CN104572272B - A kind of method for scheduling task, apparatus and system - Google Patents
A kind of method for scheduling task, apparatus and system Download PDFInfo
- Publication number
- CN104572272B CN104572272B CN201310476104.5A CN201310476104A CN104572272B CN 104572272 B CN104572272 B CN 104572272B CN 201310476104 A CN201310476104 A CN 201310476104A CN 104572272 B CN104572272 B CN 104572272B
- Authority
- CN
- China
- Prior art keywords
- efficiency ratio
- energy efficiency
- nucleoid
- node
- value
- Prior art date
- Legal status (The legal status is an assumption and is not a legal conclusion. Google has not performed a legal analysis and makes no representation as to the accuracy of the status listed.)
- Active
Links
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
- G06F9/50—Allocation of resources, e.g. of the central processing unit [CPU]
- G06F9/5094—Allocation of resources, e.g. of the central processing unit [CPU] where the allocation takes into account power or heat criteria
-
- Y—GENERAL TAGGING OF NEW TECHNOLOGICAL DEVELOPMENTS; GENERAL TAGGING OF CROSS-SECTIONAL TECHNOLOGIES SPANNING OVER SEVERAL SECTIONS OF THE IPC; TECHNICAL SUBJECTS COVERED BY FORMER USPC CROSS-REFERENCE ART COLLECTIONS [XRACs] AND DIGESTS
- Y02—TECHNOLOGIES OR APPLICATIONS FOR MITIGATION OR ADAPTATION AGAINST CLIMATE CHANGE
- Y02D—CLIMATE CHANGE MITIGATION TECHNOLOGIES IN INFORMATION AND COMMUNICATION TECHNOLOGIES [ICT], I.E. INFORMATION AND COMMUNICATION TECHNOLOGIES AIMING AT THE REDUCTION OF THEIR OWN ENERGY USE
- Y02D10/00—Energy efficient computing, e.g. low power processors, power management or thermal management
Abstract
The present invention is applied to field of computer technology, there is provided a kind of method for scheduling task, apparatus and system, methods described include:The Energy Efficiency Ratio information of respectively all first nucleoids from node is obtained, and the Energy Efficiency Ratio information is ranked up according to the size order of the value of Energy Efficiency Ratio, the Energy Efficiency Ratio information result collection after generation sequence;Obtain all processes in being run on current second nucleoid for task;Since most not busy process, concentrated in the Energy Efficiency Ratio information result, search the first nucleoid for being adapted to each process of operation successively since the efficiency ratio information of minimum;If being concentrated in the Energy Efficiency Ratio information result and finding the first nucleoid for being adapted to the operation specified process, will be run on the specified process migration to first nucleoid.The present invention, fully using first kind nuclear resource, because the first nucleoid has relatively high Energy Efficiency Ratio, it is possible to reduce the energy consumption cost at whole task data center.
Description
Technical field
The invention belongs to field of computer technology, more particularly to a kind of method for scheduling task, apparatus and system.
Background technology
The development of data center substantially experienced several stages.In early days, the huge computer floor of computer realm is data
The blank at center.In the 1990s, the computation schema of client/server is progressively popularized, server starts to be located separately,
Data center comes into vogue.At the beginning of 21 century, with the development of internet, Internet data center (Internet Data
Center, IDC) rapid rising, the service that it can be provided constantly upgrades.
In recent years, the problems such as climate warming, energy shortage, being faced with data center reduces energy consumption, cost-effective tight
High challenge, data center's energy-conservation are had historically been a concern, and the concept of " green data center " is arisen at the historic moment, and will turn into number
According to the new developing direction in center, market value is unprecedentedly huge here.
ARM companies ARMv8 is especially proposed big small nut (the big cores of A57, A53 small nuts) scheme, to adapt to the change of data center
Leather, the particularly processing to big data.For example, some access time sections, data traffic very little and not urgent, if this when
Only it is treated with big verification data in waste, if only handled with small nut data, some periods occurs again
Situation of the small nut to data processing not in time.
In consideration of it, prior art use in a Socket, data traffic very little and it is not urgent when, by big core appoint
Business, which is dispatched to above small nut, goes to perform to realize energy-conservation, wherein, Socket refers to " physical cpu ", is installed on general terminal device
There are several fans, just there are several Socket on the terminal device.
For example, in the case of single Socket, there is a big core from node-node 1 in cluster, a small nut, can incite somebody to action
The course allocation that big core in node 1 needs to perform replaces big core to perform task to small nut by small nut.Although pass through this size
Core switching method, can cause the energy consumption of system to reduce, but this switching simply switching between the big small nut of intra-node,
That is it is to carry out big small nut switching based on terminal device or in a server internal, does not account for cluster internal
Switching between the big small nut in node, such as, the switching between the small nut in big core and node 2 in node 1, therefore do not have
There is the Energy Efficiency Ratio in view of each small nut, can not really play the effect of the high small nut of Energy Efficiency Ratio.
The content of the invention
The embodiments of the invention provide a kind of method for scheduling task, apparatus and system, it is intended to solves prior art and is carrying out
During task scheduling, it is impossible to the problem of fully playing the effect of the high core of Energy Efficiency Ratio.
