CN106598693A - Energy consumption and load aware virtual machine integration method based on time delay strategy - Google Patents
Energy consumption and load aware virtual machine integration method based on time delay strategy Download PDFInfo
- Publication number
- CN106598693A CN106598693A CN201610310483.4A CN201610310483A CN106598693A CN 106598693 A CN106598693 A CN 106598693A CN 201610310483 A CN201610310483 A CN 201610310483A CN 106598693 A CN106598693 A CN 106598693A
- Authority
- CN
- China
- Prior art keywords
- server
- virtual machine
- load
- migration
- energy consumption
- 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
- G06F9/44—Arrangements for executing specific programs
- G06F9/455—Emulation; Interpretation; Software simulation, e.g. virtualisation or emulation of application or operating system execution engines
- G06F9/45533—Hypervisors; Virtual machine monitors
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06F—ELECTRIC DIGITAL DATA PROCESSING
- G06F1/00—Details not covered by groups G06F3/00 - G06F13/00 and G06F21/00
- G06F1/26—Power supply means, e.g. regulation thereof
- G06F1/32—Means for saving power
- G06F1/3203—Power management, i.e. event-based initiation of a power-saving mode
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06F—ELECTRIC DIGITAL DATA PROCESSING
- G06F1/00—Details not covered by groups G06F3/00 - G06F13/00 and G06F21/00
- G06F1/26—Power supply means, e.g. regulation thereof
- G06F1/32—Means for saving power
- G06F1/3203—Power management, i.e. event-based initiation of a power-saving mode
- G06F1/3234—Power saving characterised by the action undertaken
- G06F1/329—Power saving characterised by the action undertaken by task scheduling
-
- 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
- G06F9/44—Arrangements for executing specific programs
- G06F9/455—Emulation; Interpretation; Software simulation, e.g. virtualisation or emulation of application or operating system execution engines
- G06F9/45533—Hypervisors; Virtual machine monitors
- G06F9/45558—Hypervisor-specific management and integration aspects
- G06F2009/4557—Distribution of virtual machine instances; Migration and load balancing
-
- 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 invention provides an energy consumption and load aware virtual machine integration method based on a time delay strategy. The method comprises the following steps: setting a light load threshold value, an overload threshold value and a mark threshold value; periodically forecasting a load of a server; and selecting a server for re-operating a migrated virtual machine. According to the method, a migration operation is marked as emergency or non-emergency based on a special load threshold value; if a migration operation is marked as emergency, the migration operation is executed immediately; and otherwise, if the migration operation is marked as non-emergency, the migration operation is postponed. Therefore, via a short period of time delay, a migration alarm caused by an error of a forecasting model disappears as time goes on, and the corresponding migration operation cannot be executed, so that extra energy consumption and adverse effect on the performance of a system are avoided. Meanwhile, during selection of the server for operating the migrated virtual machine, a load detection mechanism is introduced to avoid a new virtual machine migration operation caused by an overload due to the fact that the server operates the migrated virtual machine.
Description
Technical field
The present invention relates to a kind of virtual machine integration method, in particular it relates to a kind of energy consumption and load based on delaying policy
Sensitive virtual machine integration method, belongs to virtual machine migration technology field.
Background technology
Data center scale with computing technique and big data technology fast-developing continuous expansion while data
The energy consumption at center is also being skyrocketed through.As a example by possessing the data center of 1,000,000 servers, the only energy resource consumption of server
The 50% of whole data center's operation and maintenance expenses is account for just, although these server mosts of the time are only with 10%～50% peak value
Computing capability is in work.Even if a server is in idling conditions, it can be also consumed equivalent to during full-load operation 70%
Power consumption.
Virtual machine migration technology is applied in terms of data center's energy-conservation.In the art, first underloading is taken
Virtual machine on business device is all removed makes underloading server in idling conditions, is then switched to the server that these dally low
Power consumption mode (such as park mode), so as to reduce the energy resource consumption of idle running server, and finally reduces whole data center
Energy consumption.In order to ensure the service quality of data center, periodically the load of server is monitored using forecast model, once
It was found that node overload, some virtual machines are migrated out immediately from these servers so that the load of these servers can be maintained at
Within normal range, so as to avoid service-level agreement from breaking rules.
In order to ensure the service quality of data center, load detecting operation periodically can be performed on each server.
