CN106681839A - Elasticity calculation dynamic allocation method - Google Patents
Elasticity calculation dynamic allocation method Download PDFInfo
- Publication number
- CN106681839A CN106681839A CN201611264538.9A CN201611264538A CN106681839A CN 106681839 A CN106681839 A CN 106681839A CN 201611264538 A CN201611264538 A CN 201611264538A CN 106681839 A CN106681839 A CN 106681839A
- Authority
- CN
- China
- Prior art keywords
- physical host
- virtual machine
- cpu
- physical
- internal memory
- Prior art date
- Legal status (The legal status is an assumption and is not a legal conclusion. Google has not performed a legal analysis and makes no representation as to the accuracy of the status listed.)
- Granted
Links
Classifications
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06F—ELECTRIC DIGITAL DATA PROCESSING
- G06F9/00—Arrangements for program control, e.g. control units
- G06F9/06—Arrangements for program control, e.g. control units using stored programs, i.e. using an internal store of processing equipment to receive or retain programs
- G06F9/46—Multiprogramming arrangements
- G06F9/50—Allocation of resources, e.g. of the central processing unit [CPU]
- G06F9/5083—Techniques for rebalancing the load in a distributed system
- G06F9/5088—Techniques for rebalancing the load in a distributed system involving task migration
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06F—ELECTRIC DIGITAL DATA PROCESSING
- G06F9/00—Arrangements for program control, e.g. control units
- G06F9/06—Arrangements for program control, e.g. control units using stored programs, i.e. using an internal store of processing equipment to receive or retain programs
- G06F9/46—Multiprogramming arrangements
- G06F9/48—Program initiating; Program switching, e.g. by interrupt
- G06F9/4806—Task transfer initiation or dispatching
- G06F9/4843—Task transfer initiation or dispatching by program, e.g. task dispatcher, supervisor, operating system
- G06F9/485—Task life-cycle, e.g. stopping, restarting, resuming execution
- G06F9/4856—Task life-cycle, e.g. stopping, restarting, resuming execution resumption being on a different machine, e.g. task migration, virtual machine migration
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06F—ELECTRIC DIGITAL DATA PROCESSING
- G06F2209/00—Indexing scheme relating to G06F9/00
- G06F2209/50—Indexing scheme relating to G06F9/50
- G06F2209/5022—Workload threshold
Abstract
The invention discloses an elasticity calculation dynamic allocation method. The method includes the following steps that 1, when system load unbalance is detected, a physical host with the highest load is selected as a to-be-processed physical host; 2, virtual machines on the to-be-processed physical host are selected, the selected virtual machines are transferred to a target physical host, particularly, VCPU performance data of all the virtual machines on the to-be-processed physical host is acquired, the VCPU performance data is sorted in a descending mode, the virtual machine with the highest VCPU use rate is transferred to a physical host with the lowest CPU use rate, and the virtual machines on the to-be-processed physical host are transferred sequentially according to the VCPU performance data arrangement; 3, when the load of the to-be-processed physical host in the step 1 is smaller than a threshold value or no virtual machine can be transferred out, next physical host with the load exceeding the threshold value is selected as a candidate physical host, and the step 1 is executed again. The purpose of flexible resource deploying is achieved.
Description
Technical field
The present invention relates to field of cloud calculation, more particularly to a kind of elastic calculation dynamic allocation method.
Background technology
Cloud computing is a kind of infrastructure is externally provided as resource brand-new resource provider formula by network.Cloud meter
The key problem of calculation technology is exactly the resource management of platform, and its target is exactly, by the rational management of resource, to make cloud computing platform
User task can efficiently be processed.In cloud computing environment, during in face of the mission requirements of magnanimity, it may appear that resource allocation does not conform to
The situation of reason.Therefore, cloud computing needs this key mechanism of load balancing to solve to distribute asking for task in its server resource
Topic, to reach optimization resource utilization, increases the target of data-handling capacity.Rational load balancing framework and method are to improve
The key factor of cloud computing platform operational efficiency.
