CN109308193A - A kind of multi-tenant function calculates the concurrency control method of service - Google Patents
A kind of multi-tenant function calculates the concurrency control method of service Download PDFInfo
- Publication number
- CN109308193A CN109308193A CN201811035589.3A CN201811035589A CN109308193A CN 109308193 A CN109308193 A CN 109308193A CN 201811035589 A CN201811035589 A CN 201811035589A CN 109308193 A CN109308193 A CN 109308193A
- Authority
- CN
- China
- Prior art keywords
- request
- tenant
- function
- time
- calculate node
- 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
- G06F9/30—Arrangements for executing machine instructions, e.g. instruction decode
- G06F9/38—Concurrent instruction execution, e.g. pipeline, look ahead
- G06F9/3818—Decoding for concurrent execution
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06F—ELECTRIC DIGITAL DATA PROCESSING
- G06F9/00—Arrangements for program control, e.g. control units
- G06F9/06—Arrangements for program control, e.g. control units using stored programs, i.e. using an internal store of processing equipment to receive or retain programs
- G06F9/46—Multiprogramming arrangements
- G06F9/50—Allocation of resources, e.g. of the central processing unit [CPU]
- G06F9/5005—Allocation of resources, e.g. of the central processing unit [CPU] to service a request
Abstract
The invention discloses the concurrency control methods that a kind of multi-tenant function calculates service to be identified request when including the following steps: the request for the operation function that calculate node tenant in receiving cloud platform sends;Con current control service cluster runs quantity to first respectively according to the request after mark and the second operation quantity carries out add-one operation, and runs quantity to second after the first operation quantity and operation after calculate node return operation;First operation quantity of the calculate node after determining operation is less than the first amount threshold, and after operation second operation quantity less than the second amount threshold when, corresponding function is run according to request, and returns to operation result to corresponding tenant, realizes that function calculates the con current control of service.The operational efficiency that function can be effectively improved using concurrency control method of the invention, the function for being precisely controlled each tenant's distribution under synchronization calculate Service Source equilibrium.
Description
Technical field
The present invention relates to the con current control sides that field of cloud computer technology more particularly to a kind of multi-tenant function calculate service
Method.
Background technique
It is a kind of based on event driven function trusteeship service that function, which calculates service,.The logic that it can allow user that will write
Trustship is to cloud host in the form of function for code, so that the scheduling in the request for the operation function for receiving tenant's transmission of cloud host
Calculate node runs function, and returns to operation result to tenant;Wherein, calculate node is deployed in calculation server.
As shown in Figure 1, being the model structure schematic diagram that function calculates service in the prior art.In the model structure, meter
It calculates server and scheduler is deployed on cloud host, when scheduler receives the request of tenant's transmission, scheduler is according to pre-
If scheduling strategy forward the request to calculate node.In order to realize the available computational resources of each tenant by can be by setting in advance
Fixed size is allocated, existing calculating service be usually provided on scheduler RPS (Request Per Second, often
Second receive request) number of requests threshold value, and then each tenant function numbers per second that can newly run can be limited.However, due to
The operation of function is a time lasting process, and the duration is different in size, though the function of new operation per second is by asking
It asks amount threshold to be limited, but since the time of operation function release computing resource is not fixed, tenant is caused to occupy
Computing resource may be more than preset size.Existing method can not accurately realize con current control, so that user experience
It is poor.
Summary of the invention
In view of the above-mentioned problems, a kind of multi-tenant function of the invention calculates the concurrency control method of service, can effectively improve
The operational efficiency of function, the function for being precisely controlled each tenant's distribution under synchronization calculate Service Source equilibrium, improve user
Usage experience.
