CN104852964A - Multifunctional server scheduling method - Google Patents
Multifunctional server scheduling method Download PDFInfo
- Publication number
- CN104852964A CN104852964A CN201510169232.4A CN201510169232A CN104852964A CN 104852964 A CN104852964 A CN 104852964A CN 201510169232 A CN201510169232 A CN 201510169232A CN 104852964 A CN104852964 A CN 104852964A
- Authority
- CN
- China
- Prior art keywords
- server
- client
- group
- idle
- pointer
- 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
- H04L67/1017—Server selection for load balancing based on a round robin mechanism
-
- H—ELECTRICITY
- H04—ELECTRIC COMMUNICATION TECHNIQUE
- H04L—TRANSMISSION OF DIGITAL INFORMATION, e.g. TELEGRAPHIC COMMUNICATION
- H04L47/00—Traffic control in data switching networks
- H04L47/50—Queue scheduling
-
- 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
- H04L67/50—Network services
- H04L67/60—Scheduling or organising the servicing of application requests, e.g. requests for application data transmissions using the analysis and optimisation of the required network resources
Abstract
The invention discloses a multifunctional server scheduling method comprising the following steps of: 1) dividing servers into multiple server groups according to functions; 2) establishing an index table indicating a corresponding relation between the server groups and client services; 3) according to the client service, comparing the index table in order to select the server group; and 4) determining a server in the server group corresponding to the service of a client by using a polling scheduling way. According to the method, the servers are divided according to functions and the index table is established such that a client request may quickly position server resources; the same server group uses the polling scheduling algorithm so that the client requests can be uniformly distributed to different servers and a single server is prevented from being overloaded or idle.
Description
Technical field
The present invention relates to the communications field, particularly relate to a kind of fat server dispatching method.
Background technology
On a server cluster, there is the server of a lot of difference in functionality, these function servers are independent separately, process the business of oneself, comprehensive network service needed for enterprise is provided, comprise: the Internet service demand that the Email, enterprise web site, Web service, ftp file transmission, file-sharing, project management system, Network Office System etc. of enterprise are common, seamless integration can be carried out with enterprise existing system simultaneously, ensure the investment of enterprise and the utilization of original system.Multiple safe mechanism, for solving network security and network management services problem, ensures enterprise online safety and data security.In order to increase the ability of this class business parallel processing, often adopt the server computing simultaneously of multiple stage identical function, the request of client is assigned to different machines and gets on by service routine at this moment with regard to needing a scheduling of resource uniformly.Common servo load balance method is user's manual selection process and DNS polling mode.These two kinds of method reliabilities are low, load distributes unbalanced, cause a few station server load very low, and another a few station server load is very high, process slowly, to configure the request that high server-assignment arrives few, and the request configuring low server-assignment is many, cause the situation that individual server pressure is overweight or idle.
Summary of the invention
For above-mentioned technical problem, the invention provides a kind of fat server dispatching method, step comprises:
1) server is divided into multiple server group by function;
2) concordance list that server group is corresponding with client traffic is set up;
3) according to client traffic, server group selected by concordance list described in comparison;
4) method of polling dispatching is adopted, for client determines the server in the server group corresponding with its business.
Preferred version is: the server with group lines up a queue.
Preferred version is: the method step of described polling dispatching comprises:
A) extract client traffic ID, in concordance list, then search the server group corresponding with this traffic ID;
B) judge that pointed by corresponding server group pointer, whether server is idle backward after the server pointer of this server group being moved a unit, if server free, return client " this server resource can be used "; If server is busy, described pointer moves after continuing, continue to judge that whether server is idle, if arrive the tail of the queue of described queue still without idle server, then described pointer jumps to team's head of described queue, restart to judge that whether server idle, until described pointer movement to the position that pointer points at first still without idle server, then return client " without available server resource ".
Preferred version is: described server all has a spendable mark.
Preferred version is: the original position of client-access server group is the next position of client last visit server group pointed position.
Preferred version is: described server is divided into multiple groups by function, and often group performs identical function.