First aspect, there is provided a kind of method for scheduling task, methods described include:
Obtain respectively from node all first nucleoids Energy Efficiency Ratio information, and according to Energy Efficiency Ratio value size order to institute
State Energy Efficiency Ratio information to be ranked up, the Energy Efficiency Ratio information result collection after generation sequence, the value of the Energy Efficiency Ratio of first nucleoid is high
In default first efficiency ratio;
All processes in being run on current second nucleoid for task are obtained, the value of the Energy Efficiency Ratio of second nucleoid is less than
Default second efficiency ratio, the first efficiency ratio are more than the second efficiency ratio;
Since most not busy process, concentrate in the Energy Efficiency Ratio information result, looked into successively since the efficiency ratio information of minimum
Look for the first nucleoid for being adapted to each process of operation;
Wherein, it is adapted to operation to specify the first nucleoid of process to refer to concentrate what is found from the Energy Efficiency Ratio information result
Core, the disposal ability of the core are more than the load of the specified process;
, will if being concentrated in the Energy Efficiency Ratio information result and finding the first nucleoid for being adapted to the operation specified process
Run on the specified process migration to first nucleoid.
In the first possible implementation, with reference in a first aspect, since most not busy process, in the Energy Efficiency Ratio information
In result set, search and be adapted to after the first nucleoid of each process of operation successively since the efficiency ratio information of minimum, in addition to:
The first nucleoid for being adapted to the operation specified process is not found if concentrated from the Energy Efficiency Ratio information result,
Then reduce the cpu frequency of current second nucleoid.
In second of possible implementation, with reference to first aspect or according to the first possible implementation,
It is described acquisition respectively from node all first nucleoids Energy Efficiency Ratio information, and according to Energy Efficiency Ratio value size order to the energy
Effect is ranked up than information, before the Energy Efficiency Ratio information result collection after generation sequence, in addition to:
Obtain respectively from the cpu frequency of node;
To being respectively weighted from the cpu frequency of node, CPU service condition value is obtained;
If the service condition value is less than default first processing ability value, the acquisition respectively institute from node is performed
There is the Energy Efficiency Ratio information of the first nucleoid, and the Energy Efficiency Ratio information is ranked up according to the size order of the value of Energy Efficiency Ratio, it is raw
The step of into Energy Efficiency Ratio information result collection after sequence;
First processing ability value is according to respectively set in advance from the disposal ability of all first nucleoids in node.
In the third possible implementation, according to second of possible implementation, described to respectively from node
Cpu frequency is weighted, after obtaining CPU service condition value, in addition to:
If the service condition value is more than default second processing ability value, continuation is run on second nucleoid
All processes in the task;
If the service condition value is more than default first processing ability value, less than default second processing ability value,
It will then be run on the first nucleoid on the node where the task immigration run on second nucleoid to second nucleoid;
The second processing ability value is according to respectively set in advance from the disposal ability of all second nucleoids in node.
Second aspect, there is provided a kind of task scheduling apparatus, it is characterised in that the task scheduling apparatus includes:
Energy Efficiency Ratio acquiring unit, for obtaining the Energy Efficiency Ratio information of respectively all first nucleoids from node, and according to efficiency
The size order of the value of ratio is ranked up to the Energy Efficiency Ratio information, the Energy Efficiency Ratio information result collection after generation sequence, and described the
The value of the Energy Efficiency Ratio of one nucleoid is higher than default first efficiency ratio;
Process acquiring unit, for obtaining all processes in being run on current second nucleoid for task, second class
The value of the Energy Efficiency Ratio of core is less than default second efficiency ratio, and the first efficiency ratio is more than the second efficiency ratio;
Core searching unit, for since most not busy process, being concentrated in the Energy Efficiency Ratio information result, from the efficiency ratio of minimum
Information starts to search the first nucleoid for being adapted to each process of operation successively, wherein, it is adapted to the first nucleoid of the specified process of operation to be
Refer to the core concentrated and found from the Energy Efficiency Ratio information result, the disposal ability of the core is more than the load of the specified process;
Core migration units, if being adapted to the operation specified process for concentrating to find in the Energy Efficiency Ratio information result
The first nucleoid, then will be run on the specified process migration to first nucleoid.
In the first possible implementation, with reference to second aspect, the task scheduling apparatus also includes:
Frequency adjustment unit, if being adapted to the operation finger for concentrating not finding from the Energy Efficiency Ratio information result
Determine the first nucleoid of process, then reduce the cpu frequency of current second nucleoid.
In second of possible implementation, with reference to second aspect or according to the first possible implementation, institute
Stating task scheduling apparatus also includes:
Frequency acquisition unit, for obtaining respectively from the cpu frequency of node;
Service condition value acquiring unit, for being respectively weighted from the cpu frequency of node, obtaining CPU service condition
Value;
First processing units, if being less than default first processing ability value for the service condition value, call efficiency
Than acquiring unit, the Energy Efficiency Ratio information for obtaining respectively all first nucleoids from node is performed by the Energy Efficiency Ratio acquiring unit, and
Size order according to the value of Energy Efficiency Ratio is ranked up to the Energy Efficiency Ratio information, the Energy Efficiency Ratio information result collection after generation sequence
The step of;
First processing ability value is according to respectively set in advance from the disposal ability of all first nucleoids in node.
