CN109324906A - The method of selection processing node in cloud computing platform - Google Patents
The method of selection processing node in cloud computing platform Download PDFInfo
- Publication number
- CN109324906A CN109324906A CN201811227212.8A CN201811227212A CN109324906A CN 109324906 A CN109324906 A CN 109324906A CN 201811227212 A CN201811227212 A CN 201811227212A CN 109324906 A CN109324906 A CN 109324906A
- Authority
- CN
- China
- Prior art keywords
- node
- information
- class
- nodes
- measurement
- 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.)
- Pending
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]
- G06F9/5005—Allocation of resources, e.g. of the central processing unit [CPU] to service a request
- G06F9/5011—Allocation of resources, e.g. of the central processing unit [CPU] to service a request the resources being hardware resources other than CPUs, Servers and Terminals
- G06F9/5016—Allocation of resources, e.g. of the central processing unit [CPU] to service a request the resources being hardware resources other than CPUs, Servers and Terminals the resource being the memory
-
- 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]
- G06F9/5005—Allocation of resources, e.g. of the central processing unit [CPU] to service a request
- G06F9/5027—Allocation of resources, e.g. of the central processing unit [CPU] to service a request the resource being a machine, e.g. CPUs, Servers, Terminals
Landscapes
- Engineering & Computer Science (AREA)
- Software Systems (AREA)
- Theoretical Computer Science (AREA)
- Physics & Mathematics (AREA)
- General Engineering & Computer Science (AREA)
- General Physics & Mathematics (AREA)
- Computer And Data Communications (AREA)
Abstract
The method of selection processing node in cloud computing platform, comprising: each C class node computing resource is five-tuple SRM=(M, a C in cloud computing platform, D, N, I): 1.M represents node memory information, including total memory, caching and memory information that oneself uses;2.C represents CPU information, the design parameter information including CPU and the current CPU that oneself uses;3.D represents all external memory information, total size and the current size that oneself uses including all external memory;4.N represents network state information, mainly includes network bandwidth and the network delay information with central node;5.I represents the information with the long-run average of the network delay of all G class nodes.
Description
Technical field
The present invention relates to the data processing fields of cloud computing platform.
Background technique
Cloud computing was achieving rapid development in recent years, using cloud platform carry various extensive services oneself through becoming
The trend of the times of the information industry development.In addition, also having been emerged with the progress of multimedia technology, on internet a large amount of novel more
Media business is simultaneously popularized in users.Under such two overall background, how cloud computing platform bearing multimedia is used
Business, become in order to one get a good eye value the problem of.
In order to better describe the resource information and state in system, we define computing resource model, physically
Realize a specific function.Each node computing resource is a four-tuple SRM=(M, C, D, N) in cloud computing platform:
1. M represents node memory information, including total memory, caching and memory information that oneself uses.
2. C represents CPU information, the design parameter information including CPU and the current CPU that oneself uses.
3. D represents all external memory information, total size and the current size that oneself uses including all external memory.
It mainly include network bandwidth and the network delay information with central node 4. N represents network state information.
For time t, the calculation formula of the computing resource metric of each node are as follows:
Wherein,For the weight factor of each resource type,For the metric function of each resource type;
When there is a new operation then, according to its job content, the resource requirement of operation in total can be calculated, then by
Centralized resource manager is in current record information, according to computing resource metric, finds the node group of a collection of candidate, finally by
These node groups have selected current most effective node group, start to calculate this task.
By the analysis to the prior art, the prior art is defective:
Existing cloud computing platform is general lack of the targeted design for being directed to multimedia service, thus directly in the middle part of these platforms
Many problems are faced with when affixing one's name to multimedia service.Be primarily present following problems: (1) most of cloud computing products lack to multimedia
The considerations of special hardware platform needed for business is supported;(2) real-time of multimedia service and to service quality sensitivity to net
Network characteristic proposes requirements at the higher level.
Our team finds that the data traffic in current China Mobile 87% belongs to more matchmakers of video class by research
Body.Current mainstream cloud computing platform is provided to general cloud calculation service and establishes, and what their resource was dominant is CPU energy
Power and storage capacity.But video class multimedia is but difficult to meet.