Cloud computing platform framework is mostly host node and concentrates to the pattern that is managed from node, the load balancing plan of early stage
Slightly it is deployed on the host node of cloud computing platform, the method is referred to as centralized load-balancing method.Centralized load balancing
Method has the advantages that simple deployment, efficiency high, is easy to safeguard.However, host node also becomes the bottleneck of platform property, especially
Be user task demand it is big, from number of nodes it is many in the case of host node easily paralyse.
According to the characteristics of cloud computing platform, load-balancing method can be ground in terms of task scheduling and resource allocation two
Study carefully.Resource allocation methods are mainly concentrated in dynamically disposing and migrate virtual machine according to the operation conditions of physical node.Appoint
Business scheduling strategy virtual machine loading condition current when then being distributed according to task, adds some prioris and is scheduled, such
The advantage of method is that the system overhead needed for load balancing is small.
The content of the invention
In order to overcome the deficiencies in the prior art, it is an object of the invention to provide a kind of elastic calculation dynamic allocation method,
The advantage of cloud platform service performance problem is occurred without when flexible configuration and guarantee resource allocation that it can realize resource elasticity allotment.
The purpose of the present invention is realized using following technical scheme：
A kind of elastic calculation dynamic allocation method, it is characterised in that comprise the following steps：
Step 1, when detecting system load and being unbalance, select pack heaviest a physical host as pending physics
Main frame；
Step 2, the virtual machine selected on the pending physical host, and the virtual machine (vm) migration that will be selected is to target physical master
On machine, the VCPU performance datas of all virtual machines on the pending physical host are specially obtained, and by the VCPU performance numbers
According to carrying out descending arrangement, by the maximum virtual machine (vm) migration of VCPU utilization rates to a minimum physical host of CPU usage, so
The virtual machine on the pending physical host is migrated successively according to the arrangement of VCPU performance datas afterwards；
Step 3, do not migrated out less than threshold values or without virtual function when the pending physical host load in above-mentioned steps 1
When, then select next load more than the physical host of threshold values as candidate physical main frame, and return to step 1.
Preferably, a physical host of pack heaviest is selected in above-mentioned steps 1 as pending physical host, using greedy
Load of the greedy algorithm to physical host is calculated.
Preferably, system load is unbalance in above-mentioned steps 1, selects a physical host of pack heaviest as pending thing
Managing main frame is：
For the physical host in the resource pool for enabling resource elasticity allotment, when performance indications exceed threshold value, according to thing
The index for managing the historical performance assessment resource of main frame calculates average value, specially：Between the triggered time in setting configuration interface
T is divided into, history average utilization index is calculated using following equation：
AVG (used)=A × Wa+B × Wb+C × Wc,
Wherein, A is that, beyond threshold time section average value, B is the average value in T time before threshold time section, and C is
The average value within the 24+T time before threshold time section, Wa, Wb, Wc is respectively A, the corresponding weighted value of B, C.
Preferably, the virtual machine on the pending physical host is selected in the step 2, specially：Consider the void
Plan machine is to the utilization rate of CPU and the occupancy situation of internal memory, preferential to calculate CPU if CPU and internal memory exceed threshold values, while
Consider internal memory；If only internal memory exceedes threshold values, internal memory is only calculated.
Preferably, virtual machine is as follows to the utilization rate computational methods of CPU：
The proportion that virtual machine i accounts for physical host CPU is：
Wherein, C (t) is CPU average utilization of the physical host in the t time periods, and W is the weight of current virtual machine, and N is to work as
The CPU numbers of preceding virtual machine, W (i) is the weight of virtual machine i, and N (i) is the CPU numbers of virtual machine i, and j is on physical host
Run the number of virtual machine；
After virtual machine i migrations, physical host CPU usage is changed into：
C (c) n+1=C (c) n-C (i),
Wherein, C (c) n are the value of the CPU usage of n-th change, as n=0, C (c) n=C (t).