Once a server is detected as overload, several virtual machines on this server will be migrated away.Conventional detection
Load of the method by some model prediction servers, the such as method based on threshold value and the method for returning.However, due to prediction
Method generally existing error, inevitable certain server can be detected as overload by mistake, so as to cause these servers
On some virtual machines by unnecessary migration, and these migrations will cause extra energy consumption, and bring unfavorable to systematicness
Affect.
During virtual machine (vm) migration, it is determined that virtual it is necessary to select to rerun these after virtual machine to be migrated
The server of machine.To reduce the energy consumption that causes of migration, have scholar propose a kind of optimal adaptation method that energy consumption is perceived-for each
Virtual machine selects a minimum server of energy consumption growth to rerun it.It is obvious that this method can be increased with minimum energy consumption
It is long to migrate each virtual machine.But, due to have ignored the potential growth of load, this method may cause a server opening
Soon also become overload after moving a virtual machine, so as to need to perform new virtual machine (vm) migration operation to alleviate this server
Load pressure, this also brings along extra energy consumption expense.
The content of the invention
For the migration operation for avoiding these from performing, the present invention proposes a kind of energy consumption based on delaying policy and load
Sensitive virtual machine integration method, comprises the following steps that：
Step one, setting underloading threshold value, overloading threshold and index threshold；
The load of step 2, cyclic forecast server, if prediction load meets or exceeds overloading threshold, labelling should
Server is Overloaded Servers, and carries out migration judgement；
Step 3, it is above-mentioned carry out migration judge when, carry out according to following rule：
If 1) present load of Overloaded Servers is less than index threshold, then it is assumed that the server is mistaken for overload,
Virtual machine (vm) migration thereon is labeled as into " non-emergent "；And before next predetermined period arrives, forbid on that server
Create virtual machine；
If 2) present load of Overloaded Servers meets or exceeds index threshold, then it is assumed that the server can transship, will
Migration operation on the server is labeled as " urgent ", and some virtual machine (vm) migrations are selected from the server；
Step 4, selection rerun the server of migration virtual machine：
1) determine the server for reruning migration virtual machine：Wherein total n server, m virtual machine to be migrated,
M, n are the natural number more than or equal to 1；
If a) the present load sum of the present load of i-th virtual machine and j-th server is less than index threshold,
Using j-th server as a candidate server for running the virtual machine, wherein i, j are natural number, and 1≤i≤m, 1≤
J≤n,；
If b) the present load sum of the present load of i-th virtual machine and jth server meets or exceeds mark threshold
Value, then do not know whether j-th server is transshipped during i-th virtual machine is run, j-th server be classified as time
Before election server, need to do further load detecting；During load detecting, if the present load of j-th server and i-th
Decision threshold of the prediction load sum of individual virtual machine less than overload, then using j-th server as i-th void to be migrated of operation
The candidate server of plan machine；Otherwise, j-th server is excluded from candidate server；
2) for each candidate server, one virtual machine of computation migration is to the energy consumption caused by the server；Therefrom
The server for selecting energy consumption minimum, reruns virtual machine to be migrated.
More specifically, migration judges to adoptCarry out, wherein x represents the current negative of server
Carry, y represents prediction load, TheavyFor the decision threshold of overload, TflagFor the decision threshold of unnecessary migration, f (x, y) is only the
K server skPrediction loadMore than or equal to TheavyIn the case of solve, wherein k ≠ j, k are natural number, and 1≤k≤
n；If skActual loadingNot less than Tflag, functionValue will be greater than less than 1；Conversely,It is worth and is
0；IfNot from skAny virtual machine is migrated out, while also forbidding new virtual machine in skUpper execution；IfFrom skOn migrate out some virtual machines.
More specifically, when selecting to rerun the server of migration virtual machine, if j-th server sjMeetThen by sjAs reruning i-th virtual machine viA candidate server；Otherwise, by sjIt is defined as performing
viA candidate server before, carry out load detecting.
More specifically, when carrying out load detecting, if sjWith viMeetBy sjAs operation viOne time
Election server；Otherwise, by sjFrom operation viCandidate server in exclude.
In the method, according to a special load threshold, migration operation is labeled as urgent and non-emergent.If one
Individual migration operation is marked as promptly, and it should be immediately performed., whereas if being marked as non-emergent, it will be postponed execution.