The way compromised at present is to carry out dimension-reduction treatment to image data on the mobile terminal of isomery to adapt to terminal, this
The way of sample only realizes display image data on mobile terminals reluctantly.But bring very big defect.One be due to
Data are handled in mobile terminal, processing capacity is wretched insufficiency, therefore the effect that image is presented is bad, another can disappear significantly
Consume the processing capacity and electric power of mobile terminal.Therefore, it is necessary to arrange the processing node based on GPU ability.
Solve the problems, such as be: with CPU class node and in the environment of GPU class node, how to calculate measure out this two
The resource of class node.
Summary of the invention
The method of selection processing node in cloud computing platform, comprising:
Step 1:
Each C class node computing resource is a five-tuple SRM=(M, C, D, N, I) in cloud computing platform:
1. M represents node memory information, including total memory, caching and memory information that oneself uses;
2. C represents CPU information, the design parameter information including CPU and the current CPU that oneself uses;
3. D represents all external memory information, total size and the current size that oneself uses including all external memory;
It mainly include network bandwidth and the network delay information with central node 4. N represents network state information;
5. I represents the information with the long-run average of the network delay of all G class nodes;
For time t, the calculation formula of the computing resource metric of each C class node are as follows:
;
Wherein,For the weight factor of each resource type,For the metric function of each resource type;
This measurement is the measurement done in the range of all nodes;
Step 2:
Each G class node computing resource is a five-tuple SRM=(M, G, D, N, I) in cloud computing platform:
1. M represents node memory information, including total memory, caching and memory information that oneself uses;
2. G represents GPU information, the design parameter information including GPU and the current GPU that oneself uses;
3. D represents all external memory information, total size and the current size that oneself uses including all external memory;
It mainly include network bandwidth and the network delay information with central node 4. N represents network state information;
5. I represents the information with the long-run average of the network delay of all C class nodes;
For time t, the calculation formula of the computing resource metric of each C class node are as follows:
Wherein,For the power of each resource type
Repeated factor,For the metric function of each resource type;
This measurement is the measurement done in the range of all nodes;
Step 3: Core server receives task, and task is determined to need to participate in the number of the C class node of task based on the received
Amount isQuantity with G class node is;
Step 4: the metric obtained according to step 1 and 2 determines the candidate C class node for meeting the task and candidate's G class section
Point;
Step 5: metrics range being limited to both candidate nodes set, re-execute the steps 1 measurement, since this measurement is again true
Fixed range of nodes, so, this measurement be re-determined be parameter I value, i.e., range of nodes it is more accurate after, need weight
The new network condition determined between C class node and G class node;
Step 6: metrics range being limited to both candidate nodes set, re-execute the steps 2 measurement, since this measurement is again true
Fixed range of nodes, so, this measurement be re-determined be parameter I value, i.e., range of nodes it is more accurate after, need weight
The new network condition determined between G class node and C class node;
Step 7: the metric obtained according to step 6 and 7 selects resource metric highest from candidate C class nodeA C class
Node selects resource metric highest from candidate G class nodeA G class node;WithFor preset fixed value;
Step 8: node group being formed according to the node of selection, starts to calculate this task.