In order to solve the above technical problems, a kind of multi-tenant function of the invention calculates the concurrency control method of service, including
Following steps:
When the request for the operation function that calculate node tenant in receiving cloud platform sends, the request is identified
And the request after identifying is sent to con current control service cluster;
The con current control service cluster is according to the request after the mark respectively to the first operation quantity and the second operation
Quantity carries out add-one operation, and runs number to second after the first operation quantity and operation after calculate node return operation
Amount;The first operation quantity is the function operation quantity of tenant, and the second operation quantity is the fortune of the function of tenant request
Row sum, the function of the tenant request is by tenant's trustship into cloud platform;
First operation quantity of the calculate node after determining operation is less than the first amount threshold, and second after operation
When running quantity less than the second amount threshold, the function of the tenant request is run according to the request, and to corresponding tenant
Operation result is returned, realizes that function calculates the con current control of service.
Compared with prior art, in concurrency control method of the invention, calculate node tenant in receiving cloud platform is sent out
When the request of the operation function sent, con current control is sent to after request progress unique identification, then by the request after mark
Service cluster;Con current control service cluster is according to the request after mark respectively to the function operation quantity of tenant and tenant request
The operation sum of function carries out add-one operation, so that the function of function operation quantity and tenant request that each tenant is currently running
Operation sum be all recorded in con current control service in, enable calculate node after determining operation first run quantity it is small
In the first amount threshold, and when the second operation quantity is less than the second amount threshold, just runs and run the request of function to corresponding
Function.It is preset with the first amount threshold due to being directed to each tenant in calculate node, and is preset with second for function
Amount threshold, the function of the function operation quantity and the tenant request that can run to each tenant in synchronization is in synchronization
Operation sum limited, can be precisely controlled under synchronization to each tenant carry out function calculate service resource allocation, make
It obtains function and calculates service resource allocation equilibrium, improve user experience.
As an improvement of the above scheme, the request includes time-out time;
The con current control service cluster is according to the request after the mark respectively to the first operation quantity and the second operation
Further include following steps before quantity carries out add-one operation:
The con current control service cluster receives the request after the mark, and the sequence ascending according to time-out time
Request after the mark is stored in request queue.
As an improvement of the above scheme, when calculate node runs the function of the tenant request according to the request,
Further include following steps:
The time-out time is updated according to the preset update cycle, and after sending to the con current control service cluster and updating
Time-out time;The update cycle includes multiple judgement periods;
Work as when the con current control service cluster determines that the updated time-out time is greater than within any judgement period
When the preceding time, carry out subtracting 1 operation to first number of calculations and the second operation quantity respectively, and from the request queue
The middle request deleted after the mark.
As an improvement of the above scheme, the con current control service cluster judges described updated super as follows
When the time whether be greater than current time:
Within each judgement period, successively judge in the request queue identify after request time-out time whether
Greater than the current time;
When the time-out time for determining the request after any mark is greater than the current time, any mark is determined
The time-out time of request of the request afterwards into the request queue after the last one mark is all larger than current time.
As an improvement of the above scheme, the request includes tenant ID and function ID;
It is right as follows when the request for the operation function that calculate node tenant in receiving cloud platform sends
The request is identified:
The tenant ID, the function ID and the time-out time are obtained, and generates the mark operation request at random
Request ID;
The request ID, the tenant ID, the function ID and the time-out time are encapsulated as marking by the calculate node
Request after knowledge.
As an improvement of the above scheme, the con current control service cluster is provided with add-one operation interface and more new interface;
When the request for the operation function that calculate node tenant in receiving cloud platform sends, pass through described plus 1 fortune
Calculate the request after interface sends mark to the con current control service cluster;
Alternatively, after the calculate node sends the update to the con current control service cluster by the more new interface
Time-out time.
As an improvement of the above scheme, the con current control service cluster is provided with the operation interface that subtracts 1;
The calculate node is after returning to operation result to corresponding tenant, by the operation interface that subtracts 1 to described
The transmission of con current control service cluster subtracts operational order, so that the con current control service cluster subtracts operational order according to, point
It is other to first number of calculations and the second operation quantity carries out subtracting 1 operation, and described in deleting from the request queue
Request after mark.