Thus, by a bit of time delay, As time goes on can be disappeared by the migration alarm that the error of forecast model causes
Lose, corresponding migration operation will not be performed, so as to the unfavorable shadow for avoiding extra energy resource consumption and bring to systematic function
Ring.Additionally, when the virtual machine of server operation migration is selected, introducing load detecting mechanism, it is to avoid a server is because of fortune
The virtual machine moved into of row and overload, and then cause new virtual machine (vm) migration to operate.
Specific embodiment
If a server overload, it is difficult to provide service by the quality that it promises to undertake.To ensure service quality, it is necessary to will
Some virtual machines are migrated out from this server.It is provided with an overloading threshold Theavy, examined according to this threshold period
Look into whether a server transships.If the load of a server reaches or has exceeded Theavy, then this server mistake
Carry.
To avoid the grade of service from breaking rules, after a server overload is predicted, just should as early as possible from this server
Move out several virtual machines.But, due to the presence of forecast error, migrated some virtual machines that originally need not be migrated.To avoid
These unnecessary migration operations, invention introduces index threshold Tflag.By comparing with index threshold, such migration is grasped
Filter out, only retain the migration operation for having to perform.
One main points of present invention work are from all servers for being predicted as and transshipping to find those mispredicted
Take device.For the ease of searching, the present invention devises two migration rules.This two rule is determined when a server is predicted
For the measure that should be adopted after overload.Migrating decision rule isCarry out, wherein x represents server
Present load, y represent prediction load, TheavyFor the decision threshold of overload, TflagFor the decision threshold of unnecessary migration, f (x,
Y) only in k-th server skPrediction loadMore than or equal to TheavyIn the case of solve, wherein k ≠ j, k are natural number,
And 1≤k≤n；If skActual loadingNot less than Tflag, functionValue will be greater than less than 1；Conversely,It is worth for 0；IfNot from skAny virtual machine is migrated out, while also forbidding new virtual machine in skOn
Perform；IfFrom skOn migrate out some virtual machines.
Although skOverload is predicted to be, but any virtual machine will not be gone out from the server migration.AssumeLess than Tflag,,
Mean skPredicting the outcome for overload is likely to be to be affected by error prediction model.That is, skPossibly by wrong pre-
Survey for overload, from skIt is probably unnecessary to migrate out virtual machine.These migration operations should be marked as " non-emergent ", and
Postpone and performing.Meanwhile, to avoid skOn load it is not outside influences, should forbid creating on that server or again
Perform other virtual machines.If skStrictly mispredicted for overload, after a period of delay, its load is still
Normal level is will remain in, virtual machine originally to be moved out need not also be moved out, so as to avoid the migration operation that originally need not be performed.
If skOn migration operation be marked as " non-emergent ", and skLoad again growth, s as predictionkAlso not
Overload can identify because meeting such server and can be migrated rule soon.According to migration rules, in skIt is predicted
After for overload, if functionMore than 0, skOn some servers should move out immediately and avoid its load further
Increase.More than 0,T is exceededflag, skThere is great probability to increase for overload.To ensure service quality, migration behaviour
Work should be labeled as " urgent ", and as early as possible from skExecution is moved out some virtual machines.
If a server is predicted to be overload, and its current load more than Tflag, should take from this immediately
Migrate out some virtual machines to alleviate load pressure on business device.If the server for reruning migration virtual machine selects improper,
After virtual machine of moving into, server may transship quickly, re-start virtual machine (vm) migration again.In a worse case, service
It is improper that device is selected, it is also possible to can cause a series of virtual machine (vm) migration.These operations will cause extra energy consumption, and to systematicness
Can be negatively affected.To eliminate these migration operations, the present invention needs careful selection virtual machine.
For ease of selecting, server is divided into two classes by the present invention.First kind server is when newly assigned virtual machine is run
Will not transship, and whether Equations of The Second Kind server does not then know to transship.If a server belongs to the first kind, can directly by
Its candidate server as one virtual machine of operation.But, if belonging to Equations of The Second Kind, needed to do further before selecting
Load detecting.According to this analysis, the present invention devises following two rule to distinguish this two classes server.
Server selects rule 1:If sjMeet inequalityBy sjAs reruning viCandidate clothes
Business device.
Server selects rule 2:If sjAbove-mentioned inequality is unsatisfactory for, by sjIt is defined as performing viA candidate service
Before device, load detecting is carried out.
