CN101719865A - Data processing method for centralized login of a plurality of users - Google Patents
Data processing method for centralized login of a plurality of users Download PDFInfo
- Publication number
- CN101719865A CN101719865A CN200910198696A CN200910198696A CN101719865A CN 101719865 A CN101719865 A CN 101719865A CN 200910198696 A CN200910198696 A CN 200910198696A CN 200910198696 A CN200910198696 A CN 200910198696A CN 101719865 A CN101719865 A CN 101719865A
- Authority
- CN
- China
- Prior art keywords
- application server
- server
- request
- allotting
- resource allotting
- 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
Images
Landscapes
- Computer And Data Communications (AREA)
Abstract
The invention discloses a data processing method for centralized login of a plurality of users. Each application server periodically reports the current self service load to a resource allotting server, and the resource allotting server generates a resource allotting list according to service loads reported by all the application servers; the resource allotting server calculates the time for processing a login request by each application server according to the resource allotting list and generates a time list; the first application server receives the login request of a client end connected with the first application server and reports the login request to the resource allotting server; the resource allotting server receives the login request, looks up the time list to search a second application server with shortest time for processing login, transmits the login request to the second application server to process and feeds back a processing result to the resource allotting server; the resource allotting server feeds back the processing result to the first application server, and the processing result is fed back to the client end by the first application server.
Description
Technical field
The present invention relates to network technology, relate in particular to the data processing method that user's a large number of users is concentrated login.
Background technology
Large-scale computer network system has a large amount of server that is distributed in everywhere, and usually, these servers can independent running, the service request of responsible one's respective area separately.When the regional area a large number of users concurrent operations in the system, can cause the operand of certain or certain server to increase sharply, thereby reduce the efficient that single user obtains to serve.As: when up to a hundred users carry out register simultaneously, application server that this is regional and database server will be born the computing of up to a hundred logon process in the same period, comprise data search, results set combination, display page generation etc.With respect to tens users' concurrent operations, response efficiency can reduce an order of magnitude.So the system service ability the when large user measures concurrent operations becomes the bottleneck of system's operation.
But in fact, when the certain server operand rolled up, other servers in the network did not increase a lot of operands, and their service ability also has the suitable free time.Because a large number of users concurrent operations occurs in regional area, therefore caused the imbalance of offered load, over-burden for some servers, other are then relatively more idle, if can give full play to the operational capability of these idle servers, can improve the response efficiency of system and the efficiency of service that the user obtains to a great extent.
Summary of the invention
The invention provides a kind of data processing method when concentrating login, can shunt login request of users effectively, improve whole efficient at a large number of users.
According to the present invention, a kind of data processing method that a large number of users is concentrated login that is used for is proposed, be used for a computer network, this computer network comprises one group of application server, an one group of client and a resource allotting server, wherein each client in one group of client logs on the application server request application service that it connects, and this group of client and one group of application server all are connected to resource allotting server, and this method comprises:
Each application server periodically reports current self business load to resource allotting server, resource allotting server generates resource allocation table according to the business load that all application servers report, wherein, resource allocation table periodically upgrades according to the business load that application server newly reports;
Resource allotting server is calculated each application server processes according to resource allocation table and is landed the needed time of request, the rise time table, and timetable is updated periodically according to the renewal of resource allocation table;
First application server receives the request of landing of its client that connects, and this request of landing is reported resource allotting server;
Resource allotting server receives this land request after, search timetable, seek second application server, second application server has processing and lands the needed shortest time, the resource allotting server request of will landing sends to second application server and handles, and the result of processing feeds back to resource allotting server;
The result that resource allotting server will be handled feeds back to first application server, feeds back to client by first application server.
Wherein, if second application server is exactly first application server itself, then resource allotting server sends a message to first application server, and first application server is directly handled this request of landing and fed back to client.
Perhaps, resource allotting server is periodically to current self the business load of each application server queries application server, and the business load that obtains according to inquiry generates resource allocation table.
