CN110099083A - A kind of load equilibration scheduling method and device for server cluster - Google Patents
A kind of load equilibration scheduling method and device for server cluster Download PDFInfo
- Publication number
- CN110099083A CN110099083A CN201810089039.3A CN201810089039A CN110099083A CN 110099083 A CN110099083 A CN 110099083A CN 201810089039 A CN201810089039 A CN 201810089039A CN 110099083 A CN110099083 A CN 110099083A
- Authority
- CN
- China
- Prior art keywords
- cluster
- server
- task
- load
- load balance
- 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
-
- H—ELECTRICITY
- H04—ELECTRIC COMMUNICATION TECHNIQUE
- H04L—TRANSMISSION OF DIGITAL INFORMATION, e.g. TELEGRAPHIC COMMUNICATION
- H04L67/00—Network arrangements or protocols for supporting network services or applications
- H04L67/01—Protocols
- H04L67/10—Protocols in which an application is distributed across nodes in the network
- H04L67/1001—Protocols in which an application is distributed across nodes in the network for accessing one among a plurality of replicated servers
- H04L67/1004—Server selection for load balancing
- H04L67/1008—Server selection for load balancing based on parameters of servers, e.g. available memory or workload
Landscapes
- Engineering & Computer Science (AREA)
- Computer Hardware Design (AREA)
- General Engineering & Computer Science (AREA)
- Computer Networks & Wireless Communication (AREA)
- Signal Processing (AREA)
- Computer And Data Communications (AREA)
Abstract
The invention discloses a kind of load equilibration scheduling methods and device for server cluster.This method includes: step 1: when server cluster carries out load balance scheduling for the first time, carrying out load balance scheduling using the first load balance scheduling algorithm；Step 2: after server cluster carries out load balance scheduling for the first time, the overall operation state of server cluster and the individual operating status of each cluster server are obtained, is recycled based on overall operation state and individual operating status and has been distributed but being not carried out of the task and redistributed.More intelligentized load balance scheduling may be implemented in this method, solves the problems, such as to dispatch for the first time unbalanced；Server resource can more reasonably be utilized；The flexibility and controllability of load balancing can be increased.
Description
Technical field
The present invention relates to content distributing network field more particularly to a kind of load balance scheduling sides for server cluster
Method and device.
Background technique
With the continuous development of computer network and communication network, number of users constantly increases, customer access network service
The request of device sharply increases.In order to guarantee to provide normal network service for mass users, there has been proposed server cluster systems
System and distributed server cluster system, to guarantee the parallel processing capability and redundancy ability and task point of identical services server
Class processing capacity.Key technology in server cluster system is for guaranteeing to carry out task balance scheduling between servers
Load balancing (scheduling) technology.
In the prior art, when between each cluster server for carrying out execute server cluster internal using load balancer
Load balance scheduling task when, usually only with certain specific conventional load equalization algorithm (such as Hash, poll, weighting take turns
Ask scheduling algorithm), but these conventional load equalization algorithms are often only focused on and how to be allocated to task, and cannot be according to current collection
The actual conditions of group, which are adaptively adjusted allocated task, (cannot especially redistribute the allocated mistake but have not been performed
Task), can not be scheduled in conjunction with algorithms of different, therefore, it is very easy to cause the final result of load imbalance.
Least for solving the above problems, need to propose new technical solution.
Summary of the invention
The present invention provides a kind of load equilibration scheduling methods for server cluster, comprising:
Step 1: server cluster carry out load balance scheduling for the first time when, using the first load balance scheduling algorithm come into
Row load balance scheduling；
Step 2: after server cluster carries out load balance scheduling for the first time, obtaining the overall operation state of server cluster
With the individual operating status of each cluster server, is recycled and distributed but not based on overall operation state and individual operating status
The task of execution is simultaneously redistributed.
According to the above method of the present invention, step 2 includes:
Step 2-1: in the individual operating status for getting cluster server be high load and overall operation state is low negative
When load, recycles having distributed but being not carried out for the task of the cluster server and reassigned to different cluster servers.
According to the above method of the present invention, in its step 2-1, task is realized using the second load balance scheduling algorithm
Redistribute.
