CN100391178C - Method for selecting server in network - Google Patents
Method for selecting server in network Download PDFInfo
- Publication number
- CN100391178C CN100391178C CNB031536654A CN03153665A CN100391178C CN 100391178 C CN100391178 C CN 100391178C CN B031536654 A CNB031536654 A CN B031536654A CN 03153665 A CN03153665 A CN 03153665A CN 100391178 C CN100391178 C CN 100391178C
- Authority
- CN
- China
- Prior art keywords
- edge server
- server
- user
- standby
- group
- 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.)
- Expired - Fee Related
Links
Images
Landscapes
- Computer And Data Communications (AREA)
- Information Transfer Between Computers (AREA)
Abstract
The present invention discloses a method for selecting servers in a network, which determines a selecting strategy according to user information. If the selecting strategy is a static strategy, the optimal edge server is selected by judging whether or not an edge server is overloaded; if the selecting strategy is a dynamic strategy, a group of alternate select edge servers are selected and determined by judging whether or not the edge server is overloaded and has a content and a service type requested by a user; then, a connecting state between the group of edge servers and a user access server; consequently, the optimal edge server is selected. The method for selecting servers in a network not only can rapidly and accurately give a selecting result, but also can ensure a good content hit rate. Load distribution among the servers is reasonable. No influence is caused on user application, and the network flow of user sides can not be increased.
Description
Technical field
The present invention relates to content distribution network (CDN, Content Delivery Network) technology, relate in particular to a kind of method of in CDN, selecting the optimal edge server.
Background technology
Content distribution network is in order to improve the service quality (QoS of internet (Internet), Qualityof Service) technology that grows up, as shown in Figure 1, it generally comprises three kind equipments: the Edge Server that is distributed in network edge, content is sent to the centre point discharge device of Edge Server and user's request scheduling system.Wherein user's request scheduling system is a class load-balancing device, its function is: when the user arrives user's request scheduling system to requests for content, one of the selection among the Edge Server that distributes of this equipment is only, allows this Edge Server provide content to the user then.Can alleviate the processing pressure of source server like this, reduced the bandwidth consumption of backbone network simultaneously, reduced after the key route of packet process because negative effects such as the packet loss that many " jumping " transmission brings, delay, distortion have improved the view quality of looking content of multimedia such as audio frequency, animation on the broadband network in a large number greatly.The index of above-mentioned this " only " is diversified, may be that the Edge Server load is the lightest, and is nearest from the user on the network topology, the network bandwidth maximum between user and this Edge Server, and between the user transmission time of IP message the shortest or the like.User's request scheduling system mainly contains the selection mode of Edge Server in the prior art:
One, based on the mode of domain name resolution server (DNS) agreement of expansion.As shown in Figure 2, equipment 1 is the mandate dns server of CDN system among the figure, handles all DNS that points to CDN network requests.The database table of a domain name and corresponding Edge Server IP in this plant maintenance, and comprise all Edge Servers running status, handle information such as load.When the user imports a standard resource address (URL) that comprises domain name in browser, owing to do not determine to which edge of table server to initiate to connect at last as yet, so at first send domain name to dns server by DNS Protocol, ask the IP address of the Edge Server corresponding with this domain name, after dns server is received this DNS request analysis, select the IP address of the light server of a health and load to return to user browser according to the health of all Edge Servers of grasping and load message, end user's browser initiates to connect to this IP.
The shortcoming of the technical program at first derives from DNS Protocol itself: only comprise the domain name that will inquire about in the DNS data query bag of standard, and do not pay close attention to agreement and the object content that request is used, such as file path, filename, authentication information etc., therefore the dns server Edge Server that can only select a physics to use, angle from application layer, it can not guarantee that this Edge Server can provide the user desired business, such as in Fig. 2, dns server is the common ftp server of ip1 for the user selects the IP address to the request of streaming medium content probably, and can not guarantee has the content that needs on this Edge Server; Secondly, it does not consider the network connection state between user and the server, the user therefore may occur by serving from his a far station server, and service quality is not improved.Therefore, this selection mode is just useful to alleviating server stress, and the user capture quality is not had too big improvement.
Two, static IP address binding mode.Determine in advance the corresponding relation of an IP address field and Edge Server, all Edge Servers of asking all to transfer to a correspondence from the user in certain or certain several IP address ranges are handled.Edge Server increases content " drop-down (pull) " function initiatively, promptly as user during to the non-existent content of Edge Server request, Edge Server can be automatically downloaded content buffer memory to the local device from source server, provides service for the user in data flow cache.
The shortcoming of this mode is: similar with first kind of mode, can not guarantee the content that corresponding Edge Server has the user to ask; In addition, because binding relationship fixes, if the network planning is improper, or the user in a certain zone might cause some Edge Server overload and the situation of other Edge Server free time when asking to concentrate, also can be higher to the requirement of Edge Server.And the webpage making person must increase field description in URL, points out should not go drop-down content from which origin url when having content on the Edge Server, and this has also increased the restriction to webpage making.