In the third possible implementation, according to second of possible implementation, the task scheduling apparatus is also
Including:
Second processing unit, if being more than default second processing ability value for the service condition value, continue
All processes in the task are run on second nucleoid;
3rd processing unit, if being more than default first processing ability value for the service condition value, less than default
Second processing ability value, then by the node where the task immigration to second nucleoid run on second nucleoid
Run on first nucleoid;
The second processing ability value is according to respectively set in advance from the disposal ability of all second nucleoids in node.
The third aspect, there is provided a kind of task scheduling system, the task scheduling system include host node and with the main section
At least one of point connection includes task scheduling apparatus as described above from node, the host node.
As can be seen that the embodiment of the present invention passes through the idle process of the comparison that will be performed on the second nucleoid from such scheme
Move to and run on the first suitable nucleoid, fully using first kind nuclear resource, because the first nucleoid has relatively high energy
Imitate ratio, it is possible to the energy consumption cost at whole task data center is reduced, also, more hardware resources can also be used, than
Such as hard disk, internal memory.
Brief description of the drawings
Fig. 1 is the implementation process figure for the method for scheduling task that the embodiment of the present invention one provides;
Fig. 2 is the implementation process figure for the method for scheduling task that the embodiment of the present invention two provides;
Fig. 3 is the structured flowchart for the task scheduling apparatus that the embodiment of the present invention three provides;
Fig. 4 is the structured flowchart for the task scheduling apparatus that the embodiment of the present invention four provides.
Embodiment
In order to make the purpose , technical scheme and advantage of the present invention be clearer, it is right below in conjunction with drawings and Examples
The present invention is further elaborated.It should be appreciated that the specific embodiments described herein are merely illustrative of the present invention, and
It is not used in the restriction present invention.
In embodiments of the present invention, by the idle process migration of the comparison performed on the second nucleoid to the first suitable nucleoid
Upper operation, fully using first kind nuclear resource, because the first nucleoid has relatively high Energy Efficiency Ratio, it is possible to reduce whole
The energy consumption cost at task data center, also, more hardware resources can also be used, such as hard disk, internal memory etc..
It is described in detail below in conjunction with realization of the specific embodiment to the present invention:
Embodiment one
Fig. 1 shows the implementation process for the method for scheduling task that the embodiment of the present invention one provides, with task scheduling system
Host node side exemplified by illustrate, details are as follows:
In step S101, the Energy Efficiency Ratio information of respectively all first nucleoids from node is obtained, and according to the value of Energy Efficiency Ratio
Size order the Energy Efficiency Ratio information is ranked up, generation sequence after Energy Efficiency Ratio information result collection, first nucleoid
The value of Energy Efficiency Ratio be higher than default first efficiency ratio.
In the embodiment of the present invention, the mark of Energy Efficiency Ratio information including each the first nucleoid and with each the first nucleoid
Identify the value of one-to-one Energy Efficiency Ratio.Task scheduling system is connected multiple from node by a host node and with the host node
Composition, wherein, by taking ARM frameworks as an example, each there are big core and small nut from node.Wherein, it is in ARM frameworks that Energy Efficiency Ratio is high
Core is referred to as small nut, and the low core of Energy Efficiency Ratio is referred to as big core, and in Intel Architecture, the class of core is only distinguished by the height of Energy Efficiency Ratio
Type, the title of not big small nut, so, in the present embodiment, the core in task scheduling system is divided into the first nucleoid and the second class
Core, wherein the first nucleoid refers to the higher core of value of Energy Efficiency Ratio, the second nucleoid refers to the value of Energy Efficiency Ratio than relatively low core.
Illustrated below by taking ARM frameworks as an example, wherein the first nucleoid refers to small nut, the second nucleoid refers to greatly
Core.
Each big core and small nut from node can be big core and small nut or ARMv8 under ARMv7 frameworks
Big core and small nut under framework.Each belong to same framework from the big core and small nut in node, but different two are from node
The big core and small nut belonged under two kinds of different frameworks can be possessed, such as, it is to belong to ARMv7 from the big core and small nut in node 1
Big core and small nut under framework, it is big the core and small nut belonged under ARMv8 frameworks from the big core and small nut in node 2, that is,
Say, from node, core under two kinds of frameworks of ARMv7 and ARMv8 is simultaneously deposited, the Energy Efficiency Ratio between big small nut under two kinds of frameworks
Ratio be different.
Table 1 shows the Energy Efficiency Ratio of the Energy Efficiency Ratio information, as can be seen from Table 1 big small nut of the big small nut under ARMv7 frameworks
Ratio be " big core:Small nut ≈ 11:16”.The ratio difference of ARMv8 A57 and A53 Energy Efficiency Ratio is bigger, their Energy Efficiency Ratio
Ratio is " big core:Small nut ≈ 1:4”.Parameters in specific table 1, and the acquisition methods of parameters value are existing skill
Art, it will not be repeated here.