According to the above method of the present invention, step 2 includes:
Step 2-2: in the individual operating status for getting cluster server be low-load and overall operation state is low negative
When load, will be recycled from the cluster server and distributed it is into the task of different cluster servers, distributed but be not carried out
Task recycle again and be reassigned to the cluster server.
According to the above method of the present invention, in step 2, the overall operation of server cluster is obtained according to following information
The individual operating status of state and each cluster server:
The basic information of each cluster server, the basic information include cpu utilization rate, payload size, are handling
Number of tasks.
According to the above method of the present invention, in step 2, the individual of each cluster server is obtained according to following strategy
Operating status:
It is scored by the consuming capacity that following formula weighted sum calculates each cluster server: w1*cpu utilization rate+
W2* payload size+w3* is handling number of tasks, then recognizes when the consuming capacity scoring of cluster server reaches given threshold or more
High load condition is in for cluster server, wherein w1, w2, w3 are to correspond respectively to cpu utilization rate, payload size, locating
Manage the weighted value of number of tasks；
Calculate each cluster server at the appointed time in section, the increment of number of tasks that is being handled in the unit time,
When increment continues to increase, then it is assumed that cluster server is in high load condition.
The present invention also provides a kind of load balance scheduling devices for server cluster, comprising:
Load balance scheduling module for the first time is used for when server cluster carries out load balance scheduling for the first time, using first
Load balance scheduling algorithm carries out load balance scheduling；
Subsequent load balance scheduling module, for obtaining service after server cluster carries out load balance scheduling for the first time
The overall operation state of device cluster and the individual operating status of each cluster server, based on overall operation state and running body
State has been distributed but being not carried out of the task and has been redistributed to recycle.
Above-mentioned apparatus according to the present invention, subsequent load balance scheduling module include:
First task reallocation module, for being high load and totality in the individual operating status for getting cluster server
When operating status is low-load, recycles having distributed but being not carried out for the task of the cluster server and reassigned to difference
Cluster server.
Above-mentioned apparatus according to the present invention, subsequent load balance scheduling module further include:
Second task reallocation module, for being low-load and totality in the individual operating status for getting cluster server
Operating status be low-load when, will from the cluster server recycle and distributed it is into the task of different cluster servers,
It has distributed but being not carried out for task recycles again and is reassigned to the cluster server.
The present invention also provides another load balance scheduling devices for being used for server cluster, including memory, processing
On a memory and the computer program that can run on a processor, when processor execution program, is realized described above for device and storage
Method the step of.
Above-mentioned technical proposal according to the present invention may be implemented more intelligentized load balance scheduling, solve for the first time
Dispatch unbalanced problem；More reasonably server resource can be utilized (for example, hot spot task list is shared in peak period
It loads on low machine, the task list called away is rescheduled reduction by trough period)；The flexible of load balancing can be increased
Property and controllability.
Detailed description of the invention
It is incorporated into specification and the attached drawing for constituting part of specification shows the embodiment of the present invention, and with
Relevant verbal description principle for explaining the present invention together.In the drawings, similar appended drawing reference is for indicating class
As element.Drawings in the following description are some embodiments of the invention, rather than whole embodiments.It is common for this field
For technical staff, without creative efforts, other drawings may be obtained according to these drawings without any creative labor.
Fig. 1 schematically illustrates the signal stream of the load equilibration scheduling method according to the present invention for server cluster
Cheng Tu.
Fig. 2 schematically illustrates the schematic block of the load balance scheduling device according to the present invention for server cluster
Figure.
Fig. 3 schematically illustrates the schematic block diagram that single machine task status according to the present invention obtains module.
Specific embodiment
In order to make the object, technical scheme and advantages of the embodiment of the invention clearer, below in conjunction with the embodiment of the present invention
In attached drawing, technical scheme in the embodiment of the invention is clearly and completely described, it is clear that described embodiment is
A part of the embodiments of the present invention, instead of all the embodiments.Based on the embodiments of the present invention, ordinary skill people
Member's every other embodiment obtained without making creative work, shall fall within the protection scope of the present invention.It needs
It is noted that in the absence of conflict, the features in the embodiments and the embodiments of the present application can mutual any combination.