As an improvement of the above scheme, the con current control cluster using RAFT agreement as distributed consensus agreement,
To provide con current control service.
As an improvement of the above scheme, the operation function that tenant sends in receiving cloud platform in the calculate node
Further include following steps before request:
When scheduler receives the request, asked according to preset scheduling mode to described in the forwarding of corresponding calculate node
It asks.
As an improvement of the above scheme, the scheduling mode includes:
The request received is distributed in turn to corresponding calculate node according to preset calculate node sequencing;
Alternatively, the request received is distributed in turn according to the size order of preset calculate node weight score
To corresponding calculate node.
Detailed description of the invention
Fig. 1 is the model structure schematic diagram that function calculates service in the prior art.
Fig. 2 is the model structure schematic diagram that function calculates service in the present invention.
Fig. 3 is the flow diagram that a kind of multi-tenant function of the embodiment of the present invention 1 calculates the concurrency control method of service.
Fig. 4 is the flow diagram that a kind of multi-tenant function of the embodiment of the present invention 2 calculates the concurrency control method of service.
Specific embodiment
In the following description, numerous specific details are set forth in order to facilitate a full understanding of the present invention.But the present invention can be with
It is different from the other modes of this description much to implement, those skilled in the art can be without violating the connotation of the present invention
Similar popularization is done, therefore the present invention is not limited by the specific embodiments disclosed below.
As shown in Fig. 2, being the model structure schematic diagram that function calculates service in the present invention.
In the present invention, con current control service cluster is made of at least two con current control servers, con current control clothes
Business cluster uses RAFT (Replicated And Fault Tolerant) agreement as distributed consensus agreement to provide simultaneously
Hair control service, avoids con current control service of the invention from Single Point of Faliure problem occur.Calculate node can be to con current control service
Any con current control server in cluster is communicated.
Clear, complete description is carried out to technical solution of the present invention with attached drawing combined with specific embodiments below.
Embodiment 1
As shown in figure 3, being the process that a kind of multi-tenant function of the embodiment of the present invention 1 calculates the concurrency control method of service
Schematic diagram.
A kind of concurrency control method of multi-tenant function calculating service, includes the following steps:
When the request for the operation function that S1, the calculate node tenant in receiving cloud platform send, which is marked
Know, and sends the request after identifying to con current control service cluster;
Wherein, the request for the operation function that tenant sends includes tenant ID, function ID and time-out time, in step sl,
Calculate node is as follows identified the request:
S11, tenant ID, function ID and time-out time are obtained, and generates the request ID of mark operation request at random;
S12, ID, tenant ID, function ID and time-out time will be requested to be encapsulated as the request after identifying.
S2, con current control service cluster according to the request after mark respectively to first operation quantity and second operation quantity into
Row add-one operation, and quantity is run to second after the first operation quantity and operation after calculate node return operation;First operation
Quantity is the function operation quantity of tenant, and the second operation quantity is the operation sum of the function of tenant request, the letter of tenant request
Number is by tenant's trustship into cloud platform;
The the first operation quantity of S3, calculate node after determining operation is less than the first amount threshold, and second after operation
When running quantity less than the second amount threshold, the function of tenant request is run according to the request, and return and transport to corresponding tenant
Row is as a result, realize that function calculates the con current control of service.
Below with reference to a specific example, the concurrency control method of embodiment 1 is illustrated.
The function of an entitled Function is created in cloud platform in tenant A, and sets the first quantity threshold of tenant A
Value is in the case that the 100, second amount threshold is 10, and when tenant A sends the request of operation Function, and calculate node exists
When receiving the request, unique request ID is identified to the request, and send the request after identifying to con current control service cluster;
Con current control service cluster carries out add-one operation to the first operation quantity and the second operation quantity respectively according to the request after mark,
And quantity is run to second after the first operation quantity and operation after calculate node transmission operation;Calculate node is determining operation
The first operation quantity afterwards runs Function, and to tenant A less than 100, and when the second operation quantity after operation is less than 10
Return to corresponding operation result.