Adopt technical scheme of the present invention, can in time new service request be assigned on the application server the most idle in the system and handle, but for client, remain by original application server and feed back therefore, new service request can access and respond the most in time and serve, client can not feel inconvenient yet simultaneously, makes the whole performance of system obtain effective raising.
Description of drawings
Fig. 1 has disclosed the flow chart of a kind of data processing method when concentrating login at a large number of users of the present invention.
Embodiment
The following design concept of method design of the present invention:
Load balancing: load balancing is based upon on the existing network infrastructure, a large amount of concurrent visits or data traffic can be shared processing respectively on many node server equipment, reduces the time of user's wait-for-response; Simultaneously, single heavy duty computing shared on many node server equipment do parallel processing, each node server device processes after finishing gathers the result, returns to the user, and system processing power is greatly improved.Adopt load-balancing technique, to different servers, the service that each server only is responsible for certain customers provides with the volume of services reasonable distribution, reduces the pressure of heavy duty server.Thereby make the user can at utmost obtain quick fastest response, simultaneously from the angle of server, can improve server stability, prevent system exception.
Gridding computing: the various resources in the network (computer, database, software, other system etc.) are connected to form integral body, from client, whole network is just as same huge incomparable computer, provide incorporate service to each user, because the computer software and hardware in all networks all participates in computing, so can reduce load of central server greatly, can possess service function and the performance that provides powerful simultaneously.Adopt the gridding computing, on different servers and other computers, each client and server all may be born the computational load of part with the volume of services reasonable distribution, reduces the pressure of heavy duty server.Thereby make the user can at utmost obtain quick fastest response, simultaneously from the angle of server, also can improve server stability, prevent system exception.
With reference to shown in Figure 1, disclosed according to a kind of data processing method that a large number of users is concentrated login that is used for of the present invention.Need to prove that said " login " contained generalized concept herein, comprise the user for the visit of the page, for the request of certain service, call certain interface or the like.Method of the present invention is used for a computer network, this computer network comprises one group of application server, an one group of client and a resource allotting server, wherein each client in one group of client logs on the application server request application service that it connects, one group of client and one group of application server all are connected to resource allotting server, and this method comprises:
10. each application server periodically reports current self business load to resource allotting server, resource allotting server generates resource allocation table according to the business load that all application servers report, wherein, resource allocation table periodically upgrades according to the business load that application server newly reports.Perhaps step 10 also can be can't help each application server and periodically reported current self business load to resource allotting server, but resource allotting server is periodically to current self the business load of each application server queries application server, and the business load that obtains according to inquiry generates resource allocation table.The purpose of two kinds of implementations is the same, all is in order to obtain the load Distribution situation of up-to-date application server, and resource allocation table periodically upgrades according to the new data that obtain.
11. resource allotting server is calculated each application server processes according to resource allocation table and is landed the needed time of request, the rise time table, and timetable is updated periodically according to the renewal of resource allocation table.Shown in the schematic exemplary reference table 1 of timetable:
Table 1
Application server | Load (standard %) | Processing time (unit interval) |
Application server 1 | ??156% | ??379 |
Application server 2 | ??101% | ??103 |
Application server 3 | ??77% | ??45 |
Application server 4 | ??120% | ??172 |
Application server 5 | ??98% | ??94 |
Wherein, resource allotting server can be provided with a standard termination, and current business load capacity and this standard negative carrying capacity of each application server of being obtained compared, and draws a percentage, second hurdle in the reference table 1.Further, land the processing time of the required cost of request according to one of each application server processes of this percentage calculation, same, processing time also is to be standard with the unit interval, and setting one of application server processes with standard negative carrying capacity, to land the needed time of request be 100 unit interval.Need to prove that in different application, the processing time of different loads amount correspondence also can be different, the processing time herein is just as an example.Resource allotting server is concerned about is not processing time of each application service implement body can what be, and only is under current load, and which application server has the shortest processing time, and which has the longest processing time.Therefore, how the concrete mode the present invention according to standard negative carrying capacity calculating treatmenting time does not limit, and any account form all is feasible, as long as this account form can objectively respond the relation of load size and processing time length.Perhaps, timetable will upgrade, and timetable can be updated periodically according to the renewal of resource allocation table.