Fig. 1 schematically illustrates showing for the load equilibration scheduling method 100 according to the present invention for server cluster
Meaning flow chart.
As shown in Figure 1, for server cluster load equilibration scheduling method 100 the following steps are included:
Step S102: when server cluster carries out load balance scheduling for the first time, using the first load balance scheduling algorithm
To carry out load balance scheduling；
Step S104: after server cluster carries out load balance scheduling for the first time, the overall operation of server cluster is obtained
The individual operating status (that is, state of individual server or machine) of state (that is, cluster state) and each cluster server, base
It has distributed but being not carried out of the task and has redistributed to recycle in overall operation state and individual operating status.
For example, center (scheduling) server being specially arranged can be used in such a way that active inquiry or reception report
Come obtain server (for example, Edge Server) cluster overall operation state (that is, cluster state) and each cluster server
Individual operating status (that is, state of individual server or machine).Optionally, when reporting mode in use, each cluster clothes
Business device regularly can report respective individual operating status to center (scheduling) server.Central (scheduling) server is according to each
The individual operating status (including high load, stabilization, low-load) of a cluster server obtains the overall operation of server cluster
State (including high load, stabilization, low-load).For example, when the server for being more than 1/4 is in high load condition, then it is assumed that its
The overall operation state of affiliated server cluster is in high load condition.
For example, each cluster server can determine in the following manner the individual operating status of itself and regularly to
Central (scheduling) server reports:
(operation) of single machine (that is, individual server) can be judged by (for example, customized task overstocks model)
State: high load, low-load, stable state.When incoming task number and output number of tasks within a period of time (for example, 1 hour) not
Timing, the number of tasks in buffer area can be fluctuated constantly, when incoming task number is greater than output number of tasks, indicate that single machine consumes energy
Power (that is, processing capacity) is insufficient, and the number of tasks in buffer area can increase, and single machine is in high load condition at this time；Work as incoming task
When number is less than output number of tasks, indicate that single machine consuming capacity is sufficient, the number of tasks decline in buffer area, single machine is in low negative at this time
Load state；When incoming task number is equal to output number of tasks, single machine enters stable state.Wherein, the representative of incoming task number newly enters
Number of tasks, output number of tasks represent the number of tasks of release, and the number of tasks in buffer area represents the number of tasks handled.
According to the above method of the present invention, it is loaded using the adjustment mechanism of negative-feedback, based on machine (that is, cluster server)
Come dynamically judge single machine task processing state, the adaptive operation of scheduling method is realized by automatically controlling.And it is formed
The negative feedback control of closed loop, can acquire machine loading, and dynamically adjust the input of each machine by feeding back, to reach
To the purpose for the load for controlling each machine.
For example, each cluster server can be determined in the following manner and itself distributed but being not carried out of the task and fixed
Phase Xiang Zhongyang (scheduling) server report:
The task queue that is being handled by (for example, schedule element module of backlog) for self buffer and
Itself task handles log (including historic task handling duration) and carries out integrated decision-making, and obtains itself need to rescheduling for task
List.
It optionally, can also by (for example, single machine quality obtain module) monitoring speed of download alarm log and periodically
Detection speed of download comes Comprehensive Evaluation single machine quality state, the important reference indicator as above-mentioned integrated decision-making.
It is alternatively possible to (for example, passing through single machine information management module) storage basis (related with above-mentioned integrated decision-making)
Information.Basic information include: task input state, task output state, buffer state, speed of download, scheduler task list,
Result of decision etc..
For example, central (scheduling) server can determine the overall operation state of server cluster in the following manner:
As described above, the client (application program) on server (for example, Edge Server) can be by itself decision
As a result and part basis information reporting is to central schedule server.Central schedule server can integrate all Edge Servers
Information carries out final decision.For example, if consumption (processing) ability of whole group (that is, the server cluster) is also inadequate
(for example, being more than 1/4 machine loading height), then without scheduling, if the consuming capacity of whole group is sufficient, central schedule server
Scheduler task list can be reported into load balancing according to the load ranking and backlog ranking integrated decision-making of Edge Server
The scheduling that device is oriented.
Optionally, step S104 includes:
Step 2-1 (is not shown in Fig. 1): being high load and totality in the individual operating status for getting cluster server
When operating status is low-load, recycles having distributed but being not carried out for the task of the cluster server and reassigned to difference
Cluster server.