Preferably, the selection of the target physical main frame is specifically selected using focus equation of equilibrium, and the focus is put down
Weighing apparatus formula be：
Ni=1/ (1-C (c)) * 1/ (1-M (i)),
Wherein, after C (i) is for prediction virtual machine (vm) migration to target physical main frame to be selected, the CPU of target physical main frame to be selected
Utilization rate；C (c) is the value of the CPU usage of change；M (i) is the internal memory for being migrated virtual machine i；Ni values are smaller, show target
The CPU of physical host and the balanced utilization rate of internal memory are low, and performance is better.
Compared to existing technology, the beneficial effects of the present invention are：
The technical program can dynamically realize the online migration of virtual machine, with reality according to the migration strategy from setting
The purpose of existing resource elasticity allotment.
The technical program focuses on the formulation of migration strategy, selection strategy and positioning strategy, calculates virtual machine suitable
Migration results, with this realize resource elasticity allotment.
At the initial stage of migration, service is in source host operation.When migration proceeds to certain phase, destination host has had been provided with fortune
The necessary resource of row system.By a very of short duration switching, source host services in mesh control right transfer to destination host
Main frame on continue to run with.Due to switch time it is very of short duration, thus the imperceptible service of user interruption, optimization resource
Also ensure the reliability of system while utilization rate.
Brief description of the drawings
Fig. 1 is the flow chart of elastic calculation dynamic allocation method during the present invention is implemented；
Fig. 2 is the selection flow that virtual machine is migrated during the present invention is implemented；
Fig. 3 is the positioning flow of target physical main frame during the present invention is implemented.
Specific embodiment
Below, with reference to accompanying drawing and specific embodiment, the present invention is described further：
The present invention discloses a kind of elastic calculation dynamic allocation method, and user can be with flexible configuration automatic load dynamically distributes plan
Slightly, including CPU and internal memory mobility threshold set, set away from property (can set wherein two virtual machines be away from attribute,
When automatic load dynamically distributes are migrated, the migration of two virtual machine separation properties all the time will not Autonomic Migration Framework operate in same physical machine
On), polymerism set (wherein many virtual machines can be set for Aggregate attribute, it is selected when automatic load dynamically distributes migrates
Many virtual machine polymerisms migration all the time can synchronous migration operate in same physical machine), calculated using greedy algorithm
Migration results, finally migrate virtual machine to Desirable physical main frame, until the CPU and EMS memory occupation of all physical hosts according to result
Rate is differed within 20% each other.
Resource elasticity allocation process generally comprises four committed steps：
1st, the determination (migration strategy) on opportunity is migrated：When physical host migrates most preferably；
2nd, suitable migration entity (selection strategy) is selected：There may be multiple in a physical host for the migration that is triggered
Virtual machine, selecting which virtual machine to carry out migration can make the resource capability of migration overhead minimum, release more, that is, be migrated void
The selection of plan machine；
3rd, suitable migration target (positioning strategy) is selected：In a resource pool, the quantity of physical host may compare
It is many, by the effect of virtual machine (vm) migration to which physical host preferably, i.e. the positioning of target physical main frame；
4th, online migration：Copy system state finally rebuilds virtual machine state to destination host in destination host, recovers to hold
OK.In order to ensure the available of virtual machine service in transition process, transition process only has very of short duration downtime.