According to the 1st rule, whenWithAnd be less than TflagWhen, by server sjIt is included into first kind server.Due to front
Put forward condition Tflag<TheavyPresence, sjIn operation viWhen, load is not up to or more than Theavy.According to the second rule, ifWithAnd be more than or equal to Tflag, then do not know sjIn operation viDuring whether transship, needs do further load inspection
Survey.
The present invention devises two rules and carries out load detecting.If sjMeet first load detecting rule, i.e.,With
Sum is less than Theavy, then in operation viS afterwardsjLoad be still below Theavy, i.e. sjWill not transship, can be used as operation viOne
Candidate server.According to Article 2 load detecting rule, if sjIt is unsatisfactory forIn operation viS afterwardsjLoad may
More than Theavy, i.e. sjMay overload.To avoid running viCause new virtual machine (vm) migration afterwards, should be by sjFrom operation viCandidate
Exclude in server.
Load detecting rule 1：If sjWith viMeetBy sjAs operation viA candidate server.
Load detecting rule 2：If sjWith viIt is unsatisfactory forBy sjFrom operation viCandidate server in exclude.
The present invention designs the virtual machine placement method of an energy consumption and load-sensitive based on optimal adaptation decreasing strategy.The party
Order sequence of the method first by all virtual machines to be migrated according to cpu busy percentage from high to low.Then, for each is virtual
Machine, travels through all of server.For each meets the server of resources of virtual machine demand, perform load detecting to guarantee this
Server will not be transshipped when designated virtual machine is run.For by the server of load detecting, calculating each server by moving
The energy consumption that shifting causes increases, and therefrom selects energy consumption to increase minimum server to rerun virtual machine.
Each migration operation is labeled as " urgent " or " non-emergent " according to a load threshold by the method.The overwhelming majority is not
Necessary migration operation belongs to " non-emergent " class.It is unnecessary that the method is avoided by postponing " non-emergent " migration operation
Migration.Unnecessary migration operation be due to node by mistake be predicted as transship what is caused.Postpone the migration behaviour of " non-emergent "
After work, if they are strictly what is caused by error, As time goes on, as the load of server is not as that of prediction
Sample transships, and is held in normal level, and these migration operations will become invalid.When for virtual machine select re-execute its clothes
During business device, the method according to the energy consumption and load that may be caused by the virtual machine, when server re-executes virtual machine only
On the premise of load, the minimum server of prioritizing selection energy consumption.
Claims (4)
1. a kind of virtual machine integration method of the energy consumption and load-sensitive based on delaying policy, it is characterised in that：The method includes
Following steps：
Step one, setting underloading threshold value, overloading threshold and index threshold；
The load of step 2, cyclic forecast server, if prediction load meets or exceeds overloading threshold, the labelling service
Device is Overloaded Servers, and carries out migration judgement；
Step 3, it is above-mentioned carry out migration judge when, carry out according to following rule：
If 1) present load of Overloaded Servers is less than index threshold, then it is assumed that the server is mistaken for overload, by which
On virtual machine (vm) migration be labeled as " non-emergent "；And before next predetermined period arrives, forbid creating on that server
Virtual machine；
If 2) present load of Overloaded Servers meets or exceeds index threshold, then it is assumed that the server can transship, and should
Migration operation on server is labeled as " urgent ", and some virtual machine (vm) migrations are selected from the server；
Step 4, selection rerun the server of migration virtual machine：
1) determine the server for reruning migration virtual machine：N server, m virtual machine to be migrated, m, n are had wherein
It is the natural number more than or equal to 1；
If a) i-th virtual machine viPresent load and j-th server sjPresent load sum be less than index threshold, then
By sjUsed as a candidate server for running the virtual machine, wherein i, j are natural number, and 1≤i≤m, 1≤j≤n,；
If b) viPresent load and sjPresent load sum meet or exceed index threshold, then do not know sjIn operation vi's
During whether transship, by sjBefore being classified as candidate server, need to do further load detecting；During load detecting, such as
Fruit sjPresent load and viPrediction load sum less than overload decision threshold, then by sjAs operation viCandidate service
Device；Otherwise, by sjExclude from candidate server；
2) for each candidate server, one virtual machine of computation migration is to the energy consumption caused by the server；Therefrom select
The minimum server of energy consumption, reruns virtual machine to be migrated.