The method of selection processing node in cloud computing platform, comprising:
Step 1:
Each C class node computing resource is a five-tuple SRM=(M, C, D, N, I) in cloud computing platform:
1. M represents node memory information, including total memory, caching and memory information that oneself uses;
2. C represents CPU information, the design parameter information including CPU and the current CPU that oneself uses;
3. D represents all external memory information, total size and the current size that oneself uses including all external memory;
It mainly include network bandwidth and the network delay information with central node 4. N represents network state information;
5. I represents the information with the long-run average of the network delay of all G class nodes;
For time t, the calculation formula of the computing resource metric of each C class node are as follows:
;
Wherein,For the weight factor of each resource type,For the metric function of each resource type;
This measurement is the measurement done in the range of all nodes;
Step 2:
Each G class node computing resource is a five-tuple SRM=(M, G, D, N, I) in cloud computing platform:
1. M represents node memory information, including total memory, caching and memory information that oneself uses;
2. G represents GPU information, the design parameter information including GPU and the current GPU that oneself uses;
3. D represents all external memory information, total size and the current size that oneself uses including all external memory;
It mainly include network bandwidth and the network delay information with central node 4. N represents network state information;
5. I represents the information with the long-run average of the network delay of all C class nodes;
For time t, the calculation formula of the computing resource metric of each C class node are as follows:
Wherein,For the power of each resource type
Repeated factor,For the metric function of each resource type;
This measurement is the measurement done in the range of all nodes;
Step 3: Core server receives task, and task is determined to need to participate in the number of the C class node of task based on the received
Amount isQuantity with G class node is;
Step 4: the metric obtained according to step 1 and 2 determines the candidate C class node for meeting the task and candidate's G class section
Point;
Step 5: metrics range being limited to both candidate nodes set, re-execute the steps 1 measurement, since this measurement is again true
Fixed range of nodes, so, this measurement be re-determined be parameter I value, i.e., range of nodes it is more accurate after, need weight
The new network condition determined between C class node and G class node;
Step 6: metrics range being limited to both candidate nodes set, re-execute the steps 2 measurement, since this measurement is again true
Fixed range of nodes, so, this measurement be re-determined be parameter I value, i.e., range of nodes it is more accurate after, need weight
The new network condition determined between G class node and C class node;
Step 7: the metric obtained according to step 6 and 7 selects resource metric highest from candidate C class nodeA C class
Node selects resource metric highest from candidate G class nodeA G class node;WithFor according to required by task stock number
And the value being dynamically determined;
Step 8: node group being formed according to the node of selection, starts to calculate this task.
Inventive point 1: proposing the thought that calculate node is classified, and is divided into C class node and G class node.According to task
Demand chooses these two types of nodes respectively.
Inventive point 2: since these two types of nodes need cooperation to execute task, measure in element and increase two class sections
Network condition between point.
Inventive point 3: using two step measures, first determines general scope, then precisive again.
This case proposes the new framework of one kind of cloud computing platform, and calculate node is divided into C class node and G class node, is being spent
When measuring node resource, the Network status between two class nodes, and the method for using two steps measurement are increased, is chosen most suitable
Node participate in execution task.
Specific embodiment
Embodiment 1
Edge Server is divided to two classes, and one kind is the Edge Server based on cpu resource, we are known as C class processing node, another
Class is the Edge Server based on GPU resource, we are known as G class processing node.
Since in the specific business of processing video class, C class processing node and G class processing node will be cooperated jointly to handle,
Therefore, it just needs to consider the situation that these two types of nodes are mutual when measurement handles node capacity.Measure will become
It is very complicated.
Under step of the invention:
In order to better describe the resource information and state in system, we define computing resource model, physically realize
One specific function.
Each C class node computing resource is a five-tuple SRM=(M, C, D, N, I) in cloud computing platform:
1. M represents node memory information, including total memory, caching and memory information that oneself uses.
2. C represents CPU information, the design parameter information including CPU and the current CPU that oneself uses.
3. D represents all external memory information, total size and the current size that oneself uses including all external memory.
It mainly include network bandwidth and the network delay information with central node 4. N represents network state information.
5. I represents the information with the long-run average of the network delay of all G class nodes.
Each G class node computing resource is a five-tuple SRM=(M, G, D, N, I) in cloud computing platform:
1. M represents node memory information, including total memory, caching and memory information that oneself uses.
2. G represents GPU information, the design parameter information including GPU and the current GPU that oneself uses.
3. D represents all external memory information, total size and the current size that oneself uses including all external memory.
It mainly include network bandwidth and the network delay information with central node 4. N represents network state information.