Compared with prior art, in concurrency control method of the invention, calculate node tenant in receiving cloud platform is sent out
When the request of the operation function sent, con current control is sent to after request progress unique identification, then by the request after mark
Service cluster;Con current control service cluster is according to the request after mark respectively to the function operation quantity of tenant and the operation of function
Sum carries out add-one operation, so that the operation sum of the function of function operation quantity and tenant request that each tenant is currently running
It is all recorded in con current control service, enables first operation quantity of the calculate node after determining operation less than the first quantity
Threshold value, and when the second operation quantity is less than the second amount threshold, just operation with run the corresponding function of the request of function.Due to
It is preset with the first amount threshold for each tenant in calculate node, and is preset with the second number for the function of tenant request
Threshold value is measured, the fortune of the function of the function operation quantity and tenant request that can run to each tenant in synchronization in synchronization
Row sum is limited, and can be precisely controlled under synchronization and be carried out function calculating service resource allocation to each tenant, so that letter
Number calculates service resource allocation equilibrium, improves user experience.
Embodiment 2
As shown in figure 4, being the process that a kind of multi-tenant function of the embodiment of the present invention 2 calculates the concurrency control method of service
Schematic diagram.
The concurrency control method is in addition to including in embodiment 1 other than Overall Steps, in con current control service cluster according to mark
Further include following steps before request after knowledge carries out add-one operation to the first operation quantity and the second operation quantity respectively:
S21, con current control service cluster receive the request after mark, and will mark according to the ascending sequence of time-out time
Request after knowledge is stored in request queue;
Wherein, in the step s 21, con current control service cluster can deposit the request after mark according to following form
Storage:
In step s3, further include following steps when calculate node runs corresponding function according to request:
S31, time-out time is updated according to the preset update cycle, and send updated surpass to con current control service cluster
When the time;Update cycle includes multiple judgement periods;
In step S31, the update cycle be may be configured as 5 seconds, judge that the period may be configured as 1 second, the time of update cycle is long
Degree should be greater than judging the time span in period.
S32, when con current control service cluster determine within any judgement period updated time-out time be greater than it is current when
Between when, carry out subtracting 1 operation, and asking after deleting mark in request queue to the first number of calculations and the second operation quantity respectively
It asks;
Wherein, in step s 32, whether con current control service cluster judges updated time-out time as follows
Greater than current time:
S321, within each judgement period, successively judge in request queue identify after request time-out time it is whether big
In current time;
S322, the time-out time for determining the request after any mark be greater than current time when, after determining any mark
The time-out time of request of the request into request queue after the last one mark be all larger than current time.
S33, when con current control service cluster determine within any judgement period updated time-out time be less than it is current when
Between when, then update corresponding time-out time in request queue.
In step S33, con current control service cluster updates corresponding time-out time in request queue as follows:
S331, the request after obtaining corresponding mark in request queue;
S332, by updated time-out time replacement mark after request in time-out time.
In the concurrency control method of embodiment 2, calculate node updates time-out time, con current control clothes according to the update cycle
Business cluster according to judge the request of period real-time judge whether time-out, and in the case where any judgement period decision request time-out,
Carry out subtracting 1 operation, and the request after deleting mark in request queue to the first number of calculations and the second operation quantity respectively.By
The time-out time checking mechanism of request operation is established in calculate node and con current control service cluster, so that calculate node is instant
Go offline, delay machine or in the case where calculating can not be provided, can accurate Tuning function calculate service.
Further, in the above embodiment of the invention, con current control service cluster is provided with add-one operation interface and more
New interface;
Calculate node in receiving cloud platform tenant send operation function request when, by add-one operation interface to
Con current control service cluster sends the request after mark;
Alternatively, calculate node sends updated time-out time to con current control service cluster by more new interface.