12. first application server receives the request of landing of its client that connects, and this request of landing is reported resource allotting server.Suppose that the current application server that newly receives the logging request of client is first application server, need to prove that first application server is just narrated and the title of appointment for convenience.
13. resource allotting server receives this land request after, search timetable, seek second application server, second application server has processing and lands the needed shortest time, the resource allotting server request of will landing sends to second application server and handles, and the result of processing feeds back to resource allotting server.Equally, second application server also is to narrate and the title of appointment for convenience, and second application server refers to the server that has SPT in timetable.For example, in table 1, second application server is exactly an application server 3.
14. the result that resource allotting server will be handled feeds back to first application server, feeds back to client by first application server.For example, application server 1 has received new logging request, and application server 1 is exactly first application server.Actual logging request resource allotting server can be handed to application server 3 (second application server) and handle, after finishing dealing with, the result that application server 3 will be handled feeds back to resource allotting server, resource allotting server feeds back to the result application server 1 again, feeds back to client by application server 1.Like this, for client, it finally also is to be fed back by application server 1 to the logging request that application server 1 proposes, and client can not felt any inconvenient part.
A kind of situation is also arranged, and application server 3 has received new logging request, and at this moment, application server 3 is first application servers.After through the inquiry of resource allotting server to timetable, second application server also is an application server 3.At this moment, second application server is exactly first application server itself, and then resource allotting server directly sends a message to first application server (application server 3), notifies it directly to handle.First application server (application server 3) is directly handled this request of landing and is fed back to client.
Adopt technical scheme of the present invention, can in time new service request be assigned on the application server the most idle in the system and handle, but for client, remain by original application server and feed back therefore, new service request can access and respond the most in time and serve, client can not feel inconvenient yet simultaneously, makes the whole performance of system obtain effective raising.
Claims (3)
1. one kind is used for the data processing method that a large number of users is concentrated login, it is characterized in that, be used for a computer network, this computer network comprises one group of application server, an one group of client and a resource allotting server, each client in wherein said one group of client logs on its application server request application service that connects, and described one group of client and one group of application server all are connected to described resource allotting server, and described method comprises:
Each application server periodically reports current self business load to described resource allotting server, described resource allotting server generates resource allocation table according to the business load that all application servers report, wherein, described resource allocation table periodically upgrades according to the business load that application server newly reports;
Described resource allotting server is calculated each application server processes according to described resource allocation table and is landed the needed time of request, the rise time table, and described timetable is updated periodically according to the renewal of resource allocation table;
One first application server receives the request of landing of its client that connects, and this request of landing is reported described resource allotting server;
Described resource allotting server receives this land request after, search described timetable, seek second application server, second application server has processing and lands the needed shortest time, resource allotting server sends to second application server with the described request of landing and handles, and the result of processing feeds back to described resource allotting server;
The result that described resource allotting server will be handled feeds back to first application server, feeds back to client by first application server.
2. a large number of users that is used for as claimed in claim 1 is concentrated the data processing method of logining, it is characterized in that,
If described second application server is exactly first application server itself, then described resource allotting server sends a message to described first application server, and first application server is directly handled this request of landing and fed back to client.
3. a large number of users that is used for as claimed in claim 1 is concentrated the data processing method of logining, it is characterized in that,
Described resource allotting server is periodically to current self the business load of each application server queries application server, and the business load that obtains according to inquiry generates resource allocation table.