5. I represents the information with the long-run average of the network delay of all C class nodes.
The method of selection processing node in cloud computing platform, comprising:
Step 1:
Each C class node computing resource is a five-tuple SRM=(M, C, D, N, I) in cloud computing platform:
1. M represents node memory information, including total memory, caching and memory information that oneself uses;
2. C represents CPU information, the design parameter information including CPU and the current CPU that oneself uses;
3. D represents all external memory information, total size and the current size that oneself uses including all external memory;
It mainly include network bandwidth and the network delay information with central node 4. N represents network state information;
5. I represents the information with the long-run average of the network delay of all G class nodes;
For time t, the calculation formula of the computing resource metric of each C class node are as follows:
;
Wherein,For the weight factor of each resource type,For the metric function of each resource type;
This measurement is the measurement done in the range of all nodes;
Step 2:
Each G class node computing resource is a five-tuple SRM=(M, G, D, N, I) in cloud computing platform:
1. M represents node memory information, including total memory, caching and memory information that oneself uses;
2. G represents GPU information, the design parameter information including GPU and the current GPU that oneself uses;
3. D represents all external memory information, total size and the current size that oneself uses including all external memory;
It mainly include network bandwidth and the network delay information with central node 4. N represents network state information;
5. I represents the information with the long-run average of the network delay of all C class nodes;
For time t, the calculation formula of the computing resource metric of each C class node are as follows:
Wherein,For the power of each resource type
Repeated factor,For the metric function of each resource type;
This measurement is the measurement done in the range of all nodes;
Step 3: Core server receives task, and task is determined to need to participate in the number of the C class node of task based on the received
Amount isQuantity with G class node is;
Step 4: the metric obtained according to step 1 and 2 determines the candidate C class node for meeting the task and candidate's G class section
Point;
Step 5: metrics range being limited to both candidate nodes set, re-execute the steps 1 measurement, since this measurement is again true
Fixed range of nodes, so, this measurement be re-determined be parameter I value, i.e., range of nodes it is more accurate after, need weight
The new network condition determined between C class node and G class node;
Step 6: metrics range being limited to both candidate nodes set, re-execute the steps 2 measurement, since this measurement is again true
Fixed range of nodes, so, this measurement be re-determined be parameter I value, i.e., range of nodes it is more accurate after, need weight
The new network condition determined between G class node and C class node;
Step 7: the metric obtained according to step 6 and 7 selects resource metric highest from candidate C class nodeA C class
Node selects resource metric highest from candidate G class nodeA G class node;WithFor preset fixed value;
Step 8: node group being formed according to the node of selection, starts to calculate this task.
Embodiment 2
The method of selection processing node in cloud computing platform, comprising:
Step 1:
Each C class node computing resource is a five-tuple SRM=(M, C, D, N, I) in cloud computing platform:
1. M represents node memory information, including total memory, caching and memory information that oneself uses;
2. C represents CPU information, the design parameter information including CPU and the current CPU that oneself uses;
3. D represents all external memory information, total size and the current size that oneself uses including all external memory;
It mainly include network bandwidth and the network delay information with central node 4. N represents network state information;
5. I represents the information with the long-run average of the network delay of all G class nodes;
For time t, the calculation formula of the computing resource metric of each C class node are as follows:
;
Wherein,For the weight factor of each resource type,For the metric function of each resource type;
This measurement is the measurement done in the range of all nodes;
Step 2:
Each G class node computing resource is a five-tuple SRM=(M, G, D, N, I) in cloud computing platform:
1. M represents node memory information, including total memory, caching and memory information that oneself uses;
2. G represents GPU information, the design parameter information including GPU and the current GPU that oneself uses;
3. D represents all external memory information, total size and the current size that oneself uses including all external memory;
It mainly include network bandwidth and the network delay information with central node 4. N represents network state information;
5. I represents the information with the long-run average of the network delay of all C class nodes;
For time t, the calculation formula of the computing resource metric of each C class node are as follows:
Wherein,For the power of each resource type
Repeated factor,For the metric function of each resource type;
This measurement is the measurement done in the range of all nodes;
Step 3: Core server receives task, and task is determined to need to participate in the number of the C class node of task based on the received
Amount isQuantity with G class node is;
Step 4: the metric obtained according to step 1 and 2 determines the candidate C class node for meeting the task and candidate's G class section
Point;
Step 5: metrics range being limited to both candidate nodes set, re-execute the steps 1 measurement, since this measurement is again true
Fixed range of nodes, so, this measurement be re-determined be parameter I value, i.e., range of nodes it is more accurate after, need weight
The new network condition determined between C class node and G class node;
Step 6: metrics range being limited to both candidate nodes set, re-execute the steps 2 measurement, since this measurement is again true
Fixed range of nodes, so, this measurement be re-determined be parameter I value, i.e., range of nodes it is more accurate after, need weight
The new network condition determined between G class node and C class node;
Step 7: the metric obtained according to step 6 and 7 selects resource metric highest from candidate C class nodeA C class
Node selects resource metric highest from candidate G class nodeA G class node;WithFor according to required by task stock number
And the value being dynamically determined;
Step 8: node group being formed according to the node of selection, starts to calculate this task.