Table 1
After getting the Energy Efficiency Ratio information of each small nut from node, order that can be to the Energy Efficiency Ratio information according to size
It is ranked up, can be specifically sorted according to order from big to small, can also sorts according to order from small to large, not make herein
Limitation, the result set obtained after finally this can be sorted are stored, and specific storage mode is not also restricted herein, this implementation
Example in, for convenience, the Energy Efficiency Ratio information result collection in the form of chained list after memory sequencing, and Energy Efficiency Ratio information be according to by
As low as big sequential storage is into chained list.
In step s 102, all processes in being run on current second nucleoid for task, the efficiency of second nucleoid
The value of ratio is less than default second efficiency ratio, and the first efficiency ratio is more than the second efficiency ratio.
In the embodiment of the present invention, the current task run on big core can include one or more process, and host node can
To obtain all processes in being run on current big core for task, then these processes got are carried out according to idle routine
Sequence, generate process list.During specific sequence, most not busy process can be placed on to the foremost of list, can also will most not busy process
The end of process list is placed on, this is not restricted.
In step s 103, since most not busy process, concentrate in the Energy Efficiency Ratio information result, believe from the efficiency ratio of minimum
Breath starts to search the first nucleoid for being adapted to each process of operation successively, wherein, it is adapted to operation to specify the first nucleoid of process to refer to
The core found is concentrated from the Energy Efficiency Ratio information result, the disposal ability of the core is more than the load of the specified process, if
Concentrated in the Energy Efficiency Ratio information result and find the first nucleoid for being adapted to the operation specified process, then perform step S104;
Otherwise step S105 is performed.
In the embodiment of the present invention, it is adapted to operation to specify the small nut of process to refer to concentrate what is found from Energy Efficiency Ratio information result
The disposal ability of small nut corresponding to Energy Efficiency Ratio information is more than the load for specifying process, if host node, which is found, is adapted to operation institute
The small nut of specified process is stated, then performs step S104, wherein, the process specified refers to enter from one of them in big core
Journey, each process in big core is needed to perform and concentrated in Energy Efficiency Ratio information result, since the efficiency ratio information of minimum successively
Search the operation for the first nucleoid for being adapted to run the process;If searched all over whole Energy Efficiency Ratio information result collection, again without looking into
The small nut for being adapted to operation to specify process is found, then performs step S105.
In step S104, it will be run on the specified process migration to first nucleoid.
In step S105, the cpu frequency of current second nucleoid is reduced.
In the embodiment of the present invention, host node judges whether the cpu frequency of current big core is less than predeterminated frequency value, such as
50%, if it is, whether judge the cpu frequency of the current big core can reduce, if having been reduced to it is minimum, should not be again
Reduce, otherwise continue to reduce the cpu frequency of the current big core.
, can also be from the before the cpu frequency of current second nucleoid is reduced as a preferred embodiment of the present invention
Efficiency ratio is selected to be higher than the core of current second nucleoid in two nucleoids, then by the CPU of task scheduling to the core, if
The core that efficiency ratio is less than current second nucleoid is not found, then just performs the CPU frequencies for reducing current second nucleoid
The step of rate.
The present embodiment, it will be run on the idle process migration of the comparison performed on the second nucleoid to the first suitable nucleoid,
First kind nuclear resource is fully utilized, because the first nucleoid has relatively high Energy Efficiency Ratio, it is possible to reduce whole number of tasks
According to the energy consumption cost at center, also, more hardware resources can also be used, such as hard disk, internal memory etc..
Can one of ordinary skill in the art will appreciate that realizing that all or part of step in the various embodiments described above method is
To instruct the hardware of correlation to complete by program, corresponding program can be stored in a computer read/write memory medium
In, described storage medium, such as ROM/RAM, disk or CD.
Embodiment two
Fig. 2 shows the implementation process for the method for scheduling task that the embodiment of the present invention two provides, and enters by taking host node side as an example
Row explanation, the present embodiment perform on the basis of embodiment one, and details are as follows:
In step s 201, obtain respectively from the cpu frequency in node.
In embodiments of the present invention, host node T time to it is each from node send preset times inquiry cpu frequency
Request, respectively after node receives the request, respective cpu frequency is sent to host node.Host node to preset times, than
Query Result such as 3 times is averaged, using the average value as final cpu frequency.
In addition, each can also actively send respective cpu frequency to host node from node, specifically, how host node obtains
Take respectively from the cpu frequency of node, this is not restricted.
In step S202, each cpu frequency is weighted, obtains cpu usage value.
In step S203, the service condition value is judged, if the service condition value is less than default the
One processing ability value, then step S204 to S208 is performed successively;If the service condition value is more than default second processing energy
Force value, then perform step S209;If the service condition value is more than default first processing ability value, less than default second
Processing ability value, then perform step S210.
In embodiments of the present invention, can according to respectively from node CPU frequency come between big small nut switching realize
Grading control.
Wherein, the first processing ability value is, this reality set in advance according to the disposal ability from all small nuts in node
Applying the first processing ability value in example is after being weighted to the disposal ability of all small nuts, then is averaging, and what is finally obtained is averaged
The 80% of value.
