on few physical server.After compression, reach a stable state, also by virtual machine now and
Its physical server run maps, and writes down mapping relations table now, corresponding to night mode simultaneously.
When operation system is switched to day mode by night mode, scheduler is according to the mould on daytime recorded before
Virtual machine under formula and the mapping relations table of physical server, directly by virtual machine (vm) migration to the physics belonged to
On server.Such scheduling, is equivalent to restore the system to the lower state of last time.Due to existing
Miscellaneous service system including telecommunication service system, it is continuous print and tool in being supported on a period of time
There are specific periodicity and regularity, so the stable state of yesterday generally can be predicted, are also steady for today
State, or perhaps maximum of probability stable state, this is the most basic principle of embodiment of the present invention dispatching method.
In like manner, for day mode when night mode switches, it is also possible to according to the night mode recorded before
Under stable state time mapping relations table, carry out the migration scheduling of virtual machine.
The dispatching method that the embodiment of the present invention proposes, was distributed with virtual machine according to the load in a upper cycle, and came
Load and the virtual machine of prediction next cycle be distributed, it is contemplated that historic load situation, it is possible to the most accurate
Direct virtual machine is migrated to as early as possible a stable state, its dispatching algorithm is extremely simple, and the effect dispatched is accurate
The most efficiently.Refer to Fig. 1, the dispatching method of the operation system based on cloud computing that the embodiment of the present invention provides,
Can apply to a multiple services operation system, this operation system includes multiple physical server and multiple void
Plan machine, described virtual machine is for processing the different business in operation system, and this dispatching method comprises the following steps:
Step 11, when described operation system switches to a scene mode, obtains the described field pre-saved
Mapping relations table between virtual machine and the physical server of described operation system under scape pattern, described mapping is closed
It is that in table, record has described operation system the last time to adjust according to predetermined virtual machine under described scene mode
Spending virtual machine and the mapping relations of physical server that algorithmic dispatching obtains, described scene mode is corresponding to described
Operation system in a periodic time period.
Here, the working time section of described operation system includes busy that is set in advance and that periodically present
Period and idle period.It is busy scene corresponding to scene during the described busy period, corresponding to described idle
The scene during period is idle scene.Concrete, the scene mode in above-mentioned steps 11 can be hereinbefore
The day mode mentioned, i.e. operation system are at busy scene mode corresponding to busy time period, it is also possible to be night
Evening, pattern, i.e. operation system were at idle scene mode corresponding to idle time period.
Step 12, according to the mapping relations in described mapping relations table, by the virtual machine of described operation system
It is dispatched to the physical server of correspondence.
Here, in above-mentioned steps 12, perform dispatch deal according to the mapping relations in described mapping relations table
Specifically may include that and the physical server of record in described mapping relations table is waken up up, if this physical services
Device already at wake-up states, then can directly perform next step, if this physical server be in lower electricity or
Resting state, then wake up this physical server up;Then, according to the mapping relations in described mapping relations table,
By the scheduling virtual machine of record in described mapping relations table to corresponding physical server;After scheduling completes,
By in described operation system except in described mapping relations table record physical server in addition to residue physics clothes
Business device, carries out lower electricity or dormancy processes, with energy efficient.
By above step, the present embodiment is switched to another scene mode in operation system from a scene mode
Time, according to the mapping relations table under this another scene mode of pre-recorded the last time, quickly by business
The virtual machine of system and the mapping relations of physical server revert to the situation under this scene mode the last,
I.e. it is rapidly achieved a stable state, such that it is able to simplify the process of dispatching algorithm, and scheduling jitter is reduced or avoided
Stable state.
In above step 11, if when described operation system switches to described scene mode, having not been able to look for
Described mapping relations table under the described scene mode preserved, then show to be probably first this scene mould of entrance
Formula, now can be based on virtual machine current in operation system and the membership of physical server, periodically
Gather the occupation condition of each physical server in described operation system, and according to predetermined virtual machine
Dispatching algorithm, performs the migration scheduling of virtual machine, is allowed to arrive a stable state.Further, moving every time
After transposition degree, according to the mapping relations between the virtual machine after migration scheduling and physical server, record and protect
Deposit the described mapping relations table under described scene mode, if there is described mapping relations table, then according to migration
The mapping relations between virtual machine and physical server after scheduling, update described mapping relations table.