Optionally, in step 2-1, redistributing for task is realized using the second load balance scheduling algorithm.
Optionally, step S104 includes:
Step 2-2 (is not shown in Fig. 1): being low-load and totality in the individual operating status for getting cluster server
Operating status be low-load when, will from the cluster server recycle and distributed it is into the task of different cluster servers,
It has distributed but being not carried out for task recycles again and is reassigned to the cluster server.
Optionally, in step S104, the overall operation state of server cluster and each is obtained according to following information
The individual operating status of cluster server:
The basic information of each cluster server, the basic information include cpu utilization rate, payload size, are handling
Number of tasks.
Optionally, in step S104, the individual operating status of each cluster server is obtained according to following strategy:
It is scored by the consuming capacity that following formula weighted sum calculates each cluster server: w1*cpu utilization rate+
W2* payload size+w3* is handling number of tasks, then recognizes when the consuming capacity scoring of cluster server reaches given threshold or more
High load condition is in for cluster server, wherein w1, w2, w3 are to correspond respectively to cpu utilization rate, payload size, locating
Manage the weighted value of number of tasks；
Calculate each cluster server at the appointed time in section (for example, 5 minutes), in the unit time (for example, 1 minute)
The increment of the number of tasks handled, when increment continues to increase (for example, more than 50/minute), then it is assumed that cluster server
In high load condition.
By the above method, (for example, center machine or center cluster server) carries out synthesis according to both the above strategy and determines
Plan obtains corresponding newest scheduling instruction, and (that is, newest scheduling path, Main Design Principles are by the task schedule of high load machine
To the machine of low-load), then by scheduling Centralized path notification to every machine in cluster server.
Optionally, the individual machine in cluster server can actively be initiated to target machine according to the scheduling path that decision goes out
The request for obtaining the also untreated task of target machine, is added in the task queue of oneself and handles.
At this point, (for example, center machine) has been distributed without recycling but being not carried out of the task and has been redistributed.This scheme avoids
(for example, center machine) needs to dispatch bring expense again, reduces the processing logic of (for example, center machine), improves and appoints
The efficiency of business scheduling.
Optionally, in step 2-2, redistributing for task is realized using the first load balance scheduling algorithm.
Optionally, the first load balance scheduling algorithm and the second load balance scheduling algorithm are selected from the collection including following algorithm
It closes: Hash, poll, weighted polling.
For example, the first load balance scheduling algorithm is hash algorithm, the second load balance scheduling algorithm be polling algorithm or
Weighted Round Robin.
Fig. 2 schematically illustrates showing for the load balance scheduling device 200 according to the present invention for server cluster
Meaning block diagram.
As shown in Fig. 2, the load balance scheduling device 200 for server cluster includes:
Load balance scheduling module 201 for the first time, for when server cluster carries out load balance scheduling for the first time, using the
One load balance scheduling algorithm carries out load balance scheduling；
Subsequent load balance scheduling module 203, for obtaining clothes after server cluster carries out load balance scheduling for the first time
The overall operation state of device cluster of being engaged in and the individual operating status of each cluster server, based on overall operation state and individual fortune
Row state has been distributed but being not carried out of the task and has been redistributed to recycle.
For example, subsequent load balance scheduling module 203 can obtain the overall fortune of server cluster according to following information
The individual operating status of row state and each cluster server:
Basic information (including the cpu utilization rate u, payload size load, just of each cluster server (that is, every machine)
In processing number of tasks req).
For example, the operating status of (that is, decision) each cluster server can be obtained parallel according to following two strategy:
Strategy one: (for example, in center machine) subsequent load balance scheduling module 203 passes through following formula weighted sum meter
Calculate the consuming capacity scoring of every machine (that is, cluster server): score=w1*u+w2*load+w3*req, (weighted value
W1, w2, w3 are set according to cluster actual conditions), then think that the machine is in height when scoring reaches given threshold or more
Load condition.
Strategy two: (for example, in center machine) subsequent load balance scheduling module 203 is being handled according to every machine appoints
The Long-term change trend situation of business number carries out decision, if task processing quantity is constantly in rising situation (for example, 5 in a period of time
Minute, the average rate of climb was greater than 50/minute), then it is assumed that machine consuming capacity is insufficient, is in high load condition.