Further, in the above embodiment of the invention, con current control service cluster is provided with the operation interface that subtracts 1;
Calculate node is after returning to operation result to corresponding tenant, by the operation interface that subtracts 1 to con current control service
Collection pocket transmission subtracts operational order, so that con current control service cluster is according to operational order is subtracted, respectively to the first number of calculations and the
Two operation quantity carry out subtracting 1 operation, and the request after deleting mark in request queue.
Preferably, in the above embodiment of the invention, calculate node is obtaining tenant ID, function ID and time-out time,
Further include following steps and after generating the request ID of mark operation request at random:
Request ID, tenant ID, function ID and time-out time are resolved into call parameters, and then by request ID as tune
With the mode of parameter call respectively the add-one operation interface of con current control service cluster, more new interface or the operation interface that subtracts 1 come real
Now communication between the two.
Preferably, before the request for the operation function that tenant sends in receiving cloud platform in calculate node, further include
Following steps:
When scheduler receives request, which is forwarded to corresponding calculate node according to preset scheduling mode.
Wherein, scheduling mode include: according to preset calculate node sequencing by the request received distribute in turn to
Corresponding calculate node;
Alternatively, according to the size order of preset calculate node weight score, by the request received by weight wheel flow point
It is assigned to corresponding calculate node.
The above described is only a preferred embodiment of the present invention, limitation in any form not is done to the present invention, therefore
All contents without departing from technical solution of the present invention, it is made to the above embodiment according to the technical essence of the invention any simply to repair
Change, equivalent variations and modification, all of which are still within the scope of the technical scheme of the invention.
Claims (10)
1. the concurrency control method that a kind of multi-tenant function calculates service, which comprises the steps of:
Calculate node in receiving cloud platform tenant send operation function request when, to it is described request be identified and to
Con current control service cluster sends the request after mark;
The con current control service cluster is according to the request after the mark respectively to the first operation quantity and the second operation quantity
Add-one operation is carried out, and runs quantity to second after the first operation quantity and operation after calculate node return operation;Institute
The function operation quantity that the first operation quantity is tenant is stated, the second operation quantity is that the operation of the function of tenant request is total
Number, the function of the tenant request is by tenant's trustship into cloud platform;
First operation quantity of the calculate node after determining operation is less than the first amount threshold, and the second operation after operation
When quantity is less than the second amount threshold, the function of the tenant request is run according to the request, and return to corresponding tenant
Operation result realizes that function calculates the con current control of service.
2. concurrency control method as described in claim 1, which is characterized in that the request includes time-out time;
The con current control service cluster is according to the request after the mark respectively to the first operation quantity and the second operation quantity
Further include following steps before carrying out add-one operation:
The con current control service cluster receives the request after the mark, and according to the ascending sequence of time-out time by institute
Request after stating mark is stored in request queue.
3. concurrency control method as claimed in claim 2, which is characterized in that in calculate node according to request operation
Further include following steps when the function of tenant request:
The time-out time is updated according to the preset update cycle, and sends updated surpass to the con current control service cluster
When the time;The update cycle includes multiple judgement periods;
When the con current control service cluster determines that the updated time-out time is greater than current within any judgement period
Between when, carry out subtracting 1 operation to first number of calculations and the second operation quantity respectively, and delete from the request queue
Except the request after the mark.
4. concurrency control method as claimed in claim 3, which is characterized in that the con current control service cluster by walking as follows
Suddenly judge whether the updated time-out time is greater than current time:
Within each judgement period, successively judge whether the time-out time of the request after identifying in the request queue is greater than
The current time;
When the time-out time for determining the request after any mark is greater than the current time, after determining any mark
The time-out time of request after requesting into the request queue the last one mark is all larger than current time.