2. based on the claims 1 a kind of energy consumption and load-sensitive based on delaying policy virtual machine integration method, its
It is characterised by：In above-mentioned steps three, migration judges to adoptCarry out, wherein x represents server
Present load, y represent prediction load, TheavyFor the decision threshold of overload, TflagFor the decision threshold of unnecessary migration, f (x, y)
Only in k-th server skPrediction loadMore than or equal to TheavyIn the case of solve, wherein k ≠ j, k are natural number, and 1
≤k≤n；If skActual loadingNot less than Tflag, functionValue will be greater than less than 1；Conversely,
It is worth for 0；IfNot from skAny virtual machine is migrated out, while also forbidding new virtual machine in skUpper execution；IfFrom skOn migrate out some virtual machines.
3. based on the claims 1 a kind of energy consumption and load-sensitive based on delaying policy virtual machine integration method, its
It is characterised by：In above-mentioned steps four, when the server of migration virtual machine is reruned in selection, if sjMeetThen will
sjAs reruning viA candidate server；Otherwise, then by sjIt is defined as performing viA candidate server it
Before, carry out load detecting.
4. based on the claims 3 a kind of energy consumption and load-sensitive based on delaying policy virtual machine integration method, its
It is characterised by：When carrying out load detecting, if sjWith viMeetBy sjAs operation viA candidate server；
Otherwise, by sjFrom operation viCandidate server in exclude.
Priority Applications (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN201610310483.4A CN106598693B (en) | 2016-05-11 | 2016-05-11 | Delay strategy-based energy consumption and load sensitive virtual machine integration method |
Applications Claiming Priority (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN201610310483.4A CN106598693B (en) | 2016-05-11 | 2016-05-11 | Delay strategy-based energy consumption and load sensitive virtual machine integration method |
Publications (2)
| Publication Number | Publication Date |
|---|---|
| CN106598693A true CN106598693A (en) | 2017-04-26 |
| CN106598693B CN106598693B (en) | 2020-07-28 |
Family
ID=58555970
Family Applications (1)
| Application Number | Title | Priority Date | Filing Date |
|---|---|---|---|
| CN201610310483.4A Expired - Fee Related CN106598693B (en) | 2016-05-11 | 2016-05-11 | Delay strategy-based energy consumption and load sensitive virtual machine integration method |
Country Status (1)
| Country | Link |
|---|---|
| CN (1) | CN106598693B (en) |
Cited By (4)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN109992392A (en) * | 2017-12-29 | 2019-07-09 | 中移（杭州）信息技术有限公司 | A kind of calculation resource disposition method, device and Resource Server |
| CN112256438A (en) * | 2020-06-28 | 2021-01-22 | 腾讯科技（深圳）有限公司 | Load balancing control method and device, storage medium and electronic equipment |
| CN114827281A (en) * | 2022-03-10 | 2022-07-29 | 北京百度网讯科技有限公司 | Method, system and device for sending and receiving network request |
| CN115643419A (en) * | 2021-07-19 | 2023-01-24 | 北京字节跳动网络技术有限公司 | Web live broadcast page display method and device and client |
Citations (5)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN102790793A (en) * | 2012-05-08 | 2012-11-21 | 北京邮电大学 | Decision method and control module facing to cloud computing virtual machine migration |
| WO2013097151A1 (en) * | 2011-12-29 | 2013-07-04 | 华为技术有限公司 | Resource scheduling method and device |
| CN104199736A (en) * | 2014-06-30 | 2014-12-10 | 浙江大学苏州工业技术研究院 | Method for saving energy of data center under cloud environment |
| CN105302632A (en) * | 2015-11-19 | 2016-02-03 | 国家电网公司 | Cloud computing working load dynamic integration method |
| CN105430083A (en) * | 2015-11-27 | 2016-03-23 | 成都微讯云通科技有限公司 | Cloud platform deployment method |
-
2016
- 2016-05-11 CN CN201610310483.4A patent/CN106598693B/en not_active Expired - Fee Related
Patent Citations (5)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| WO2013097151A1 (en) * | 2011-12-29 | 2013-07-04 | 华为技术有限公司 | Resource scheduling method and device |
| CN102790793A (en) * | 2012-05-08 | 2012-11-21 | 北京邮电大学 | Decision method and control module facing to cloud computing virtual machine migration |