Second processing ability value is set in advance according to the disposal ability from all big cores in node, in the present embodiment
Second processing ability value is after being weighted to the disposal ability of all big cores, then is averaging, the average value finally obtained
80%.
In step S204, the Energy Efficiency Ratio information of respectively all first nucleoids from node is obtained, and according to the value of Energy Efficiency Ratio
Size order the Energy Efficiency Ratio information is ranked up, generation sequence after Energy Efficiency Ratio information result collection, first nucleoid
The value of Energy Efficiency Ratio be higher than default first efficiency ratio.
In step S205, all processes in being run on current second nucleoid for task are obtained, second nucleoid
The value of Energy Efficiency Ratio is less than default second efficiency ratio, and the first efficiency ratio is more than the second efficiency ratio.
In step S206, since most not busy process, concentrate in the Energy Efficiency Ratio information result, believe from the efficiency ratio of minimum
Breath starts to search the first nucleoid for being adapted to each process of operation successively, wherein, it is adapted to operation to specify the first nucleoid of process to refer to
The core found is concentrated from the Energy Efficiency Ratio information result, the disposal ability of the core is more than the load of the specified process, if
Concentrated in the Energy Efficiency Ratio information result and find the first nucleoid for being adapted to the operation specified process, then perform step S207;
Otherwise step S208 is performed.
In step S207, it will be run on the specified process migration to first nucleoid.
In step S208, the cpu frequency of second nucleoid is reduced.
In step S209, continue all processes in running the task on current second nucleoid.
In step S210, by task immigration to the first nucleoid of local node run on current second nucleoid
Operation.
The present embodiment, host node are obtained respectively from the cpu frequency of node, and the task in the second nucleoid is determined according to the frequency
Migration model, the grading control of the first nucleoid and the switching of the second nucleoid is realized, can further control the first nucleoid and second
Switching frequency between nucleoid, improve the availability of task scheduling system.
Embodiment three
Fig. 3 shows the concrete structure block diagram for the task scheduling apparatus that the embodiment of the present invention three provides, for convenience of description,
It illustrate only the part related to the embodiment of the present invention.The task scheduling apparatus 3 can be the software list being built in host node
The unit of member, hardware cell or software and hardware combining, wherein, host node is a node in task scheduling system, the node
It is connected with least one from node, the task scheduling apparatus 3 includes:Energy Efficiency Ratio acquiring unit 31, process acquiring unit 32,
Core searching unit 33 and core migration units 34.
Energy Efficiency Ratio acquiring unit 31, for obtaining the Energy Efficiency Ratio information of respectively all first nucleoids from node, and according to energy
The size order for imitating the value of ratio is ranked up to the Energy Efficiency Ratio information, the Energy Efficiency Ratio information result collection after generation sequence, described
The value of the Energy Efficiency Ratio of first nucleoid is higher than default first efficiency ratio;
Process acquiring unit 32, for obtaining all processes in being run on current second nucleoid for task, described second
The value of the Energy Efficiency Ratio of nucleoid is less than default second efficiency ratio, and the first efficiency ratio is more than the second efficiency ratio;
Core searching unit 33, for since most not busy process, being concentrated in the Energy Efficiency Ratio information result, from the efficiency of minimum
Start to search the first nucleoid for being adapted to each process of operation successively than information, wherein, it is adapted to operation to specify the first nucleoid of process
Refer to concentrate the core found from the Energy Efficiency Ratio information result, the disposal ability of the core is more than the load of the specified process;
Core migration units 34, if for the Energy Efficiency Ratio information result concentrate find be adapted to operation it is described specify into
First nucleoid of journey, then it will be run on the specified process migration to first nucleoid.
As a preferred embodiment of the present invention, the task scheduling apparatus 3 also includes:
Frequency adjustment unit 35, if be adapted to for concentrating not finding from the Energy Efficiency Ratio information result described in operation
First nucleoid of the process of specifying, then reduce the cpu frequency of current second nucleoid.
Task scheduling apparatus 3 provided in an embodiment of the present invention can be applied in foregoing corresponding method for scheduling task embodiment
In one, details will not be repeated here referring to the description of above-described embodiment one.
Example IV
Fig. 4 shows the concrete structure block diagram for the task scheduling apparatus that the embodiment of the present invention four provides, for convenience of description,
It illustrate only the part related to the embodiment of the present invention.The task scheduling apparatus 4 can be the software list being built in host node
The unit of member, hardware cell or software and hardware combining, wherein, host node is a node in task scheduling system, the node
It is connected with least one from node, the task scheduling apparatus 4 is except including the Energy Efficiency Ratio acquiring unit described in embodiment three
31st, process acquiring unit 32, core searching unit 33, core migration units 34, frequency adjustment unit 35, in addition to frequency acquisition list
Member 41, service condition value acquiring unit 42, first processing units 43, the processing unit 45 of second processing unit 44 and the 3rd.