Owing to the load of operation system is the most constant, the stable state under a scene mode before reverting to
After, the current situation of change according to business load may be also needed to, continue to be scheduling virtual machine migrating,
To adapt to new change.Now, the embodiment of the present invention is after above-mentioned steps 12, it is also possible to include following
Step: the periodically occupation condition of each physical server in the described operation system of collection, and according to
Described occupation condition, according to predetermined scheduling virtual machine algorithm, performs the migration scheduling of virtual machine;So
After, after the migration scheduling having performed described virtual machine, the virtual machine current according to described operation system and thing
Mapping relations between reason server, update described mapping relations table.
In the embodiment of the present invention, according to described occupation condition, according to predetermined scheduling virtual machine algorithm,
When performing the migration scheduling of virtual machine, can there is different process for different scenes, such as:
When the busy scene that described scene mode is during corresponding to the described busy period, can according to load all
Account method, by the partial virtual machine on the resources occupation rate physical server beyond predetermined threshold, migrate to institute
State on the physical server that resources occupation rate in operation system is minimum;
When the idle scene that described scene mode is during corresponding to the described idle period, can calculate according to knapsack
Method, by the virtual machine (vm) migration of described operation system to minimal number of physical server.
By above method, the embodiment of the present invention can simplify the migration scheduling of the virtual machine of operation system, subtract
Less or avoid scheduling jitter stable state, performance and the stability of operation system are improved.
Based on above method, the embodiment of the present invention additionally provides a kind of device in order to implement said method.Please
Shown in reference Fig. 2, the dispatching device of the operation system based on cloud computing that the embodiment of the present invention provides, including:
Acquiring unit 21, for when described operation system switches to a scene mode, acquisition pre-saves
Described scene mode under described operation system virtual machine and physical server between mapping relations table, institute
Stating record in mapping relations table has described operation system the last time according to predetermined under described scene mode
The virtual machine that scheduling virtual machine algorithmic dispatching obtains and the mapping relations of physical server, described scene mode pair
Operation system described in Ying Yu in a periodic time period;
First scheduling unit 22, for according to the mapping relations in described mapping relations table, by described business
The scheduling virtual machine of system is to corresponding physical server.
Further, above-mentioned dispatching device can also include:
Second scheduling unit, for when described operation system switches to described scene mode, if not yet preserving
There is the described mapping relations table under described scene mode, then periodically gather each thing in described operation system
The occupation condition of reason server, and according to predetermined scheduling virtual machine algorithm, perform the migration of virtual machine
Scheduling;
First updating block, for closing according to the mapping between the virtual machine after migration scheduling and physical server
System, records and preserves the described mapping relations table under described scene mode.
Concrete, above-mentioned first scheduling unit 22 may include that
Migration units, for waking up up the physical server of record in described mapping relations table, and according to described
Mapping relations in mapping relations table, by the scheduling virtual machine of record in described mapping relations table to corresponding thing
Reason server;And,
Power saving unit, for removing the physical server of record in described mapping relations table in described operation system
Outside residue physical server, carry out lower electricity or dormancy and process.
For adapting to the change of the business load of operation system, above-mentioned dispatching device can also include:
3rd scheduling unit, for periodically gathering the resource of each physical server in described operation system
Take situation, and according to described occupation condition, according to predetermined scheduling virtual machine algorithm, perform virtual
The migration scheduling of machine;And, after the migration scheduling having performed described virtual machine, according to described operation system
Mapping relations between current virtual machine and physical server, update described mapping relations table.
Concrete, the working time section of described operation system can include set in advance and periodically present
The busy period and the idle period.Now, described second scheduling unit or the 3rd scheduling unit, use further
In time at described scene mode being busy scene during corresponding to the described busy period, calculate according to load balancing
Method, by the partial virtual machine on the resources occupation rate physical server beyond predetermined threshold, migrates to described industry
On the physical server that in business system, resources occupation rate is minimum;And, at described scene mode for corresponding to institute
During idle scene when stating the idle period, according to knapsack algorithm, by the virtual machine (vm) migration of described operation system extremely
On minimal number of physical server.
Below, then the example by a more specifically dispatching method, the present invention is retouched in more detail
State.This dispatching method is divided into four-stage, i.e. initial stage on daytime, initial stage at night, normality rank on daytime
Section and normality stage at night.
The initial stage on daytime:
Step one: as original state, (one optional according to load balance scheduling algorithm principle for operation system
Algorithm be the algorithm that scatters at random), virtual machine (VM) is dispersed randomly to each physical server hosts
(Host), on, the mapping relations table in the stage on daytime of VM Yu Host is set up.