Through the above technical solutions, (for example, on center machine or center cluster server) subsequent load balance scheduling mould
Block 203 carries out integrated decision-making according to both the above strategy and obtains newest scheduling instruction (that is, newest scheduling path, mainly sets
Counting principle is by the machine of the task schedule of high load machine to low-load), then by newest scheduling instruction notice to cluster
In every machine.
Optionally, the machine in cluster can indicate actively to initiate to obtain target machine to target machine according to newest scheduling
The request of also untreated task, is added in the task queue of oneself and is handled.
At this point, subsequent load balance scheduling module 203 has been distributed without recycling but being not carried out of the task and has been redistributed.This
Kind scheme avoids (for example, in center machine) load balance scheduling module 203 and needs to dispatch bring expense again, reduces
The processing logic of (for example, in center machine) subsequent load balance scheduling module 203, improves the efficiency of task schedule.
Optionally, subsequent load balance scheduling module 203 includes:
First task reallocation module (is not shown) in Fig. 2, in a running body shape for getting cluster server
When state is high load and overall operation state is low-load, recycles having distributed but being not carried out for the task of the cluster server and incite somebody to action
It reassigns to different cluster servers.
Optionally, subsequent load balance scheduling module 203 further include:
Second task reallocation module (being not shown in Fig. 2), in a running body shape for getting cluster server
When state is low-load and overall operation state is low-load, it will recycle and distributed to different cluster clothes from the cluster server
Task in the task of device, having distributed but be not carried out of being engaged in recycles again and is reassigned to the cluster server.
Optionally, first task reallocation module and/or the second task reallocation module include:
Single machine task status obtains module, for obtaining the individual running state information of each cluster server.
Fig. 3 schematically illustrates the schematic block diagram that single machine task status according to the present invention obtains module.
As shown in figure 3, it includes performance estimation module, schedule element (module), quality prison that single machine task status, which obtains module,
Control (module), machine information management (module).The input parameter (being not shown in Fig. 3) of its performance estimation module includes: (each
Cluster server) newly establish number of tasks, the number of tasks that number of tasks is completed, is handling, task history log, speed of download
Index etc..
Single machine task status obtains module can determine the individual fortune of each cluster server according to above-mentioned input parameter
Row status information.For example, the operating status of cluster server may include: high load, low-load, stable state.When incoming task number
When mismatching within a period of time (for example, 1 hour) with output number of tasks, the number of tasks in buffer area can be fluctuated constantly, when
When incoming task number is greater than output number of tasks, the number of tasks that single machine consuming capacity (that is, processing capacity) is insufficient, in buffer area is indicated
It can increase, single machine is in high load condition at this time；When incoming task number is less than output number of tasks, indicate that single machine consuming capacity is filled
Sufficient, the number of tasks decline in buffer area, single machine is in low-load state at this time；When incoming task number is equal to output number of tasks,
Single machine enters stable state.Wherein, incoming task number represents the number of tasks newly entered, and output number of tasks represents the number of tasks of release, delays
The number of tasks rushed in area represents the number of tasks handled.
Optionally, single machine task status obtains module and can also be arranged on each cluster server, each cluster clothes
Being engaged in device can be regularly to first task reallocation module and/or the respective a running body shape of the second task reallocation module feedback
State information.
Optionally, first task reallocation module and/or the second task reallocation module further include: intelligent control module.
Intelligent control module is used for, and calculates the deviation of (each cluster server) present load and reference load, if
The deviation of present load and reference load is excessive, and will automatically adjust (relevant parameter for changing intelligent control module), single
Machine task status obtains module and carries out Real-time Decision according to the output result of intelligent control module and input parameter state, is appointed
The related scheduling of business, achievees the purpose that dynamic control (each cluster server) output loading.
Load balance scheduling device (being not shown in the accompanying drawings) the present invention also provides another kind for server cluster,
Including memory, processor and the computer program that can be run on a memory and on a processor is stored, processor executes institute
When stating program the step of implementation method 100.