Wherein, frequency acquisition unit 41, for obtaining respectively from the cpu frequency of node;
Service condition value acquiring unit 42, for being respectively weighted from the cpu frequency of node, obtaining CPU use feelings
Condition value;
First processing units 43, if being less than default first processing ability value for the service condition value, call energy
Effect is performed the Energy Efficiency Ratio letter for obtaining respectively all first nucleoids from node by the Energy Efficiency Ratio acquiring unit than acquiring unit 31
Breath, and the Energy Efficiency Ratio information is ranked up according to the size order of the value of Energy Efficiency Ratio, the Energy Efficiency Ratio information after generation sequence
The step of result set;
First processing ability value is according to respectively set in advance from the disposal ability of all first nucleoids in node;
Second processing unit 44, if being more than default second processing ability value for the service condition value, continue
All processes in the task are run on second nucleoid;
3rd processing unit 45, if being more than default first processing ability value for the service condition value, less than pre-
If second processing ability value, then by the node where the task immigration to second nucleoid run on second nucleoid
The first nucleoid on run;
The second processing ability value is according to respectively set in advance from the disposal ability of all second nucleoids in node.
Task scheduling apparatus 4 provided in an embodiment of the present invention can be applied in foregoing corresponding method for scheduling task embodiment
In two, details will not be repeated here referring to the description of above-described embodiment two.
It is worth noting that, in said system embodiment, included unit is simply drawn according to function logic
Point, but above-mentioned division is not limited to, as long as corresponding function can be realized;In addition, each functional unit is specific
Title is also only to facilitate mutually distinguish, the protection domain being not intended to limit the invention.
The foregoing is merely illustrative of the preferred embodiments of the present invention, is not intended to limit the invention, all essences in the present invention
All any modification, equivalent and improvement made within refreshing and principle etc., should be included in the scope of the protection.
Claims (9)
1. a kind of method for scheduling task, it is characterised in that methods described includes:
Obtain respectively from node all first nucleoids Energy Efficiency Ratio information, and according to Energy Efficiency Ratio value size order to the energy
Effect is ranked up than information, and the Energy Efficiency Ratio information result collection after generation sequence, the value of the Energy Efficiency Ratio of first nucleoid is higher than pre-
If the first efficiency ratio;
All processes in being run on current second nucleoid for task are obtained, the value of the Energy Efficiency Ratio of second nucleoid is less than default
The second efficiency ratio, the first efficiency ratio is more than the second efficiency ratio;
Since most not busy process, concentrate, searched successively since the efficiency ratio information of minimum suitable in the Energy Efficiency Ratio information result
Close the first nucleoid for running each process;
Wherein, it is adapted to operation to specify the first nucleoid of process to refer to concentrate the core found from the Energy Efficiency Ratio information result, should
The disposal ability of core is more than the load of the specified process;
The first nucleoid for being adapted to the operation specified process is found if concentrated in the Energy Efficiency Ratio information result, by described in
Specify and run on process migration to first nucleoid.
2. the method as described in claim 1, it is characterised in that since most not busy process, in the Energy Efficiency Ratio information result collection
In, search and be adapted to after the first nucleoid of each process of operation successively since the efficiency ratio information of minimum, in addition to:
The first nucleoid for being adapted to the operation specified process is not found if concentrated from the Energy Efficiency Ratio information result, is dropped
The cpu frequency of low current second nucleoid.
3. method as claimed in claim 1 or 2, it is characterised in that obtain respectively all first nucleoids from node described
Energy Efficiency Ratio information, and the Energy Efficiency Ratio information is ranked up according to the size order of the value of Energy Efficiency Ratio, the energy after generation sequence
Before effect is than information result collection, in addition to:
Obtain respectively from the cpu frequency of node;
To being respectively weighted from the cpu frequency of node, CPU service condition value is obtained;
If the service condition value is less than default first processing ability value, the acquisition is performed respectively all the from node
The Energy Efficiency Ratio information of one nucleoid, and the Energy Efficiency Ratio information is ranked up according to the size order of the value of Energy Efficiency Ratio, generation row
The step of Energy Efficiency Ratio information result collection after sequence;
First processing ability value is according to respectively set in advance from the disposal ability of all first nucleoids in node.
4. method as claimed in claim 3, it is characterised in that described to being respectively weighted from the cpu frequency of node, obtain
After CPU service condition value, in addition to:
If the service condition value is more than default second processing ability value, continue on second nucleoid described in operation
All processes in task;
If the service condition value is more than default first processing ability value, less than default second processing ability value, then will
Run on the first nucleoid on node where the task immigration run on second nucleoid to second nucleoid;
The second processing ability value is according to respectively set in advance from the disposal ability of all second nucleoids in node.