5. concurrency control method as claimed in claim 3, which is characterized in that the request includes tenant ID and function ID;
When the request for the operation function that calculate node tenant in receiving cloud platform sends, as follows to described
Request is identified:
The tenant ID, the function ID and the time-out time are obtained, and generates the request of the mark operation request at random
ID;
After the request ID, the tenant ID, the function ID and the time-out time are encapsulated as mark by the calculate node
Request.
6. concurrency control method as claimed in claim 3, which is characterized in that the con current control service cluster, which is provided with, adds 1
Operation interface and more new interface;
When the request for the operation function that calculate node tenant in receiving cloud platform sends, connect by the add-one operation
Mouth sends the request after identifying to the con current control service cluster;
Alternatively, the calculate node is described updated super to con current control service cluster transmission by the more new interface
When the time.
7. concurrency control method as claimed in claim 3, which is characterized in that the con current control service cluster, which is provided with, subtracts 1
Operation interface;
The calculate node is after returning to operation result to corresponding tenant, by the operation interface that subtracts 1 to described concurrent
Control service cluster transmission subtracts operational order, so that the con current control service cluster subtracts operational order according to, it is right respectively
First number of calculations and the second operation quantity carry out subtracting 1 operation, and the mark is deleted from the request queue
Request afterwards.
8. concurrency control method as described in claim 1, which is characterized in that the con current control cluster is made using RAFT agreement
For distributed consensus agreement, to provide con current control service.
9. concurrency control method as described in claim 1, which is characterized in that in the calculate node in receiving cloud platform
Further include following steps before the request for the operation function that tenant sends:
When scheduler receives the request, the request is forwarded to corresponding calculate node according to preset scheduling mode.
10. concurrency control method as claimed in claim 9, which is characterized in that the scheduling mode includes:
The request received is distributed in turn to corresponding calculate node according to preset calculate node sequencing;
Alternatively, according to the size order of preset calculate node weight score, by the request received by weight wheel flow point
It is assigned to corresponding calculate node.
Priority Applications (1)
Application Number | Priority Date | Filing Date | Title |
---|---|---|---|
CN201811035589.3A CN109308193B (en) | 2018-09-06 | 2018-09-06 | A kind of multi-tenant function calculates the concurrency control method of service |
Applications Claiming Priority (1)
Application Number | Priority Date | Filing Date | Title |
---|---|---|---|
CN201811035589.3A CN109308193B (en) | 2018-09-06 | 2018-09-06 | A kind of multi-tenant function calculates the concurrency control method of service |
Publications (2)
Publication Number | Publication Date |
---|---|
CN109308193A true CN109308193A (en) | 2019-02-05 |
CN109308193B CN109308193B (en) | 2019-08-09 |
Family
ID=65224481
Family Applications (1)
Application Number | Title | Priority Date | Filing Date |
---|---|---|---|
CN201811035589.3A Active CN109308193B (en) | 2018-09-06 | 2018-09-06 | A kind of multi-tenant function calculates the concurrency control method of service |
Country Status (1)
Country | Link |
---|---|
CN (1) | CN109308193B (en) |
Citations (7)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
WO2012000850A1 (en) * | 2010-06-29 | 2012-01-05 | International Business Machines Corporation | Allocating computer resources in a network data processing system |
US20130111491A1 (en) * | 2011-10-28 | 2013-05-02 | Daniel A. Gerrity | Entitlement vector with resource and/or capabilities fields |
CN103218175A (en) * | 2013-04-01 | 2013-07-24 | 无锡成电科大科技发展有限公司 | Multi-tenant cloud storage platform access control system |
CN104734946A (en) * | 2015-04-09 | 2015-06-24 | 北京易掌云峰科技有限公司 | Multi-tenant high-concurrency instant messaging cloud platform |