Claims (4)
1. the method for selection processing node in cloud computing platform, comprising:
Step 1:
Each C class node computing resource is a five-tuple SRM=(M, C, D, N, I) in cloud computing platform:
M represents node memory information, including total memory, caching and memory information that oneself uses;
C represents CPU information, the design parameter information including CPU and the current CPU that oneself uses;
D represents all external memory information, total size and the current size that oneself uses including all external memory;
N represents network state information, mainly includes network bandwidth and the network delay information with central node;
I represents the information with the long-run average of the network delay of all G class nodes;
For time t, the calculation formula of the computing resource metric of each C class node are as follows:
;
Wherein,For the weight factor of each resource type,For the metric function of each resource type;
This measurement is the measurement done in the range of all nodes;
Step 2:
Each G class node computing resource is a five-tuple SRM=(M, G, D, N, I) in cloud computing platform:
M represents node memory information, including total memory, caching and memory information that oneself uses;
G represents GPU information, the design parameter information including GPU and the current GPU that oneself uses;
D represents all external memory information, total size and the current size that oneself uses including all external memory;
N represents network state information, mainly includes network bandwidth and the network delay information with central node;
I represents the information with the long-run average of the network delay of all C class nodes;
For time t, the calculation formula of the computing resource metric of each C class node are as follows:
Wherein,For the power of each resource type
Repeated factor,For the metric function of each resource type;
This measurement is the measurement done in the range of all nodes;
Step 3: Core server receives task, and task is determined to need to participate in the number of the C class node of task based on the received
Amount isQuantity with G class node is;
Step 4: the metric obtained according to step 1 and 2 determines the candidate C class node for meeting the task and candidate's G class section
Point;
Step 5: metrics range being limited to both candidate nodes set, re-execute the steps 1 measurement, since this measurement is again true
Fixed range of nodes, so, this measurement be re-determined be parameter I value, i.e., range of nodes it is more accurate after, need weight
The new network condition determined between C class node and G class node;
Step 6: metrics range being limited to both candidate nodes set, re-execute the steps 2 measurement, since this measurement is again true
Fixed range of nodes, so, this measurement be re-determined be parameter I value, i.e., range of nodes it is more accurate after, need weight
The new network condition determined between G class node and C class node;
Step 7: the metric obtained according to step 6 and 7 selects resource metric highest from candidate C class nodeA C class
Node selects resource metric highest from candidate G class nodeA G class node;WithFor preset fixed value;
Step 8: node group being formed according to the node of selection, starts to calculate this task.