Server divides into groups according to function by the present invention, and set up index, can allow client-requested quick position server resource, same group of server adopts polling dispatching algorithm, client-requested can be allowed to be assigned to uniformly on different machines, to avoid the possibility that individual server pressure is overweight or idle.
Accompanying drawing explanation
Fig. 1 is workflow schematic diagram of the present invention.
Embodiment
On a server cluster, there is the server of a lot of difference in functionality, these function servers are independent separately, process the business of oneself, in order to increase the ability of this class business parallel processing, often adopt the server computing simultaneously of multiple stage identical function, the request of client is assigned to different machines and gets on by service routine at this moment with regard to needing a scheduling of resource uniformly.Common servo load balance method is:
1) user's manual selection process
This is a kind of comparatively ancient mode, by the mode providing different circuit, different server to connect at main website homepage entrance, realizes load balancing.
2) DNS polling mode
The parsing of many A record is supported in most domain name registration commercial city, Here it is DNS poll, and the order that analysis request records according to A by dns server, is assigned on different IP one by one, this completes simple load balancing.Although the cost of DNS poll is very low, there are two obvious shortcomings in DNS load balancing.
A. reliability is low
Suppose a domain name DNS poll multiple servers, if a station server wherein breaks down, the request of so all this servers of access can not be responded to some extent, and this is that anyone is unwilling to see.Even if remove the IP of this server from DNS, but on internet, numerous DNS deposits in the buffer by the broadband ISPs such as each department telecommunications, Netcom, and to save the access time, DNS record all comes into force and needs several hours, even more of a specified duration.So, although DNS to some extent solves problem of load balancing in turn, but there is the shortcoming that reliability is not high.
B. load distributes unbalanced
What DNS load balancing adopted is the simplest poll loading algorithm, can not the difference of Differentiated Services device, the current operating conditions of server can not be reflected, the server overabsorption request into better performances can not be accomplished, even there will be client's request situation dns server concentrated on a certain station server is according to certain hierarchical structure tissue, local dns server can cushion the domain name of the having resolved mapping to IP address, what this can cause using the user of this dns server to access within a period of time is same Web server, cause the load imbalance between Web server.In addition, user's local computer also can the buffer memory domain name of having resolved to the mapping of IP address.When all certain domain name of buffer memory is to the mapping of IP address for multiple subscriber computer, and these users continue to access the webpage under this domain name, and this is that the load of different Web service device also can be caused to distribute is unbalanced.The consequence that load imbalance may cause has: a few station server load is very low, and another a few station server load is very high, process slow; Configure the request that high server-assignment arrives few, and the request configuring the server-assignment of Fiao is many.Therefore, DNS polling mode is only applicable to the not high server cluster of some reliability requirements, such as: picture servers group, pure static Web page server cluster etc.
Present invention employs polling dispatching method, polling dispatching is the one strategy of channel dispatch in communication, and this scheduling strategy makes user use shared resource in turn, can not consider instantaneous channel conditions.Be assigned to the angle of every communication link from equal number Radio Resource (identical scheduling time section), polling dispatching is regarded as equity dispatching.
Describe the present invention below in conjunction with Fig. 1:
The invention provides a kind of fat server dispatching method, step comprises:
1) server is divided into multiple server group by function, described server is divided into multiple groups by function, and often group performs identical function;
2) set up the concordance list that server group is corresponding with client traffic, extract client traffic ID, in concordance list, then search the server group corresponding with this traffic ID;
3) according to client traffic, server group selected by concordance list described in comparison;
4) adopt the method for polling dispatching, for client determines the server in the server group corresponding with its business, the server with group lines up a queue; The method step of described polling dispatching comprises: judge after the server pointer of this server group being moved backward a unit that pointed by corresponding server group pointer, whether server is idle, if server free, return client " this server resource can be used "; If server is busy, described pointer moves after continuing, continue to judge that whether server is idle, if arrive the tail of the queue of described queue still without idle server, then described pointer jumps to team's head of described queue, restart to judge that whether server idle, until described pointer movement to the position that pointer points at first still without idle server, then return client " without available server resource ".Described server all has a spendable mark.The original position of client-access server group is the next position of client last visit server group pointed position.