Step 2: carry out dynamic dispatching according to loading condition.The strategy of telecom operators, usually Host
CPU usage more than 70%, it is necessary to carry out VM scheduling.Therefore, cloud management dispatching device is real-time
Detect the CPU usage of every Host, if any the CPU usage Host more than 70%, select on it
Take the highest VM of CPU, move on the Host of most lightly loaded.Update VM with Host white simultaneously
The mapping relations table in it stage.
Repeatedly perform step 2, until day mode is converted to night mode.
The initial stage at night:
Step one: as original state, operation system is according to (the optional calculation of energy-saving distribution algorithm principle
Method is knapsack algorithm), VM (virtual machine) is compressed on minimum Host (physical host), sets up VM
Mapping relations table with the stage at night of Host.
The normality stage on daytime
Step one: when being switched to day mode by night mode, according to VM and Host the most set up
The mapping relations table in stage on daytime, is quickly dispatched to VM on the Host of correspondence.
Step 2: the CPU usage of detecting real-time every Host, if any CPU usage more than 70%
Host, select on it, to take the highest VM of CPU, move on the Host of most lightly loaded, simultaneously the most more
The mapping relations table in new VM Yu Host stage on daytime.
Periodically perform above-mentioned steps two, until night mode conversion on daytime, till scheduling strategy switching.
The normality stage at night:
Step one: according to the mapping relations table in the stage at night of VM Yu Host set up before, quickly will
VM is dispatched on the Host of correspondence, idle Host is carried out dormancy or under electrically operated, to save electric power
With air conditioner refrigerating consumption.
In sum, the embodiment of the present invention provide dispatching algorithm and dispatching device, it is possible to effectively to operation
Business's multi-service cloud system is scheduling, on the basis of ensureing systematic function, it is possible to simplify dispatching algorithm, real
Now save energy and reduce the cost, it is provided that the Performance And Reliability of operation system, the highest practicality.
The above is the preferred embodiment of the present invention, it is noted that for the common skill of the art
For art personnel, on the premise of without departing from principle of the present invention, it is also possible to make some improvements and modifications,
These improvements and modifications also should be regarded as protection scope of the present invention.
Claims (10)
1. the dispatching method of an operation system based on cloud computing, it is characterised in that including:
When described operation system switches to a scene mode, obtain institute under the described scene mode pre-saved
State the mapping relations table between the virtual machine of operation system and physical server, record in described mapping relations table
There is described operation system the last time according to predetermined scheduling virtual machine algorithmic dispatching under described scene mode
The virtual machine obtained and the mapping relations of physical server, described scene mode is corresponding to described operation system
In a periodic time period;
According to the mapping relations in described mapping relations table, by the scheduling virtual machine of described operation system to corresponding
Physical server.
2. the method for claim 1, it is characterised in that
When described operation system switches to described scene mode, if not yet preserving under described scene mode
Described mapping relations table, then periodically gather the resource occupation of each physical server in described operation system
Situation, and according to predetermined scheduling virtual machine algorithm, perform the migration scheduling of virtual machine;
And, according to the mapping relations between the virtual machine after migration scheduling and physical server, record and protect
Deposit the described mapping relations table under described scene mode.
3. the method for claim 1, it is characterised in that described according in described mapping relations table
Mapping relations, by the scheduling virtual machine of described operation system to corresponding physical server, including:
The physical server of record in described mapping relations table is waken up up, and according in described mapping relations table
Mapping relations, by the scheduling virtual machine of record in described mapping relations table to corresponding physical server;And,
By in described operation system except in described mapping relations table record physical server in addition to residue
Reason server, carries out lower electricity or dormancy processes.
4. the method for claim 1, it is characterised in that described according to described mapping relations table
In mapping relations, by the step of the scheduling virtual machine of described operation system to corresponding physical server it
After, described method also includes:
Periodically gather the occupation condition of each physical server in described operation system, and according to institute
State occupation condition, according to predetermined scheduling virtual machine algorithm, perform the migration scheduling of virtual machine;
And, after the migration scheduling having performed described virtual machine, according to current virtual of described operation system
Mapping relations between machine and physical server, update described mapping relations table.