Three, the mode of Dynamic Selection.After user's request scheduling system receives that the user asks, allow all Edge Servers determine by Internet Control Message Protocol (ICMP) message and the user between cyberrelationship, select a shortest Edge Server of message passing time then, provide service to the user.It can obtain the network connection state between Edge Server and the user in real time, thereby guarantees the bandwidth of user capture.
The technical program just has superiority at network sender mask, but shortcoming is to increase the weight of on the one hand the burden of Edge Server, if the Edge Server number is too many on the other hand, and the network bandwidth that the user inserts too hour, may cause incorrect result, and also can impact user's application.And, this system of selection length consuming time, wait until that all Edge Servers are all obtained network condition information to the user after, just can do decision-making.In addition, it does not make full use of existing network topology, and in fact, it is very little that the user inserts the variation of network topology of side, and the general user has only a network insertion, and it is in fact unnecessary to the network between the user to measure Edge Server.
From the above mentioned as can be known: existing technical scheme when selecting Edge Server, often only in the index of " only " an index and be optimized.May not find real " only " Edge Server on the one hand, not consider the structure of existing network on the other hand, not consider that also different business should be different to some extent to the selection strategy of Edge Server, so effect is unsatisfactory.
Summary of the invention
Purpose of the present invention just provides a kind of method of more rational selection optimal edge server, can either provide selection result soon, can guarantee good content hit rate again, make the load allocating between the Edge Server more reasonable simultaneously, and can not increase the network traffics of user side.
The object of the present invention is achieved like this: a kind of method of selecting server in network is characterized in that this method may further comprise the steps:
The request that user side is received by a1, user's request scheduling system, determine the access server of this user side according to the IP address of active user's end, and further according to be provided with in advance to the scheduling strategy that each access server adopted, determine scheduling strategy to this access server adopted;
Whether a2, judgement are static policies to the scheduling strategy that this access server adopts, if then select the optimal edge server of binding in advance according to static configuration; Otherwise, then determine candidate's Edge Server of one group of appointment earlier according to static configuration, then detection of dynamic should the group Edge Server in the network connection state of each candidate's Edge Server between this subscription access server, select optimum Edge Server according to testing result.
Dispatching patcher is determined scheduling strategy to each access server adopted according to the distribution situation of network in advance.
Described static configuration comprises: comprise the content name that the user asks, the user profile of IP address at least; Whether Edge Server preserves the content of user's request; The type of service whether Edge Server can provide the user to need; Normally whether Edge Server operation; Whether the inner network interface of Edge Server transships; Whether system uses the central store server as standby server.
Described Edge Server according to static configuration selection optimum comprises:
B1, user's request scheduling system determine an Edge Server and standby Edge Server that this access server is bound according to the access server described in the step a1;
B2, according to Edge Server whether normally operation judge whether this Edge Server transships, if transship, illustrate that then all network interfaces of this Edge Server all transship, and execution in step b4; Otherwise selecting this Edge Server is the optimal edge server;
Whether each network interface of the Edge Server described in b3, the determining step b2 transships, if nonoverload then selects this network interface that the user is served, and process ends; Otherwise, return step b3, next network interface is judged;
B4, according to standby Edge Server whether normally operation judge whether this standby Edge Server transships, if overload, all network interfaces that this standby Edge Server then is described all transship, and report an error or select the central store server to come the process user request as standby server, and process ends, otherwise selecting this standby Edge Server is the optimal edge server;
B5, whether each network interface of the standby Edge Server described in the determining step b4 transships successively, if nonoverload then selects this network interface to provide service for the user, and process ends, otherwise, return step b5, next network interface is judged.
Nearby principle is pressed for adopting an access server Edge Server of binding and a standby Edge Server of static policies according to the distribution situation of network in advance by user's request scheduling system.
Describedly determine that according to static configuration one group of candidate's Edge Server may further comprise the steps:
C1, user's request scheduling system determine the pairing one group of Edge Server of this access server and one group of standby Edge Server according to the access server described in the step a1;
C2, according to Edge Server whether normally operation judge successively whether each Edge Server in this group Edge Server transships, if overload, then with Edge Server deletion from this group of overload, otherwise, with this Edge Server reservation; Judge whether whole overload of this group Edge Server, if, execution in step c4 then, otherwise, to remaining Edge Server execution in step c3 in this group;
C3, whether each network interface of each remaining Edge Server transships described in the determining step c2 successively, if nonoverload is then selected this network interface, and execution in step c6; Otherwise, return step c3, next network interface is judged;
C4, according to standby Edge Server whether normally operation judge successively whether this each standby Edge Server of organizing in the standby Edge Server transships, if overload, the standby Edge Server that then will transship is deleted from this group, otherwise level should keep by standby Edge Server; Judge that this organizes whether whole overload of standby Edge Server, if, then report an error or select the central store server to come the process user request as standby server, and process ends, otherwise, to remaining standby Edge Server execution in step c5 in this group;
C5, whether each network interface of each remaining standby Edge Server transships described in the determining step c4 successively, if nonoverload is then selected this network interface, and execution in step c6, otherwise, return step c5, next network interface is judged;
C6, with Edge Server remaining in current group as a candidate set;
Whether exist in c7, the described candidate set of determining step c6 and contain the content that the user asks and the Edge Server of type of service, if exist, then with all contain the Edge Server of content that the user asks and type of service as one group of candidate's Edge Server in this candidate set, if there is no, then will from described candidate set, select an Edge Server that load is the lightest to provide service to the user, and perhaps select the central store server to come the process user request as standby server as the optimal edge server.