Although embodiment of the present invention are open as above, but it is not restricted to listed in specification and execution mode utilization, it can be applied to various applicable the field of the invention completely, for those skilled in the art, can easily realize other amendment, therefore do not deviating under the universal that claim and equivalency range limit, the present invention is not limited to specific details and illustrates here and the legend described.
Claims (6)
1. a fat server dispatching method, is characterized in that, comprises the following steps:
1) server is divided into multiple server group by function;
2) concordance list that server group is corresponding with client traffic is set up;
3) according to client traffic, server group selected by concordance list described in comparison;
4) method of polling dispatching is adopted, for client determines the server in the server group corresponding with its business.
2. server scheduling method according to claim 1, is characterized in that, the server with group lines up a queue.
3. server scheduling method according to claim 2, is characterized in that, the method step of described polling dispatching comprises:
A) extract client traffic ID, in concordance list, then search the server group corresponding with this traffic ID;
B) judge that pointed by corresponding server group pointer, whether server is idle backward after the server pointer of this server group being moved a unit, if server free, return client " this server resource can be used "; If server is busy, described pointer moves after continuing, continue to judge that whether server is idle, if arrive the tail of the queue of described queue still without idle server, then described pointer jumps to team's head of described queue, restart to judge that whether server idle, until described pointer movement to the position that pointer points at first still without idle server, then return client " without available server resource ".
4. server scheduling method according to claim 1, is characterized in that, described server all has a spendable mark.
5. server scheduling method according to claim 4, is characterized in that, the original position of client-access server group is the next position of client last visit server group pointed position.
6. server scheduling method according to claim 1, is characterized in that, described server is divided into multiple groups by function, and often group performs identical function.
Priority Applications (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN201510169232.4A CN104852964A (en) | 2015-04-10 | 2015-04-10 | Multifunctional server scheduling method |
Applications Claiming Priority (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN201510169232.4A CN104852964A (en) | 2015-04-10 | 2015-04-10 | Multifunctional server scheduling method |
Publications (1)
| Publication Number | Publication Date |
|---|---|
| CN104852964A true CN104852964A (en) | 2015-08-19 |
Family
ID=53852315
Family Applications (1)
| Application Number | Title | Priority Date | Filing Date |
|---|---|---|---|
| CN201510169232.4A Pending CN104852964A (en) | 2015-04-10 | 2015-04-10 | Multifunctional server scheduling method |
Country Status (1)
| Country | Link |
|---|---|
| CN (1) | CN104852964A (en) |
Cited By (4)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN105959410A (en) * | 2016-06-29 | 2016-09-21 | 深圳中兴网信科技有限公司 | Load balancing method and load balancing system of server cluster |
| CN111294372A (en) * | 2018-12-07 | 2020-06-16 | 北京京东尚科信息技术有限公司 | Method, device and system for realizing cache in proxy server |
| CN113515369A (en) * | 2021-04-23 | 2021-10-19 | 深圳希施玛数据科技有限公司 | Data processing method, system, terminal and storage medium |
| CN114726907A (en) * | 2022-05-18 | 2022-07-08 | 北京结慧科技有限公司 | Access method, system, equipment and storage medium of mobile terminal product |
Citations (4)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN101621413A (en) * | 2009-08-20 | 2010-01-06 | 中兴通讯股份有限公司 | Apparatus and method for performing load balance and disaster tolerance to WEB server |
| US20120278432A1 (en) * | 2011-04-27 | 2012-11-01 | Michael Luna | System and method for making requests on behalf of a mobile device based on atomic processes for mobile network traffic relief |
| CN103491123A (en) * | 2012-06-14 | 2014-01-01 | 中国移动通信集团贵州有限公司 | Load balancing method based on domain name visiting, system and load balancer |