Above-mentioned technical proposal according to the present invention can be adaptively adjusted according to the actual conditions of current cluster and divide
Matching but having not been performed for task, additionally it is possible to be scheduled in conjunction with the advantage of algorithms of different, therefore, it is equal obtain preferably load
Weigh result.
Above-mentioned technical proposal according to the present invention may be implemented more intelligentized load balance scheduling, solve for the first time
Dispatch unbalanced problem；More reasonably server resource can be utilized (for example, hot spot task list is shared in peak period
It loads on low machine, the task list called away is rescheduled reduction by trough period)；The flexible of load balancing can be increased
Property and controllability.
Although describing above-mentioned technical proposal of the invention by taking common server group system as an example, however, of the invention
Above-mentioned technical proposal is also applied for distributed server cluster system.
Descriptions above can combine implementation individually or in various ways, and these variants all exist
Within protection scope of the present invention.
Finally, it should be noted that the above embodiments are merely illustrative of the technical solutions of the present invention, rather than its limitations.Although
Present invention has been described in detail with reference to the aforementioned embodiments, those skilled in the art should understand that: it still may be used
To modify the technical solutions described in the foregoing embodiments or equivalent replacement of some of the technical features；
And these are modified or replaceed, the spirit of the technical solution for various embodiments of the present invention that it does not separate the essence of the corresponding technical solution
And range.
Claims (10)
1. a kind of load equilibration scheduling method for server cluster, which is characterized in that the described method includes:
Step 1: the server cluster carry out load balance scheduling for the first time when, using the first load balance scheduling algorithm come into
Row load balance scheduling；
Step 2: after the server cluster carries out load balance scheduling for the first time, obtaining the overall operation of the server cluster
The individual operating status of state and each cluster server is recycled based on the overall operation state and individual operating status
It distributes but being not carried out of the task and redistributes.
2. the method as described in claim 1, which is characterized in that the step 2 includes:
Step 2-1: in the individual operating status for getting cluster server be high load and the overall operation state is low negative
When load, recycles having distributed but being not carried out for the task of the cluster server and reassigned to different cluster servers.
3. method according to claim 2, which is characterized in that in the step 2-1, calculated using the second load balance scheduling
Method realizes redistributing for task.
4. method as claimed in claim 3, which is characterized in that the step 2 includes:
Step 2-2: in the individual operating status for getting cluster server be low-load and the overall operation state is low negative
When load, will be recycled from the cluster server and distributed it is into the task of different cluster servers, distributed but be not carried out
Task recycle again and be reassigned to the cluster server.
5. the method as described in claim 1, which is characterized in that in the step 2, the clothes are obtained according to following information
The overall operation state of device cluster of being engaged in and the individual operating status of each cluster server:
The basic information of each cluster server, the basic information include cpu utilization rate, payload size, are handling task
Number.
6. method as claimed in claim 5, which is characterized in that in the step 2, each collection is obtained according to following strategy
The individual operating status of group's server:
The consuming capacity scoring of each cluster server is calculated by following formula weighted sum: w1*cpu utilization rate+w2* is negative
It carries size+w3* and is handling number of tasks, then think cluster when the consuming capacity scoring of cluster server reaches given threshold or more
Server is in high load condition, wherein w1, w2, w3 are to correspond respectively to cpu utilization rate, payload size, handling task
Several weighted values；
Calculate each cluster server at the appointed time in section, the increment of number of tasks that is being handled in the unit time, work as increasing
When amount continues to increase, then it is assumed that cluster server is in high load condition.
7. a kind of load balance scheduling device for server cluster, which is characterized in that described device includes:
Load balance scheduling module for the first time is used for when the server cluster carries out load balance scheduling for the first time, using first
Load balance scheduling algorithm carries out load balance scheduling；
Subsequent load balance scheduling module is used for after the server cluster carries out load balance scheduling for the first time, described in acquisition
The overall operation state of server cluster and the individual operating status of each cluster server, based on the overall operation state and
Individual operating status has been distributed but being not carried out of the task and has been redistributed to recycle.
8. device as claimed in claim 7, which is characterized in that the subsequent load balance scheduling module includes:
First task reallocation module, for being high load and the totality in the individual operating status for getting cluster server
When operating status is low-load, recycles having distributed but being not carried out for the task of the cluster server and reassigned to difference
Cluster server.