As shown in figure 1, a kind of elastic calculation dynamic allocation method, it is characterised in that comprise the following steps：
Step 1, when detecting the unbalance (C (t) of system load>=C (threshold)) when, one of selection pack heaviest
Used as pending physical host, C (threshold) is that cpu busy percentage migrates activation threshold value to physical host, and C (t) is detected material
CPU average utilization of the reason main frame in the t time periods；
Step 2, the virtual machine selected on the pending physical host, and the virtual machine (vm) migration that will be selected is to target physical master
On machine, the VCPU performance datas of all virtual machines on the pending physical host are specially obtained, and by the VCPU performance numbers
According to carrying out descending arrangement, by the maximum virtual machine (vm) migration of VCPU utilization rates to a minimum physical host of CPU usage, i.e.,
Choose the maximum virtual machine of VCPU utilization rates and be put on other physical hosts (physical host of the CPU usage less than threshold values),
If put down (needing to consider the CPU usage of target physical main frame whether no more than threshold values, whether the memory value of virtual machine does not surpass
Cross the available memory size of target physical host), then choose a wherein minimum physical host of CPU usage and place.Then
The virtual machine on the pending physical host is migrated successively according to the arrangement of VCPU performance datas；
Step 3, do not migrated out less than threshold values or without virtual function when the pending physical host load in above-mentioned steps 1
When, then select next load more than the physical host of threshold values as candidate physical main frame, and return to step 1.
Wherein, a physical host of pack heaviest is selected as pending physical host, using greedy algorithm to physics
The load of main frame is calculated.
A physical host for selecting pack heaviest is that migration strategy is as pending physical host：Work as cpu busy percentage
More than setting threshold value (can by the selected height of user, in, low three kinds of threshold types, every kind of threshold type corresponds to a specific number
Value scope) and just triggering migration is continued for some time, this patent devises a kind of based on history average utilization level mechanisms
Threshold triggers strategy.
For the physical host in the resource pool for enabling resource elasticity allotment, when performance indications exceed the threshold value for setting,
The index of resource is assessed according to its historical performance to calculate average value.In setting configuration interface【Triggered time is spaced】It is T, comments
After having estimated the resource utilization data collected in the following time period, history average utilization index can be calculated：
AVG (used)=A × Wa+B × Wb+C × Wc
Wherein, A, B, C are respectively beyond threshold time section average value, average within the T time before threshold time section
Value and the average value within the 24+T time before threshold time section；Wa, Wb, Wc are respectively A, the corresponding weighted value of B, C.By
The CPU or internal memory of calculating, as long as having one beyond threshold values, just triggering is migrated.
The virtual machine on the pending physical host is selected in step 2, once that is, selection strategy is as shown in Fig. 2 certain thing
Reason main frame is triggered migration, it is necessary to select which virtual machine on the physical host is migrated, with reach migration overhead it is smaller,
The release purpose such as resource is more.
When migrating objects are selected, the virtual machine must be considered to the utilization rate of CPU and the occupancy situation of internal memory.Virtually
Machine is bigger to the occupancy of internal memory, and the memory mirror of copy is bigger needed for migration, and the system consumption of migration is also bigger, but together
When, migration also releases more memory headrooms in origin node.
Assuming that the proportion that virtual machine i accounts for physical host CPU is：
Wherein, C (t) is to be detected CPU average utilization of the physical host in the t time periods, and W is the power of current virtual machine
Weight, N is the CPU numbers of current virtual machine, and W (i) is the weight of certain virtual machine i, and N (i) is the CPU numbers of virtual machine i, and j is
The number of the operation virtual machine on physical host.
After virtual machine i migrations, physical host CPU usage is changed into：
C (c) n+1=C (c) n-C (i),
Wherein, C (c) n are the value of the CPU usage of n-th change.As n=0, if C (c) n=C (t) CPU, internal memory
Exceed threshold values, preferentially calculate CPU, while considering internal memory；If only internal memory exceedes threshold values, internal memory is only calculated.
The selection of target physical main frame is positioning strategy as shown in Figure 3：In Fig. 3 Mavailable for physical host to be selected can
Use internal memory；
After virtual machine is chosen migration, in order to the use of memory source in balanced physical host and CPU computing resources,
During selection target main frame, virtual machine to be migrated must be considered with (CPU consumption, the memory consumption) of target physical main frame
Matching degree, selects suitable target physical main frame to receive the virtual machine being migrated.