Priority Applications (1)
Application Number | Priority Date | Filing Date | Title |
---|---|---|---|
CN200910198696A CN101719865A (en) | 2009-11-12 | 2009-11-12 | Data processing method for centralized login of a plurality of users |
Applications Claiming Priority (1)
Application Number | Priority Date | Filing Date | Title |
---|---|---|---|
CN200910198696A CN101719865A (en) | 2009-11-12 | 2009-11-12 | Data processing method for centralized login of a plurality of users |
Publications (1)
Publication Number | Publication Date |
---|---|
CN101719865A true CN101719865A (en) | 2010-06-02 |
Family
ID=42434385
Family Applications (1)
Application Number | Title | Priority Date | Filing Date |
---|---|---|---|
CN200910198696A Pending CN101719865A (en) | 2009-11-12 | 2009-11-12 | Data processing method for centralized login of a plurality of users |
Country Status (1)
Country | Link |
---|---|
CN (1) | CN101719865A (en) |
Cited By (3)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
CN102158475A (en) * | 2011-02-22 | 2011-08-17 | 山东大学 | System architecture based on student dormitory passageway system and system data synchronization method thereof |
CN103200225A (en) * | 2013-02-25 | 2013-07-10 | 汉柏科技有限公司 | Method and system of centralized management of services in server cluster |
CN103347062A (en) * | 2013-06-24 | 2013-10-09 | 北京汉柏科技有限公司 | Virtual computing resource distribution method |
-
2009
- 2009-11-12 CN CN200910198696A patent/CN101719865A/en active Pending
Cited By (5)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
CN102158475A (en) * | 2011-02-22 | 2011-08-17 | 山东大学 | System architecture based on student dormitory passageway system and system data synchronization method thereof |
CN102158475B (en) * | 2011-02-22 | 2014-03-19 | 山东大学 | System architecture based on student dormitory passageway system and system data synchronization method thereof |
CN103200225A (en) * | 2013-02-25 | 2013-07-10 | 汉柏科技有限公司 | Method and system of centralized management of services in server cluster |
CN103200225B (en) * | 2013-02-25 | 2016-12-28 | 汉柏科技有限公司 | The centralized management method and system of service in a kind of server cluster |
CN103347062A (en) * | 2013-06-24 | 2013-10-09 | 北京汉柏科技有限公司 | Virtual computing resource distribution method |
Similar Documents
Publication | Publication Date | Title |
---|---|---|
CN102521012B (en) | Virtual machine-based general processing unit (GPU) cluster management system | |
CN101841565B (en) | Database cluster system load balancing method and database cluster system | |
CN104850450A (en) | Load balancing method and system facing mixed cloud application | |
CN103067297B (en) | A kind of dynamic load balancing method based on resource consumption prediction and device | |
CN103401947A (en) | Method and device for allocating tasks to multiple servers | |
CN105426245A (en) | Dynamically composed compute nodes comprising disaggregated components | |
CN102447734B (en) | Cloud service method for taxation cloud computing network billing IM (Instant Messaging) online customer system | |
CN106027328A (en) | Cluster monitoring method and system based on application container deployment | |
JPH10507024A (en) | Load distribution support tool for distributed applications | |
CN105516325A (en) | Cloud load balancing method for carrying out elastic expansion and traffic distribution expansion according to application load | |
CN104601680A (en) | Resource management method and device | |
CN204740299U (en) | Electric energy quality intelligent monitoring system based on cloud calculates | |
US20100161145A1 (en) | Search engine design and computational cost analysis | |
CN101702668A (en) | Gridding load balance system | |
CN104077188A (en) | Method and device for scheduling tasks | |
CN108282526B (en) | Dynamic allocation method and system for servers between double clusters | |
CN101753359A (en) | Method and system for dynamically distributing components | |
CN105162623A (en) | Cloud processing method and cloud server | |
CN102036188B (en) | Mail proxy method, equipment and system under multi-node system | |
CN101751297A (en) | Information system to which a large number of clients can log in and method for large number of clients to log in to same | |
CN101719865A (en) | Data processing method for centralized login of a plurality of users | |
CN111131040A (en) | Route configuration method, device and system, storage medium and electronic device | |
CN107071020A (en) | Load balancing framework applied to cloud computing server | |
CN104468710A (en) | Mixed big data processing system and method | |
CN104852964A (en) | Multifunctional server scheduling method |
Legal Events
Date | Code | Title | Description |
---|---|---|---|
C06 | Publication | ||
PB01 | Publication | ||
C02 | Deemed withdrawal of patent application after publication (patent law 2001) | ||
WD01 | Invention patent application deemed withdrawn after publication |
Application publication date: 20100602 |