CN104850450A (en) * | 2015-05-14 | 2015-08-19 | 华中科技大学 | Load balancing method and system facing mixed cloud application |
CN104881325A (en) * | 2015-05-05 | 2015-09-02 | 中国联合网络通信集团有限公司 | Resource scheduling method and resource scheduling system |
CN105512264A (en) * | 2015-12-04 | 2016-04-20 | 贵州大学 | Performance prediction method of concurrency working loads in distributed database |
-
2018
- 2018-09-06 CN CN201811035589.3A patent/CN109308193B/en active Active
Patent Citations (7)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
WO2012000850A1 (en) * | 2010-06-29 | 2012-01-05 | International Business Machines Corporation | Allocating computer resources in a network data processing system |
US20130111491A1 (en) * | 2011-10-28 | 2013-05-02 | Daniel A. Gerrity | Entitlement vector with resource and/or capabilities fields |
CN103218175A (en) * | 2013-04-01 | 2013-07-24 | 无锡成电科大科技发展有限公司 | Multi-tenant cloud storage platform access control system |
CN104734946A (en) * | 2015-04-09 | 2015-06-24 | 北京易掌云峰科技有限公司 | Multi-tenant high-concurrency instant messaging cloud platform |
CN104881325A (en) * | 2015-05-05 | 2015-09-02 | 中国联合网络通信集团有限公司 | Resource scheduling method and resource scheduling system |
CN104850450A (en) * | 2015-05-14 | 2015-08-19 | 华中科技大学 | Load balancing method and system facing mixed cloud application |
CN105512264A (en) * | 2015-12-04 | 2016-04-20 | 贵州大学 | Performance prediction method of concurrency working loads in distributed database |
Also Published As
Publication number | Publication date |
---|---|
CN109308193B (en) | 2019-08-09 |
Similar Documents
Publication | Publication Date | Title |
---|---|---|
CN104994123B (en) | A kind of traffic scheduling method of CDN cloud platforms and CDN cloud platforms | |
US6986139B1 (en) | Load balancing method and system based on estimated elongation rates | |
CN105242956B (en) | Virtual functions service chaining deployment system and its dispositions method | |
CN109246229A (en) | A kind of method and apparatus of distribution resource acquisition request | |
CN104579761B (en) | A kind of nosql clusters automatic configuration system and method for automatic configuration based on cloud computing | |
CN104008015B (en) | Control device and resource control method | |
WO2018095066A1 (en) | Method and device for task grouping, electronic device, and computer storage medium | |
CN108776612A (en) | A kind of cloud computing method for allocating tasks, device, equipment and storage medium | |
CN102724103B (en) | Proxy server, hierarchical network system and distributed workload management method | |
CN107968836A (en) | A kind of task distribution method and device | |
US8458379B2 (en) | Information processing program, method, and transfer processing device | |
CN111818181B (en) | Data processing method and device based on block chain and computer readable storage medium | |
CN110795217B (en) | Task allocation method and system based on resource management platform | |
CN112217725B (en) | Delay optimization method based on edge calculation | |
CN105407162A (en) | Cloud computing Web application resource load balancing algorithm based on SLA service grade | |
CN109697637A (en) | Object type determines method, apparatus, electronic equipment and computer storage medium | |
Celaya et al. | A fair decentralized scheduler for bag-of-tasks applications on desktop grids | |
KR20200062887A (en) | Apparatus and method for assuring quality of control operations of a system based on reinforcement learning. | |
WO2015019538A1 (en) | Virtual resource control system and virtual resource control method | |
CN109308193B (en) | A kind of multi-tenant function calculates the concurrency control method of service | |
CN109657801A (en) | Shunt method, device and the readable storage medium storing program for executing of recommender system | |
Shefu et al. | Fruit fly optimization algorithm for network-aware web service composition in the cloud | |
CN110191362B (en) | Data transmission method and device, storage medium and electronic equipment | |
CN105138391B (en) | The multitasking virtual machine distribution method of cloud system justice is distributed towards wide area | |
CN110096337A (en) | A kind of enterprise applies the cloud data center resource allocation methods and system of cloud deployment |
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 |