Focus equation of equilibrium：
N i=1/ (1-C (c)) * 1/ (1-M (i))
Wherein, after C (i) is for prediction virtual machine (vm) migration to target physical main frame to be selected, the CPU of target physical main frame to be selected
Utilization rate；M (i) is the internal memory for being migrated virtual machine i；Ni values are smaller, show that the balanced utilization rate of CPU and internal memory is low, and performance is got over
It is good.
It will be apparent to those skilled in the art that technical scheme that can be as described above and design, make other various
It is corresponding to change and deformation, and all these change and deformation should all belong to the protection domain of the claims in the present invention
Within.
Claims (6)
1. a kind of elastic calculation dynamic allocation method, it is characterised in that comprise the following steps：
Step 1, when detecting system load and being unbalance, select pack heaviest a physical host as pending physics master
Machine；
Step 2, the virtual machine selected on the pending physical host, and the virtual machine (vm) migration that will be selected is to target physical main frame
On, specially obtain the VCPU performance datas of all virtual machines on the pending physical host, and by the VCPU performance datas
Descending arrangement is carried out, by the maximum virtual machine (vm) migration of VCPU utilization rates to a minimum physical host of CPU usage, then
The virtual machine on the pending physical host is migrated successively according to the arrangement of VCPU performance datas；
Step 3, when in above-mentioned steps 1 pending physical host load less than threshold values or without virtual function migrate out when,
Then select next load more than the physical host of threshold values as candidate physical main frame, and return to step 1.
2. elastic calculation dynamic allocation method according to claim 1, it is characterised in that load is selected in above-mentioned steps 1
Used as pending physical host, the load using greedy algorithm to physical host is calculated a most heavy physical host.
3. elastic calculation dynamic allocation method according to claim 2, it is characterised in that system load in above-mentioned steps 1
Unbalance, a physical host for selecting pack heaviest is as pending physical host：
For the physical host in the resource pool for enabling resource elasticity allotment, when performance indications exceed threshold value, according to physics master
The index of the historical performance assessment resource of machine calculates average value, specially：Setting configuration interface in triggered time at intervals of
T, history average utilization index is calculated using following equation：
AVG (used)=A × Wa+B × Wb+C × Wc,
Wherein, A is that, beyond threshold time section average value, B is the average value in T time before threshold time section, C be beyond
Average value in 24+T time before threshold time section, Wa, Wb, Wc is respectively A, the corresponding weighted value of B, C.
4. elastic calculation dynamic allocation method according to claim 3, it is characterised in that select this to treat in the step 2
Virtual machine on treatment physical host, specially：The virtual machine is considered to the utilization rate of CPU and the occupancy situation of internal memory,
If CPU and internal memory exceed threshold values, preferential to calculate CPU, while considering internal memory；If only internal memory exceedes threshold values, only
Calculate internal memory.
5. elastic calculation dynamic allocation method according to claim 4, it is characterised in that utilization rate of the virtual machine to CPU
Computational methods are as follows：
The proportion that virtual machine i accounts for physical host CPU is：
Wherein, C (t) is CPU average utilization of the physical host in the t time periods, and W is the weight of current virtual machine, and N is current empty
The CPU numbers of plan machine, W (i) is the weight of virtual machine i, and N (i) is the CPU numbers of virtual machine i, and j is the operation on physical host
The number of virtual machine；
After virtual machine i migrations, physical host CPU usage is changed into：
C (c) n+1=C (c) n-C (i),
Wherein, C (c) n are the value of the CPU usage of n-th change, as n=0, C (c) n=C (t).
6. elastic calculation dynamic allocation method according to claim 5, it is characterised in that the choosing of the target physical main frame
Select and specifically selected using focus equation of equilibrium,
The focus equation of equilibrium is：
Ni=1/ (1-C (c)) * 1/ (1-M (i)),
Wherein, after C (i) is for prediction virtual machine (vm) migration to target physical main frame to be selected, the CPU of target physical main frame to be selected is utilized
Rate；C (c) is the value of the CPU usage of change；M (i) is the internal memory for being migrated virtual machine i；Ni values are smaller, show target physical
The CPU of main frame and the balanced utilization rate of internal memory are low, and performance is better.