5. a kind of task scheduling apparatus, it is characterised in that the task scheduling apparatus includes:
Energy Efficiency Ratio acquiring unit, for obtaining the Energy Efficiency Ratio information of respectively all first nucleoids from node, and according to Energy Efficiency Ratio
The size order of value is ranked up to the Energy Efficiency Ratio information, the Energy Efficiency Ratio information result collection after generation sequence, the first kind
The value of the Energy Efficiency Ratio of core is higher than default first efficiency ratio;
Process acquiring unit, for obtaining all processes in being run on current second nucleoid for task, second nucleoid
The value of Energy Efficiency Ratio is less than default second efficiency ratio, and the first efficiency ratio is more than the second efficiency ratio;
Core searching unit, for since most not busy process, being concentrated in the Energy Efficiency Ratio information result, from the efficiency ratio information of minimum
Start to search the first nucleoid for being adapted to each process of operation successively, wherein, be adapted to operation specify the first nucleoid of process refer to from
The Energy Efficiency Ratio information result concentrates the core found, and the disposal ability of the core is more than the load of the specified process;
Core migration units, if finding be adapted to the operation specified process for being concentrated in the Energy Efficiency Ratio information result
One nucleoid, then it will be run on the specified process migration to first nucleoid.
6. task scheduling apparatus as claimed in claim 5, it is characterised in that the task scheduling apparatus also includes:
Frequency adjustment unit, if for from the Energy Efficiency Ratio information result concentrate do not find be adapted to operation it is described specify into
First nucleoid of journey, then reduce the cpu frequency of current second nucleoid.
7. the task scheduling apparatus as described in claim 5 or 6, it is characterised in that the task scheduling apparatus also includes:
Frequency acquisition unit, for obtaining respectively from the cpu frequency of node;
Service condition value acquiring unit, for being respectively weighted from the cpu frequency of node, obtaining CPU service condition value;
First processing units, if being less than default first processing ability value for the service condition value, Energy Efficiency Ratio is called to obtain
Take unit, by the Energy Efficiency Ratio acquiring unit perform obtain respectively from node all first nucleoids Energy Efficiency Ratio information, and according to
The size order of the value of Energy Efficiency Ratio is ranked up to the Energy Efficiency Ratio information, the step of the Energy Efficiency Ratio information result collection after generation sequence
Suddenly;
First processing ability value is according to respectively set in advance from the disposal ability of all first nucleoids in node.
8. task scheduling apparatus as claimed in claim 7, it is characterised in that the task scheduling apparatus also includes:
Second processing unit, if being more than default second processing ability value for the service condition value, continue described
All processes in the task are run on second nucleoid;
3rd processing unit, if being more than default first processing ability value for the service condition value, less than default
Two processing ability values, then by first on the node where the task immigration to second nucleoid run on second nucleoid
Run on nucleoid;
The second processing ability value is according to respectively set in advance from the disposal ability of all second nucleoids in node.
9. a kind of task scheduling system, it is characterised in that the task scheduling system includes host node and connected with the host node
What is connect is at least one from node, and the host node includes the task scheduling apparatus as described in any one of claim 5 to 8.
Priority Applications (2)
Application Number | Priority Date | Filing Date | Title |
---|---|---|---|
CN201310476104.5A CN104572272B (en) | 2013-10-12 | 2013-10-12 | A kind of method for scheduling task, apparatus and system |
PCT/CN2014/086124 WO2015051685A1 (en) | 2013-10-12 | 2014-09-09 | Task scheduling method, device and system |
Applications Claiming Priority (1)
Application Number | Priority Date | Filing Date | Title |
---|---|---|---|
CN201310476104.5A CN104572272B (en) | 2013-10-12 | 2013-10-12 | A kind of method for scheduling task, apparatus and system |
Publications (2)
Publication Number | Publication Date |
---|---|
CN104572272A CN104572272A (en) | 2015-04-29 |
CN104572272B true CN104572272B (en) | 2018-02-09 |
Family
ID=52812510
Family Applications (1)
Application Number | Title | Priority Date | Filing Date |
---|---|---|---|
CN201310476104.5A Active CN104572272B (en) | 2013-10-12 | 2013-10-12 | A kind of method for scheduling task, apparatus and system |
Country Status (2)
Country | Link |
---|---|
CN (1) | CN104572272B (en) |
WO (1) | WO2015051685A1 (en) |
Families Citing this family (8)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
CN105354084B (en) * | 2015-10-30 | 2018-12-18 | 浪潮(北京)电子信息产业有限公司 | A kind of CPU task immigration method and system based on bandwidth scheduling |
CN107066326A (en) * | 2017-03-27 | 2017-08-18 | 深圳市金立通信设备有限公司 | The method and terminal of a kind of scheduler task |
CN108021223A (en) * | 2017-11-30 | 2018-05-11 | 奇酷互联网络科技(深圳)有限公司 | Optimize method, equipment, mobile terminal and the computer-readable storage medium of stand-by power consumption |
CN110968415B (en) * | 2018-09-29 | 2022-08-05 | Oppo广东移动通信有限公司 | Scheduling method and device of multi-core processor and terminal |