9. device as claimed in claim 8, which is characterized in that the subsequent load balance scheduling module further include:
Second task reallocation module, for being low-load and the totality in the individual operating status for getting cluster server
Operating status be low-load when, will from the cluster server recycle and distributed it is into the task of different cluster servers,
It has distributed but being not carried out for task recycles again and is reassigned to the cluster server.
10. a kind of load balance scheduling device for server cluster, including memory, processor and it is stored in the storage
On device and the computer program that can run on the processor, which is characterized in that the processor executes real when described program
The step of existing method described in any one of claims 1 to 6.
Priority Applications (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN201810089039.3A CN110099083A (en) | 2018-01-30 | 2018-01-30 | A kind of load equilibration scheduling method and device for server cluster |
Applications Claiming Priority (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN201810089039.3A CN110099083A (en) | 2018-01-30 | 2018-01-30 | A kind of load equilibration scheduling method and device for server cluster |
Publications (1)
| Publication Number | Publication Date |
|---|---|
| CN110099083A true CN110099083A (en) | 2019-08-06 |
Family
ID=67442714
Family Applications (1)
| Application Number | Title | Priority Date | Filing Date |
|---|---|---|---|
| CN201810089039.3A Pending CN110099083A (en) | 2018-01-30 | 2018-01-30 | A kind of load equilibration scheduling method and device for server cluster |
Country Status (1)
| Country | Link |
|---|---|
| CN (1) | CN110099083A (en) |
Cited By (5)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN111082972A (en) * | 2019-11-26 | 2020-04-28 | 北京杰思安全科技有限公司 | Method for realizing elastic expansion based on distributed cluster and distributed cluster architecture |
| CN112395058A (en) * | 2020-11-13 | 2021-02-23 | 北京数码视讯技术有限公司 | Task regulation and control device, method and system |
| CN113485847A (en) * | 2021-08-05 | 2021-10-08 | 杭州绿城信息技术有限公司 | Resource scheduling system based on big data |
| CN113778627A (en) * | 2021-09-10 | 2021-12-10 | 龙坤(无锡)智慧科技有限公司 | Scheduling method for creating cloud resources |
| CN114726860A (en) * | 2022-03-21 | 2022-07-08 | 慧之安信息技术股份有限公司 | Load balancing system and load balancing method for streaming media transmission |
Citations (6)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN103401939A (en) * | 2013-08-08 | 2013-11-20 | 中国航天科工集团第三研究院第八三五七研究所 | Load balancing method adopting mixing scheduling strategy |
| US20140307553A1 (en) * | 2013-04-13 | 2014-10-16 | Hei Tao Fung | Network traffic load balancing |
| CN105007312A (en) * | 2015-07-03 | 2015-10-28 | 叶秀兰 | Method and system for controlling adaptive load-balancing of cloud computing server |
| CN105159240A (en) * | 2015-07-23 | 2015-12-16 | 上海极熵数据科技有限公司 | Job scheduling system of automatic industrial apparatuses |
| CN105376317A (en) * | 2015-11-19 | 2016-03-02 | 网宿科技股份有限公司 | Load balancing control method and load balancing control device |
| US20170214737A1 (en) * | 2016-01-25 | 2017-07-27 | Vmware, Inc. | Integrated application-aware load balancer incorporated within a distributed-service-application-controlled distributed computer system |
-
2018
- 2018-01-30 CN CN201810089039.3A patent/CN110099083A/en active Pending
Patent Citations (6)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| US20140307553A1 (en) * | 2013-04-13 | 2014-10-16 | Hei Tao Fung | Network traffic load balancing |
| CN103401939A (en) * | 2013-08-08 | 2013-11-20 | 中国航天科工集团第三研究院第八三五七研究所 | Load balancing method adopting mixing scheduling strategy |
| CN105007312A (en) * | 2015-07-03 | 2015-10-28 | 叶秀兰 | Method and system for controlling adaptive load-balancing of cloud computing server |
| CN105159240A (en) * | 2015-07-23 | 2015-12-16 | 上海极熵数据科技有限公司 | Job scheduling system of automatic industrial apparatuses |
| CN105376317A (en) * | 2015-11-19 | 2016-03-02 | 网宿科技股份有限公司 | Load balancing control method and load balancing control device |