2. the method for selection processing node in cloud computing platform, comprising:
Step 1:
Each C class node computing resource is a five-tuple SRM=(M, C, D, N, I) in cloud computing platform:
M represents node memory information, including total memory, caching and memory information that oneself uses;
C represents CPU information, the design parameter information including CPU and the current CPU that oneself uses;
D represents all external memory information, total size and the current size that oneself uses including all external memory;
N represents network state information, mainly includes network bandwidth and the network delay information with central node;
I represents the information with the long-run average of the network delay of all G class nodes;
For time t, the calculation formula of the computing resource metric of each C class node are as follows:
;
Wherein,For the weight factor of each resource type,For the metric function of each resource type;
This measurement is the measurement done in the range of all nodes;
Step 2:
Each G class node computing resource is a five-tuple SRM=(M, G, D, N, I) in cloud computing platform:
M represents node memory information, including total memory, caching and memory information that oneself uses;
G represents GPU information, the design parameter information including GPU and the current GPU that oneself uses;
D represents all external memory information, total size and the current size that oneself uses including all external memory;
N represents network state information, mainly includes network bandwidth and the network delay information with central node;
I represents the information with the long-run average of the network delay of all C class nodes;
For time t, the calculation formula of the computing resource metric of each C class node are as follows:
Wherein,For the power of each resource type
Repeated factor,For the metric function of each resource type;
This measurement is the measurement done in the range of all nodes;
Step 3: Core server receives task, and task is determined to need to participate in the number of the C class node of task based on the received
Amount isQuantity with G class node is;
Step 4: the metric obtained according to step 1 and 2 determines the candidate C class node for meeting the task and candidate's G class section
Point;
Step 5: metrics range being limited to both candidate nodes set, re-execute the steps 1 measurement, since this measurement is again true
Fixed range of nodes, so, this measurement be re-determined be parameter I value, i.e., range of nodes it is more accurate after, need weight
The new network condition determined between C class node and G class node;
Step 6: metrics range being limited to both candidate nodes set, re-execute the steps 2 measurement, since this measurement is again true
Fixed range of nodes, so, this measurement be re-determined be parameter I value, i.e., range of nodes it is more accurate after, need weight
The new network condition determined between G class node and C class node;
Step 7: the metric obtained according to step 6 and 7 selects resource metric highest from candidate C class nodeA C class
Node selects resource metric highest from candidate G class nodeA G class node;WithFor according to required by task stock number
And the value being dynamically determined;
Step 8: node group being formed according to the node of selection, starts to calculate this task.
3. a kind of computer program, for executing any one method in method 1-2.
4. the system of selection processing node in cloud computing platform, comprising: central processing unit, memory include on the memory
Computer program, the computer program, for executing any one method in method 1-2.
Priority Applications (1)
Application Number | Priority Date | Filing Date | Title |
---|---|---|---|
CN201811227212.8A CN109324906A (en) | 2018-10-22 | 2018-10-22 | The method of selection processing node in cloud computing platform |
Applications Claiming Priority (1)
Application Number | Priority Date | Filing Date | Title |
---|---|---|---|
CN201811227212.8A CN109324906A (en) | 2018-10-22 | 2018-10-22 | The method of selection processing node in cloud computing platform |
Publications (1)
Publication Number | Publication Date |
---|---|
CN109324906A true CN109324906A (en) | 2019-02-12 |
Family
ID=65262929
Family Applications (1)
Application Number | Title | Priority Date | Filing Date |
---|---|---|---|
CN201811227212.8A Pending CN109324906A (en) | 2018-10-22 | 2018-10-22 | The method of selection processing node in cloud computing platform |
Country Status (1)
Country | Link |
---|---|
CN (1) | CN109324906A (en) |
Cited By (3)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
CN109902822A (en) * | 2019-03-07 | 2019-06-18 | 北京航空航天大学合肥创新研究院 | Memory computing system and method based on Skyrmion racing track memory |