CN109947569B (en) * | 2019-03-15 | 2021-04-06 | Oppo广东移动通信有限公司 | Method, device, terminal and storage medium for binding core |
CN110290215B (en) * | 2019-06-28 | 2021-09-28 | 深圳前海微众银行股份有限公司 | Signal transmission method and device |
CN115016885B (en) * | 2021-11-19 | 2023-11-24 | 荣耀终端有限公司 | Virtual machine garbage recycling operation method and electronic equipment |
CN114461404B (en) * | 2022-04-01 | 2022-07-08 | 统信软件技术有限公司 | Process migration method, computing device and readable storage medium |
Citations (4)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
CN101604264A (en) * | 2009-07-08 | 2009-12-16 | 深圳先进技术研究院 | The method for scheduling task of supercomputer and system |
JP2010160565A (en) * | 2009-01-06 | 2010-07-22 | Ricoh Co Ltd | Task scheduling apparatus, control method for task scheduling, and control program for task scheduling |
CN102520785A (en) * | 2011-12-27 | 2012-06-27 | 东软集团股份有限公司 | Energy consumption management method and system for cloud data center |
CN102707996A (en) * | 2012-05-15 | 2012-10-03 | 江苏中科梦兰电子科技有限公司 | Task scheduling method for heterogeneous multi-core processor |
Family Cites Families (3)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
US8429663B2 (en) * | 2007-03-02 | 2013-04-23 | Nec Corporation | Allocating task groups to processor cores based on number of task allocated per core, tolerable execution time, distance between cores, core coordinates, performance and disposition pattern |
KR101065436B1 (en) * | 2010-12-07 | 2011-09-19 | 경상대학교산학협력단 | Stochastic scheduling of a real-time parallel task with uncertain computation amount on mulit-core processors |
CN102208986B (en) * | 2011-03-22 | 2014-01-15 | 曙光信息产业(北京)有限公司 | Cluster power consumption distribution and control method |
-
2013
- 2013-10-12 CN CN201310476104.5A patent/CN104572272B/en active Active
-
2014
- 2014-09-09 WO PCT/CN2014/086124 patent/WO2015051685A1/en active Application Filing
Patent Citations (4)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
JP2010160565A (en) * | 2009-01-06 | 2010-07-22 | Ricoh Co Ltd | Task scheduling apparatus, control method for task scheduling, and control program for task scheduling |
CN101604264A (en) * | 2009-07-08 | 2009-12-16 | 深圳先进技术研究院 | The method for scheduling task of supercomputer and system |
CN102520785A (en) * | 2011-12-27 | 2012-06-27 | 东软集团股份有限公司 | Energy consumption management method and system for cloud data center |
CN102707996A (en) * | 2012-05-15 | 2012-10-03 | 江苏中科梦兰电子科技有限公司 | Task scheduling method for heterogeneous multi-core processor |
Also Published As
Publication number | Publication date |
---|---|
WO2015051685A1 (en) | 2015-04-16 |
CN104572272A (en) | 2015-04-29 |
Similar Documents
Publication | Publication Date | Title |
---|---|---|
CN104572272B (en) | A kind of method for scheduling task, apparatus and system | |
Chen et al. | Distributed deep learning model for intelligent video surveillance systems with edge computing | |
CN101446962B (en) | Data conversion method, device thereof and data processing system | |
CN108196935B (en) | Cloud computing-oriented virtual machine energy-saving migration method | |
CN102231121B (en) | Memory mapping-based rapid parallel extraction method for big data file | |
US20150032758A1 (en) | High Performance Index Creation | |
CN113454954A (en) | Real-time event detection on social data streams | |
CN110309437A (en) | A kind of information-pushing method and device | |
WO2022105440A1 (en) | Hybrid quantum-classical cloud platform and task execution method | |
CN103023805A (en) | MapReduce system | |
CN105630810B (en) | A method of mass small documents are uploaded in distributed memory system | |
CN102880879A (en) | Distributed processing and support vector machine (SVM) classifier-based outdoor massive object recognition method and system | |
US10031947B2 (en) | Method and apparatus for performing a search operation on heterogeneous computing systems | |
CN101719148A (en) | Three-dimensional spatial information saving method, device, system and dispatching system | |
CN105426228B (en) | A kind of OpenStack virtual machine placement methods towards live streaming media and video code conversion | |
CN116126488A (en) | Self-adaptive resource scheduling method and system for server unaware computing and computer equipment | |
CN105389196A (en) | Method for processing seismic data by utilizing virtualization technology | |
Juran et al. | Data organization and retrieval on parallel air channels: Performance and energy issues | |
Zhang et al. | A locally distributed mobile computing framework for dnn based android applications | |
CN108319604B (en) | Optimization method for association of large and small tables in hive | |
CN110069319B (en) | Multi-target virtual machine scheduling method and system for cloud resource management | |
CN110888713A (en) | Trusted virtual machine migration algorithm for heterogeneous cloud data center | |
CN102497410B (en) | Method for dynamically partitioning computing resources of cloud computing system | |
CN101969402A (en) | Data exchanging method based on parallel processing | |
CN103678545A (en) | Network resource clustering method and device |
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 | ||
TR01 | Transfer of patent right | ||
TR01 | Transfer of patent right |
Effective date of registration: 20200415 Address after: 518129 Bantian HUAWEI headquarters office building, Longgang District, Guangdong, Shenzhen Patentee after: HUAWEI TECHNOLOGIES Co.,Ltd. Address before: 301, A building, room 3, building 301, foreshore Road, No. 310000, Binjiang District, Zhejiang, Hangzhou Patentee before: Huawei Technologies Co.,Ltd. |