5. the method as described in claim 2 or 4, it is characterised in that
The working time section of described operation system include the busy period that is set in advance and that periodically present and
The idle period;
Described according to described occupation condition, according to predetermined scheduling virtual machine algorithm, perform virtual machine
Migration scheduling, including:
When the busy scene that described scene mode is during corresponding to the described busy period, calculate according to load balancing
Method, by the partial virtual machine on the resources occupation rate physical server beyond predetermined threshold, migrates to described industry
On the physical server that in business system, resources occupation rate is minimum;
When the idle scene that described scene mode is during corresponding to the described idle period, according to knapsack algorithm,
By on the virtual machine (vm) migration of described operation system to minimal number of physical server.
6. the dispatching device of an operation system based on cloud computing, it is characterised in that including:
Acquiring unit, for when described operation system switches to a scene mode, obtains the institute pre-saved
State the mapping relations table between virtual machine and the physical server of described operation system under scene mode, described in reflect
Penetrate in relation table record have described operation system the last under described scene mode according to predetermined virtual
Machine dispatching algorithm dispatches the mapping relations of virtual machine and the physical server obtained, and described scene mode corresponds to
Described operation system in a periodic time period;
First scheduling unit, for according to the mapping relations in described mapping relations table, by described operation system
Scheduling virtual machine to corresponding physical server.
7. dispatching device as claimed in claim 6, it is characterised in that also include:
Second scheduling unit, for when described operation system switches to described scene mode, if not yet preserving
There is the described mapping relations table under described scene mode, then periodically gather each thing in described operation system
The occupation condition of reason server, and according to predetermined scheduling virtual machine algorithm, perform the migration of virtual machine
Scheduling;
First updating block, for closing according to the mapping between the virtual machine after migration scheduling and physical server
System, records and preserves the described mapping relations table under described scene mode.
8. dispatching device as claimed in claim 6, it is characterised in that described first scheduling unit includes:
Migration units, for waking up up the physical server of record in described mapping relations table, and according to described
Mapping relations in mapping relations table, by the scheduling virtual machine of record in described mapping relations table to corresponding thing
Reason server;And,
Power saving unit, for removing the physical server of record in described mapping relations table in described operation system
Outside residue physical server, carry out lower electricity or dormancy and process.
9. dispatching device as claimed in claim 6, it is characterised in that also include:
3rd scheduling unit, for periodically gathering the resource of each physical server in described operation system
Take situation, and according to described occupation condition, according to predetermined scheduling virtual machine algorithm, perform virtual
The migration scheduling of machine;And, after the migration scheduling having performed described virtual machine, according to described operation system
Mapping relations between current virtual machine and physical server, update described mapping relations table.
10. the dispatching device as described in claim 7 or 9, it is characterised in that the work of described operation system
Busy period that is set in advance and that periodically present and idle period is included as the time period;
Described second scheduling unit or the 3rd scheduling unit, be further used for described scene mode for corresponding to
During busy scene during the described busy period, according to load-balancing algorithm, by resources occupation rate beyond predetermined door
Partial virtual machine on the physical server of limit, migrates to the thing that in described operation system, resources occupation rate is minimum
On reason server;And, when the idle scene that described scene mode is during corresponding to the described idle period,
According to knapsack algorithm, by the virtual machine (vm) migration of described operation system to minimal number of physical server.