Priority Applications (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN201611264538.9A CN106681839B (en) | 2016-12-31 | 2016-12-31 | Elastic calculation dynamic allocation method |
Applications Claiming Priority (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN201611264538.9A CN106681839B (en) | 2016-12-31 | 2016-12-31 | Elastic calculation dynamic allocation method |
Publications (2)
| Publication Number | Publication Date |
|---|---|
| CN106681839A true CN106681839A (en) | 2017-05-17 |
| CN106681839B CN106681839B (en) | 2021-06-15 |
Family
ID=58849549
Family Applications (1)
| Application Number | Title | Priority Date | Filing Date |
|---|---|---|---|
| CN201611264538.9A Active CN106681839B (en) | 2016-12-31 | 2016-12-31 | Elastic calculation dynamic allocation method |
Country Status (1)
| Country | Link |
|---|---|
| CN (1) | CN106681839B (en) |
Cited By (6)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN108388471A (en) * | 2018-01-31 | 2018-08-10 | 山东汇贸电子口岸有限公司 | A kind of management method constraining empty machine migration based on double threshold |
| CN111427673A (en) * | 2020-03-16 | 2020-07-17 | 杭州迪普科技股份有限公司 | Load balancing method, device and equipment |
| CN112579246A (en) * | 2019-09-27 | 2021-03-30 | 阿里巴巴集团控股有限公司 | Virtual machine migration processing method and device |
| CN113032098A (en) * | 2021-03-25 | 2021-06-25 | 深信服科技股份有限公司 | Virtual machine scheduling method, device, equipment and readable storage medium |
| CN114221962A (en) * | 2021-12-09 | 2022-03-22 | 兴业银行股份有限公司 | Cloud resource redistribution method and system based on peak utilization rate |
| CN115878329A (en) * | 2023-02-02 | 2023-03-31 | 天翼云科技有限公司 | Host resource scheduling method and device, electronic equipment and storage medium |
Citations (4)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN103473139A (en) * | 2013-09-26 | 2013-12-25 | 四川中电启明星信息技术有限公司 | Virtual machine cluster resource allocation and scheduling method |
| CN103617076A (en) * | 2013-10-31 | 2014-03-05 | 中兴通讯股份有限公司 | Method and system for dispatching virtualized resources and server |
| US20150052287A1 (en) * | 2013-08-13 | 2015-02-19 | Vmware, Inc. | NUMA Scheduling Using Inter-vCPU Memory Access Estimation |
| CN105740074A (en) * | 2016-01-26 | 2016-07-06 | 中标软件有限公司 | Cloud computing based virtual machine load balancing method |
-
2016
- 2016-12-31 CN CN201611264538.9A patent/CN106681839B/en active Active
Patent Citations (4)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| US20150052287A1 (en) * | 2013-08-13 | 2015-02-19 | Vmware, Inc. | NUMA Scheduling Using Inter-vCPU Memory Access Estimation |
| CN103473139A (en) * | 2013-09-26 | 2013-12-25 | 四川中电启明星信息技术有限公司 | Virtual machine cluster resource allocation and scheduling method |
| CN103617076A (en) * | 2013-10-31 | 2014-03-05 | 中兴通讯股份有限公司 | Method and system for dispatching virtualized resources and server |
| CN105740074A (en) * | 2016-01-26 | 2016-07-06 | 中标软件有限公司 | Cloud computing based virtual machine load balancing method |
Non-Patent Citations (1)
| Title |
|---|
| 袁秋华: "基于OpenStack云平台的数字化校园设计与实现", 《CNKI》 * |
Cited By (10)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN108388471A (en) * | 2018-01-31 | 2018-08-10 | 山东汇贸电子口岸有限公司 | A kind of management method constraining empty machine migration based on double threshold |
| CN108388471B (en) * | 2018-01-31 | 2021-02-09 | 浪潮云信息技术股份公司 | Management method based on double-threshold constraint virtual machine migration |
| CN112579246A (en) * | 2019-09-27 | 2021-03-30 | 阿里巴巴集团控股有限公司 | Virtual machine migration processing method and device |
| CN111427673A (en) * | 2020-03-16 | 2020-07-17 | 杭州迪普科技股份有限公司 | Load balancing method, device and equipment |