CN110708244A (en) * | 2019-09-27 | 2020-01-17 | 杭州安恒信息技术股份有限公司 | Intelligent routing and scheduling method for cloud protection engine nodes |
CN111866054A (en) * | 2019-12-16 | 2020-10-30 | 北京小桔科技有限公司 | Cloud host building method and device, electronic equipment and readable storage medium |
Citations (5)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
CN102591712A (en) * | 2011-12-30 | 2012-07-18 | 大连理工大学 | Decoupling parallel scheduling method for rely tasks in cloud computing |
CN103067524A (en) * | 2013-01-18 | 2013-04-24 | 浪潮电子信息产业股份有限公司 | Ant colony optimization computing resource distribution method based on cloud computing environment |
US20170024259A1 (en) * | 2013-12-18 | 2017-01-26 | Telefonaktiebolaget Lm Ericsson (Publ) | Method and Network Node for Selecting a Media Processing Unit |
CN107729126A (en) * | 2016-08-12 | 2018-02-23 | 中国移动通信集团浙江有限公司 | A kind of method for scheduling task and device of container cloud |
CN107786963A (en) * | 2017-10-26 | 2018-03-09 | 北京邮电大学 | The migration strategy of calculating task in a kind of self-organizing network |
-
2018
- 2018-10-22 CN CN201811227212.8A patent/CN109324906A/en active Pending
Patent Citations (5)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
CN102591712A (en) * | 2011-12-30 | 2012-07-18 | 大连理工大学 | Decoupling parallel scheduling method for rely tasks in cloud computing |
CN103067524A (en) * | 2013-01-18 | 2013-04-24 | 浪潮电子信息产业股份有限公司 | Ant colony optimization computing resource distribution method based on cloud computing environment |
US20170024259A1 (en) * | 2013-12-18 | 2017-01-26 | Telefonaktiebolaget Lm Ericsson (Publ) | Method and Network Node for Selecting a Media Processing Unit |
CN107729126A (en) * | 2016-08-12 | 2018-02-23 | 中国移动通信集团浙江有限公司 | A kind of method for scheduling task and device of container cloud |
CN107786963A (en) * | 2017-10-26 | 2018-03-09 | 北京邮电大学 | The migration strategy of calculating task in a kind of self-organizing network |
Cited By (3)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
CN109902822A (en) * | 2019-03-07 | 2019-06-18 | 北京航空航天大学合肥创新研究院 | Memory computing system and method based on Skyrmion racing track memory |
CN110708244A (en) * | 2019-09-27 | 2020-01-17 | 杭州安恒信息技术股份有限公司 | Intelligent routing and scheduling method for cloud protection engine nodes |
CN111866054A (en) * | 2019-12-16 | 2020-10-30 | 北京小桔科技有限公司 | Cloud host building method and device, electronic equipment and readable storage medium |
Similar Documents
Publication | Publication Date | Title |
---|---|---|
CN107689925B (en) | Load balancing optimization method and device based on cloud monitoring | |
CN104270416B (en) | Control method for equalizing load and management node | |
CN104038540B (en) | Method and system for automatically selecting application proxy server | |
US20120060167A1 (en) | Method and system of simulating a data center | |
CN109324906A (en) | The method of selection processing node in cloud computing platform | |
De Maio et al. | Modelling energy consumption of network transfers and virtual machine migration | |
WO2022222755A1 (en) | Service processing method and apparatus, and storage medium | |
CN108196935B (en) | Cloud computing-oriented virtual machine energy-saving migration method | |
CN106161661A (en) | A kind of method and device of distributed load equalizing scheduling | |
CN105992020A (en) | Video conversion resource distribution method and system | |
CN104216784B (en) | Focus balance control method and relevant apparatus | |
CN107483292B (en) | Dynamic monitoring method for cloud platform | |
CN112261120B (en) | Cloud-side cooperative task unloading method and device for power distribution internet of things | |
Zhou et al. | An auction-based incentive mechanism for heterogeneous mobile clouds | |
CA3009819A1 (en) | Control device for estimation of power consumption and energy efficiency of application containers | |
CN108337699A (en) | A kind of method, apparatus of online, mobile terminal and storage medium | |
US9565060B2 (en) | Managing a network connection for use by a plurality of application program processes | |
Petri et al. | Modelling and implementing social community clouds | |
CN117135131A (en) | Task resource demand perception method for cloud edge cooperative scene | |
Mahallat | ASTAW: auto-scaling threshold-based approach for web application in cloud computing environment | |
CN105471770B (en) | A kind of message processing method and device based on multi-core processor | |
US9501321B1 (en) | Weighted service requests throttling | |
CN101986608A (en) | Method for evaluating heterogeneous overlay network load balance degree | |
CN106201711B (en) | A kind of task processing method and server | |
Yuan et al. | A DRL-Based Container Placement Scheme with Auxiliary Tasks. |
Legal Events
Date | Code | Title | Description |
---|---|---|---|
PB01 | Publication | ||
PB01 | Publication | ||
SE01 | Entry into force of request for substantive examination | ||
SE01 | Entry into force of request for substantive examination |