Priority Applications (2)
Application Number | Priority Date | Filing Date | Title |
---|---|---|---|
CN201510035570.9A CN105868004B (en) | 2015-01-23 | 2015-01-23 | Scheduling method and scheduling device of service system based on cloud computing |
PCT/CN2015/098308 WO2016115956A1 (en) | 2015-01-23 | 2015-12-22 | Scheduling method and scheduling device for business system based on cloud computing |
Applications Claiming Priority (1)
Application Number | Priority Date | Filing Date | Title |
---|---|---|---|
CN201510035570.9A CN105868004B (en) | 2015-01-23 | 2015-01-23 | Scheduling method and scheduling device of service system based on cloud computing |
Publications (2)
Publication Number | Publication Date |
---|---|
CN105868004A true CN105868004A (en) | 2016-08-17 |
CN105868004B CN105868004B (en) | 2020-10-16 |
Family
ID=56416401
Family Applications (1)
Application Number | Title | Priority Date | Filing Date |
---|---|---|---|
CN201510035570.9A Active CN105868004B (en) | 2015-01-23 | 2015-01-23 | Scheduling method and scheduling device of service system based on cloud computing |
Country Status (2)
Country | Link |
---|---|
CN (1) | CN105868004B (en) |
WO (1) | WO2016115956A1 (en) |
Cited By (6)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
CN107247617A (en) * | 2017-05-17 | 2017-10-13 | 努比亚技术有限公司 | The concocting method of resources of virtual machine, platform on probation and readable storage medium storing program for executing |
CN110609747A (en) * | 2019-08-29 | 2019-12-24 | 联想(北京)有限公司 | Information processing method and electronic equipment |
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 |
WO2021057640A1 (en) * | 2019-09-24 | 2021-04-01 | 阿里巴巴集团控股有限公司 | Process processing method and apparatus, electronic device, and computer readable storage medium |
CN112667392A (en) * | 2020-12-09 | 2021-04-16 | 南方电网数字电网研究院有限公司 | Cloud computing resource allocation method and device, computer equipment and storage medium |
CN112866131A (en) * | 2020-12-30 | 2021-05-28 | 神州绿盟成都科技有限公司 | Traffic load balancing method, device, equipment and medium |
Families Citing this family (2)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
CN107145216A (en) * | 2017-05-05 | 2017-09-08 | 北京景行锐创软件有限公司 | A kind of dispatching method |
US10417035B2 (en) | 2017-12-20 | 2019-09-17 | At&T Intellectual Property I, L.P. | Virtual redundancy for active-standby cloud applications |
Citations (5)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
CN103747085A (en) * | 2014-01-10 | 2014-04-23 | 浪潮电子信息产业股份有限公司 | Storage resource scheduling algorithm under cloud computing operation system |
CN103957231A (en) * | 2014-03-18 | 2014-07-30 | 成都盛思睿信息技术有限公司 | Virtual machine distributed task scheduling method under cloud calculating platform |
US8850434B1 (en) * | 2012-09-14 | 2014-09-30 | Adaptive Computing Enterprises, Inc. | System and method of constraining auto live migration of virtual machines using group tags |
CN104077171A (en) * | 2013-03-28 | 2014-10-01 | 华为技术有限公司 | Processing method and equipment during virtual machine scheduling |
US20150007178A1 (en) * | 2013-06-28 | 2015-01-01 | Kabushiki Kaisha Toshiba | Virtual machines management apparatus, virtual machines management method, and computer readable storage medium |
Family Cites Families (5)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
US9047083B2 (en) * | 2008-09-15 | 2015-06-02 | Vmware, Inc. | Reducing power consumption in a server cluster |
CN102546700B (en) * | 2010-12-23 | 2015-07-01 | 中国移动通信集团公司 | Resource scheduling and resource migration methods and equipment |
US20120266163A1 (en) * | 2011-04-13 | 2012-10-18 | International Business Machines Corporation | Virtual Machine Migration |
CN102890554A (en) * | 2011-07-22 | 2013-01-23 | 鸿富锦精密工业(深圳)有限公司 | Power supply management system and method |
CN103905303B (en) * | 2012-12-28 | 2018-02-23 | 中国移动通信集团公司 | Data processing method, apparatus and system after a kind of virtual machine VM across-the-wire migrations |
-
2015
- 2015-01-23 CN CN201510035570.9A patent/CN105868004B/en active Active
- 2015-12-22 WO PCT/CN2015/098308 patent/WO2016115956A1/en active Application Filing
Patent Citations (5)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
US8850434B1 (en) * | 2012-09-14 | 2014-09-30 | Adaptive Computing Enterprises, Inc. | System and method of constraining auto live migration of virtual machines using group tags |
CN104077171A (en) * | 2013-03-28 | 2014-10-01 | 华为技术有限公司 | Processing method and equipment during virtual machine scheduling |
US20150007178A1 (en) * | 2013-06-28 | 2015-01-01 | Kabushiki Kaisha Toshiba | Virtual machines management apparatus, virtual machines management method, and computer readable storage medium |
CN103747085A (en) * | 2014-01-10 | 2014-04-23 | 浪潮电子信息产业股份有限公司 | Storage resource scheduling algorithm under cloud computing operation system |