| US20170214737A1 (en) * | 2016-01-25 | 2017-07-27 | Vmware, Inc. | Integrated application-aware load balancer incorporated within a distributed-service-application-controlled distributed computer system |
Cited By (8)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN111082972A (en) * | 2019-11-26 | 2020-04-28 | 北京杰思安全科技有限公司 | Method for realizing elastic expansion based on distributed cluster and distributed cluster architecture |
| CN111082972B (en) * | 2019-11-26 | 2022-08-05 | 北京杰思安全科技有限公司 | Method for realizing elastic expansion based on distributed cluster and distributed cluster architecture |
| CN112395058A (en) * | 2020-11-13 | 2021-02-23 | 北京数码视讯技术有限公司 | Task regulation and control device, method and system |
| CN113485847A (en) * | 2021-08-05 | 2021-10-08 | 杭州绿城信息技术有限公司 | Resource scheduling system based on big data |
| CN113778627A (en) * | 2021-09-10 | 2021-12-10 | 龙坤(无锡)智慧科技有限公司 | Scheduling method for creating cloud resources |
| CN113778627B (en) * | 2021-09-10 | 2024-03-12 | 龙坤(无锡)智慧科技有限公司 | Scheduling method for creating cloud resources |
| CN114726860A (en) * | 2022-03-21 | 2022-07-08 | 慧之安信息技术股份有限公司 | Load balancing system and load balancing method for streaming media transmission |
| CN114726860B (en) * | 2022-03-21 | 2022-10-21 | 慧之安信息技术股份有限公司 | Load balancing system and load balancing method for streaming media transmission |
Similar Documents
| Publication | Publication Date | Title |
|---|---|---|
| KR102562260B1 (en) | Commitment-aware scheduler | |
| CN110099083A (en) | A kind of load equilibration scheduling method and device for server cluster | |
| Yang et al. | Joint computation partitioning and resource allocation for latency sensitive applications in mobile edge clouds | |
| US8997107B2 (en) | Elastic scaling for cloud-hosted batch applications | |
| CN102681889B (en) | Scheduling method of cloud computing open platform | |
| Hoang et al. | FBRC: Optimization of task scheduling in fog-based region and cloud | |
| US9386086B2 (en) | Dynamic scaling for multi-tiered distributed systems using payoff optimization of application classes | |
| WO2016082370A1 (en) | Distributed node intra-group task scheduling method and system | |
| CN109120715A (en) | Dynamic load balancing method under a kind of cloud environment | |
| CN103838621B (en) | Method and system for scheduling routine work and scheduling nodes | |
| US20100218005A1 (en) | Energy-aware server management | |
| CN104168332A (en) | Load balance and node state monitoring method in high performance computing | |
| CN103220337A (en) | Cloud computing resource optimizing collocation method based on self-adaptation elastic control | |
| US11496413B2 (en) | Allocating cloud computing resources in a cloud computing environment based on user predictability | |
| CN103064744B (en) | The method for optimizing resources that a kind of oriented multilayer Web based on SLA applies | |
| Ahn et al. | Competitive partial computation offloading for maximizing energy efficiency in mobile cloud computing | |
| CN109861850B (en) | SLA-based stateless cloud workflow load balancing scheduling method | |
| CN105939225A (en) | Method and device for executing service | |
| Mogouie et al. | A novel approach for optimization auto-scaling in cloud computing environment | |
| CN114938372B (en) | Federal learning-based micro-grid group request dynamic migration scheduling method and device | |
| CN109815009A (en) | Scheduling of resource and optimization method under a kind of CSP | |
| CN107122235B (en) | Public infrastructure resource scheduling method based on application priority | |
| CN112862319B (en) | Order scheduling method, device, equipment and storage medium | |
| Patel et al. | An improved approach for load balancing among heterogeneous resources in computational grids | |
| Khalifa et al. | MobiCloud: A reliable collaborative mobilecloud management 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 | ||
| RJ01 | Rejection of invention patent application after publication |
Application publication date: 20190806 |
|
| RJ01 | Rejection of invention patent application after publication |