| CN104199736A (en) * | 2014-06-30 | 2014-12-10 | 浙江大学苏州工业技术研究院 | Method for saving energy of data center under cloud environment |
| CN105302632A (en) * | 2015-11-19 | 2016-02-03 | 国家电网公司 | Cloud computing working load dynamic integration method |
| CN105430083A (en) * | 2015-11-27 | 2016-03-23 | 成都微讯云通科技有限公司 | Cloud platform deployment method |
Non-Patent Citations (2)
| Title |
|---|
| 李志红: ""云计算虚拟机集群部署与动态迁移技术研究"", 《中国优秀硕士学位论文全文数据库信息科技辑》 * |
| 罗树: ""基于优化APH的虚拟机部署与调度策略"", 《计算机工程与设计》 * |
Cited By (6)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN109992392A (en) * | 2017-12-29 | 2019-07-09 | 中移（杭州）信息技术有限公司 | A kind of calculation resource disposition method, device and Resource Server |
| CN112256438A (en) * | 2020-06-28 | 2021-01-22 | 腾讯科技（深圳）有限公司 | Load balancing control method and device, storage medium and electronic equipment |
| CN112256438B (en) * | 2020-06-28 | 2021-06-25 | 腾讯科技（深圳）有限公司 | Load balancing control method and device, storage medium and electronic equipment |
| CN115643419A (en) * | 2021-07-19 | 2023-01-24 | 北京字节跳动网络技术有限公司 | Web live broadcast page display method and device and client |
| CN114827281A (en) * | 2022-03-10 | 2022-07-29 | 北京百度网讯科技有限公司 | Method, system and device for sending and receiving network request |
| CN114827281B (en) * | 2022-03-10 | 2023-09-29 | 北京百度网讯科技有限公司 | Method, system and device for sending and receiving network request |
Also Published As
| Publication number | Publication date |
|---|---|
| CN106598693B (en) | 2020-07-28 |
Similar Documents
| Publication | Publication Date | Title |
|---|---|---|
| EP3847549B1 (en) | Minimizing impact of migrating virtual services | |
| CN108829494B (en) | Container cloud platform intelligent resource optimization method based on load prediction | |
| CN106598693A (en) | Energy consumption and load aware virtual machine integration method based on time delay strategy | |
| CN109586952B (en) | Server capacity expansion method and device | |
| US20190235922A1 (en) | Controlling Resource Allocation in a Data Center | |
| US10528378B2 (en) | System and method for load estimation of virtual machines in a cloud environment and serving node | |
| CN103607459A (en) | Dynamic resource monitoring and scheduling method of cloud computing platform IaaS layer | |
| CN102111337A (en) | Method and system for task scheduling | |
| CN103916438B (en) | Cloud testing environment scheduling method and system based on load forecast | |
| CN103595805A (en) | Data placement method based on distributed cluster | |
| CN110502431B (en) | System service evaluation method and device and electronic equipment | |
| CN108429815A (en) | Dynamic resource scheduling method based on OpenStack | |
| CN103677960A (en) | Game resetting method for virtual machines capable of controlling energy consumption | |
| CN112836911A (en) | Method and device for determining cell energy-saving parameter, electronic equipment and storage medium | |
| CN114296868B (en) | Virtual machine automatic migration decision method based on user experience in multi-cloud environment | |
| CN109976879B (en) | Cloud computing virtual machine placement method based on resource usage curve complementation | |
| CN104572279A (en) | Node binding-supporting virtual machine dynamic scheduling method | |
| CN114826770A (en) | Big data management platform for intelligent analysis of computer network | |
| Yue et al. | Microservice aging and rejuvenation | |
| CN112559440B (en) | Method and device for realizing serial service performance optimization in multi-small-chip system | |
| KR102284074B1 (en) | Multi-Server Management Method to Enhance Power Efficiency Using Workload Prediction of Virtual Machine Clustering | |
| CN113791913B (en) | Task migration method of brain-like computer operating system | |
| CN106888237A (en) | A kind of data dispatching method and system | |
| Qiu et al. | Multi-objective virtual machine consolidation | |
| CN117539613B (en) | Method for managing shared resource in distributed computing system |
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 | ||
| GR01 | Patent grant | ||
| GR01 | Patent grant | ||
| CF01 | Termination of patent right due to non-payment of annual fee | ||
| CF01 | Termination of patent right due to non-payment of annual fee |
Granted publication date: 20200728 Termination date: 20210511 |