CN103957231A (en) * | 2014-03-18 | 2014-07-30 | 成都盛思睿信息技术有限公司 | Virtual machine distributed task scheduling method under cloud calculating platform |
Cited By (10)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
CN107247617A (en) * | 2017-05-17 | 2017-10-13 | 努比亚技术有限公司 | The concocting method of resources of virtual machine, platform on probation and readable storage medium storing program for executing |
CN107247617B (en) * | 2017-05-17 | 2020-11-24 | 北京神州数码云科信息技术有限公司 | Virtual machine resource allocation method, trial platform and readable storage medium |
CN110609747A (en) * | 2019-08-29 | 2019-12-24 | 联想(北京)有限公司 | Information processing method and electronic equipment |
WO2021057640A1 (en) * | 2019-09-24 | 2021-04-01 | 阿里巴巴集团控股有限公司 | Process processing method and apparatus, electronic device, and computer readable storage medium |
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 |
CN112667392A (en) * | 2020-12-09 | 2021-04-16 | 南方电网数字电网研究院有限公司 | Cloud computing resource allocation method and device, computer equipment and storage medium |
CN112667392B (en) * | 2020-12-09 | 2024-01-23 | 南方电网数字电网研究院有限公司 | Cloud computing resource allocation method and device, computer equipment and storage medium |
CN112866131A (en) * | 2020-12-30 | 2021-05-28 | 神州绿盟成都科技有限公司 | Traffic load balancing method, device, equipment and medium |
CN112866131B (en) * | 2020-12-30 | 2023-04-28 | 神州绿盟成都科技有限公司 | Traffic load balancing method, device, equipment and medium |
Also Published As
Publication number | Publication date |
---|---|
WO2016115956A1 (en) | 2016-07-28 |
CN105868004B (en) | 2020-10-16 |
Similar Documents
Publication | Publication Date | Title |
---|---|---|
CN105868004A (en) | Cloud computing based business system scheduling method and apparatus | |
CN102546700B (en) | Resource scheduling and resource migration methods and equipment | |
CN103345298B (en) | Method of data center energy saving system based on virtual IT resource distribution technology | |
Marotta et al. | A simulated annealing based approach for power efficient virtual machines consolidation | |
CN103294521B (en) | A kind of method reducing data center's traffic load and energy consumption | |
CN103179048B (en) | Main frame qos policy transform method and the system of cloud data center | |
CN103559072A (en) | Method and system for implementing bidirectional auto scaling service of virtual machines | |
CN107861796B (en) | Virtual machine scheduling method supporting energy consumption optimization of cloud data center | |
CN106020934A (en) | Optimized deploying method based on virtual cluster online migration | |
CN103856512A (en) | Cloud computing management server, working host, idle host and resource scheduling method | |
CN104142860A (en) | Resource adjusting method and device of application service system | |
CN108429815A (en) | Dynamic resource scheduling method based on OpenStack | |
CN103188277A (en) | Load energy consumption management system, method and server | |
Di Valerio et al. | Optimal virtual machines allocation in mobile femto-cloud computing: An MDP approach | |
Razali et al. | Virtual machine migration implementation in load balancing for cloud computing | |
CN108684075B (en) | Processing resource allocation method under centralized base station architecture | |
CN103870315B (en) | A kind of JVM consumption control methods and device | |
CN106059835A (en) | High-reliability control method for low-energy-consumption computer cluster nodes | |
CN107908461A (en) | A kind of resource dynamic retractility device and implementation method based on Docker | |
CN109960579B (en) | Method and device for adjusting service container | |
CN105607943A (en) | Dynamic deployment mechanism of virtual machine under cloud environment | |
CN103108039A (en) | Service quality guarantee method in low-energy cluster environment | |
CN105847385A (en) | Cloud computing platform virtual machine dispatching method based on operation duration | |
Ajayi et al. | Multi-Class load balancing scheme for QoS and energy conservation in cloud computing | |
CN107203256A (en) | Energy-conservation distribution method and device under a kind of network function virtualization scene |
Legal Events
Date | Code | Title | Description |
---|---|---|---|
C06 | Publication | ||
PB01 | Publication | ||
SE01 | Entry into force of request for substantive examination | ||
SE01 | Entry into force of request for substantive examination | ||
TA01 | Transfer of patent application right |
Effective date of registration: 20200914 Address after: 210012 Nanjing, Yuhuatai District, South Street, Bauhinia Road, No. 68 Applicant after: Nanjing Zhongxing New Software Co.,Ltd. Address before: 518057 Nanshan District Guangdong high tech Industrial Park, South Road, science and technology, ZTE building, Ministry of Justice Applicant before: ZTE Corp. |
|
TA01 | Transfer of patent application right | ||
GR01 | Patent grant | ||
GR01 | Patent grant |