| CN104363275A (en) * | 2011-12-29 | 2015-02-18 | 北京奇虎科技有限公司 | Method, device and system for processing data by using agencies |
-
2015
- 2015-04-10 CN CN201510169232.4A patent/CN104852964A/en active Pending
Patent Citations (4)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN101621413A (en) * | 2009-08-20 | 2010-01-06 | 中兴通讯股份有限公司 | Apparatus and method for performing load balance and disaster tolerance to WEB server |
| US20120278432A1 (en) * | 2011-04-27 | 2012-11-01 | Michael Luna | System and method for making requests on behalf of a mobile device based on atomic processes for mobile network traffic relief |
| CN104363275A (en) * | 2011-12-29 | 2015-02-18 | 北京奇虎科技有限公司 | Method, device and system for processing data by using agencies |
| CN103491123A (en) * | 2012-06-14 | 2014-01-01 | 中国移动通信集团贵州有限公司 | Load balancing method based on domain name visiting, system and load balancer |
Cited By (6)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN105959410A (en) * | 2016-06-29 | 2016-09-21 | 深圳中兴网信科技有限公司 | Load balancing method and load balancing system of server cluster |
| CN105959410B (en) * | 2016-06-29 | 2019-06-04 | 深圳中兴网信科技有限公司 | The load-balancing method and SiteServer LBS of server cluster |
| CN111294372A (en) * | 2018-12-07 | 2020-06-16 | 北京京东尚科信息技术有限公司 | Method, device and system for realizing cache in proxy server |
| CN111294372B (en) * | 2018-12-07 | 2023-04-07 | 北京京东尚科信息技术有限公司 | Method, device and system for realizing cache in proxy server |
| CN113515369A (en) * | 2021-04-23 | 2021-10-19 | 深圳希施玛数据科技有限公司 | Data processing method, system, terminal and storage medium |
| CN114726907A (en) * | 2022-05-18 | 2022-07-08 | 北京结慧科技有限公司 | Access method, system, equipment and storage medium of mobile terminal product |
Similar Documents
| Publication | Publication Date | Title |
|---|---|---|
| CN108052675B (en) | Log management method, system and computer readable storage medium | |
| JP5582344B2 (en) | Connection management system and connection management server linkage method in thin client system | |
| US8918599B2 (en) | Integrated storage platform system and method thereof | |
| WO2020133967A1 (en) | Method for scheduling shared computing resources, shared computing system, server, and storage medium | |
| CN104834722A (en) | CDN (Content Delivery Network)-based content management system | |
| US20130007253A1 (en) | Method, system and corresponding device for load balancing | |
| WO2015096681A1 (en) | Method and device for processing service access | |
| US20110166952A1 (en) | Facilitating dynamic construction of clouds | |
| CN108833462A (en) | A kind of system and method found from registration service towards micro services | |
| CN110071965B (en) | Data center management system based on cloud platform | |
| CN104601702B (en) | Cluster remote procedure calling (PRC) method and system | |
| CN104852964A (en) | Multifunctional server scheduling method | |
| Rawadi et al. | Providing local cloud services to mobile devices with inter-cloudlet communication | |
| CN104320488A (en) | Proxy server system and proxy service method | |
| CN103997482A (en) | Method of user registration in desktop cloud service, and system | |
| US20110153826A1 (en) | Fault tolerant and scalable load distribution of resources | |
| CN105897865B (en) | A kind of network file service management system and method that agreement is unrelated | |
| CN110798495B (en) | Method and server for end-to-end message push in cluster architecture mode | |
| CN105515839A (en) | System and method for promoting Radius service performance | |
| US10715608B2 (en) | Automatic server cluster discovery | |
| US8533333B2 (en) | Shared hosting using host name affinity | |
| CN105025042B (en) | A kind of method and system of determining data information, proxy server | |
| JP6888478B2 (en) | Sorting system | |
| CN113596105B (en) | Content acquisition method, edge node and computer readable storage medium | |
| CN106357549B (en) | Method, apparatus, system and the load balancer that high concurrent shunts in clustered deploy(ment) |
Legal Events
| Date | Code | Title | Description |
|---|---|---|---|
| C06 | Publication | ||
| PB01 | Publication | ||
| EXSB | Decision made by sipo to initiate substantive examination | ||
| SE01 | Entry into force of request for substantive examination | ||
| RJ01 | Rejection of invention patent application after publication | ||
| RJ01 | Rejection of invention patent application after publication |
Application publication date: 20150819 |