| CN111427673B (en) * | 2020-03-16 | 2023-04-07 | 杭州迪普科技股份有限公司 | Load balancing method, device and equipment |
| CN113032098A (en) * | 2021-03-25 | 2021-06-25 | 深信服科技股份有限公司 | Virtual machine scheduling method, device, equipment and readable storage medium |
| CN113032098B (en) * | 2021-03-25 | 2024-04-09 | 深信服科技股份有限公司 | Virtual machine scheduling method, device, equipment and readable storage medium |
| CN114221962A (en) * | 2021-12-09 | 2022-03-22 | 兴业银行股份有限公司 | Cloud resource redistribution method and system based on peak utilization rate |
| CN114221962B (en) * | 2021-12-09 | 2024-02-13 | 兴业银行股份有限公司 | Cloud resource reallocation method and system based on peak utilization rate |
| CN115878329A (en) * | 2023-02-02 | 2023-03-31 | 天翼云科技有限公司 | Host resource scheduling method and device, electronic equipment and storage medium |
Also Published As
| Publication number | Publication date |
|---|---|
| CN106681839B (en) | 2021-06-15 |
Similar Documents
| Publication | Publication Date | Title |
|---|---|---|
| CN106681839A (en) | Elasticity calculation dynamic allocation method | |
| CN102232282B (en) | Method and apparatus for realizing load balance of resources in data center | |
| CN104102543B (en) | The method and apparatus of adjustment of load in a kind of cloud computing environment | |
| Jung et al. | Synchronous parallel processing of big-data analytics services to optimize performance in federated clouds | |
| CN104375897B (en) | Cloud computing resource scheduling method based on the unbalanced degree of minimum relative load | |
| CN114138486B (en) | Method, system and medium for arranging containerized micro-services for cloud edge heterogeneous environment | |
| US11496413B2 (en) | Allocating cloud computing resources in a cloud computing environment based on user predictability | |
| Tsai et al. | Two-tier multi-tenancy scaling and load balancing | |
| CN111078363A (en) | NUMA node scheduling method, device, equipment and medium for virtual machine | |
| CN104598316B (en) | A kind of storage resource distribution method and device | |
| CN103095599A (en) | Dynamic feedback weighted integration load scheduling method of cloud computing operating system | |
| CN104008018B (en) | The online moving method of virtual machine under cloud computing environment | |
| Mittal et al. | Real time contingency analysis for power grids | |
| CN104199739A (en) | Speculation type Hadoop scheduling method based on load balancing | |
| CN111880939A (en) | Container dynamic migration method and device and electronic equipment | |
| CN105491150A (en) | Load balance processing method based on time sequence and system | |
| CN103488538B (en) | Application extension device and application extension method in cloud computing system | |
| CN112463395A (en) | Resource allocation method, device, equipment and readable storage medium | |
| Keerthika et al. | A multiconstrained grid scheduling algorithm with load balancing and fault tolerance | |
| CN109992392A (en) | A kind of calculation resource disposition method, device and Resource Server | |
| CN105740077A (en) | Task assigning method applicable to cloud computing | |
| Guo | Ant colony optimization computing resource allocation algorithm based on cloud computing environment | |
| Lin et al. | A workload-driven approach to dynamic data balancing in MongoDB | |
| Chen et al. | Evaluation of counter-based dynamic load balancing schemes for massive contingency analysis on over 10,000 cores | |
| CN114237902A (en) | Service deployment method and device, electronic equipment and computer readable medium |
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 |