User's request scheduling system specifies one group Edge Server and one group standby Edge Server by nearby principle for the access server that adopts dynamic strategy according to the distribution situation of network in advance.
Connection state between each Edge Server of described detection of dynamic and the selected user access server is selected the optimal edge server according to testing result, may further comprise the steps:
Each candidate's Edge Server in candidate's Edge Server group described in c8, the indication step c7 of user's request scheduling system, detect the network connection state between this Edge Server self and the above-mentioned subscription access server respectively, according to the testing result of connection state, selection and this subscription access server are apart from the Edge Server of the shortest Edge Server as optimum;
C9, user's request scheduling system use the Edge Server and the selected network interface of this Edge Server of selected optimum among the step c8 to provide service to user side.
Select the optimal edge server by dispatching method static and that dynamically combine among the present invention, thereby can provide the result of selection rapidly and accurately; By judging that whether having content that the user asks and type of service and Edge Server whether to transship in the Edge Server screens Edge Server, thereby guaranteed good content hit rate, the load allocating between the Edge Server is also more reasonable; When the detection of dynamic network, not directly to detect the network connection of Edge Server, but detect the connection of Edge Server to the affiliated access server of user to user side, do not impact so can not use the user, also can not increase the network traffics of user side.
Description of drawings
Fig. 1 is the CDN network structure;
Fig. 2 utilizes DNS Protocol to select the conceptual scheme of optimal edge server in the CDN network;
Fig. 3 is for selecting the system construction drawing of optimal edge server;
Fig. 4 is for selecting the flow chart of optimal edge server.
Embodiment
Below in conjunction with the drawings and specific embodiments implementation method of the present invention is described.
Core concept of the present invention is: user's request scheduling system combines static and dynamic two kinds of selection strategies, determine corresponding selection strategy at concrete user, if static policies then adopts the static IP address binding mode to select optimum Edge Server; If dynamic strategy then adopts and determines one group of candidate's Edge Server according to static configuration earlier, this group Edge Server of detection of dynamic is selected optimum Edge Server to the mode of connection state between the subscription access server then.
Fig. 3 is for selecting the system construction drawing of optimal edge server.Wherein, solid line is represented content information, and dotted line is represented state information, and chain-dotted line is represented user profile.As shown in Figure 3, wherein content manager is responsible for providing end to converge content from content, behind the intelligent optimization distributed topology, reasonably is distributed to content on the Edge Server, distribution of content information is the input system data base administration on the one hand, on the one hand real-time informing user request scheduling system.The distribution of Edge Server, central store server received content receives user's request of user's request scheduling system forwards, and the final content requests of process user.Edge Server, central store server are regularly notified user's request scheduling system the status attribute information such as professional ability of the situation of self-operating, apparatus of load, equipment.
User's request scheduling system is a core of the present invention, and this user's request scheduling system comprises distribution of content topology, all centers/edge device initial condition attribute from the initialization information of system database acquisition system operation.User's request scheduling system obtains real time content information from content manager in running, obtain equipment operation information from Edge Server.
Content manager provides with content and uses Transmission Control Protocol to communicate by letter between end, the database, thereby guarantees communication reliability; Use the normal data bank interface between user's request scheduling system and the database, with the specialized protocol of central store server/edge memory employing, to raise the efficiency based on User Datagram Protoco (UDP) (UDP); Use standard hypertext transmission (HTTP) protocol communication between user's request scheduling system and the user, to guarantee versatility.Also can carry out communication between user's request scheduling system and the user by other agreement.
The user mainly contains a certain requests for content input information: the contents attribute that visit, as filename, file type, service provider's sign; User's identity information is as IP.What user expectation obtained is the IP address that the Edge Server of service can be provided.User's request scheduling system obtains necessary required parameter from user's input information, other equipment obtain the running status of distribution of content and each Edge Server from the CDN system, thereby determines optimum Edge Server.User's request scheduling system is selecting the optimal service device will consider following factor when (comprising Edge Server and central store server):
1, user profile: as content name, contents attribute, service attribute, the User IP of request;
2, user and the Edge Server static policies of clearly binding and determine candidate set is earlier selected the dynamic strategy of optimal edge server again from this candidate set;
3, whether Edge Server preserves the content of user's request;
4, whether Edge Server can provide the type of service that the user needs;
5, whether server normally moves;
6, whether the inner network interface of server transships;
7, whether the network connection state between user and Edge Server is unobstructed;
8, whether system uses the central store server as standby server;
Above-mentioned factor of judgment can be matched flexibly.
Current network ip address generally is to distribute according to wideband network access server.User's request scheduling system presses nearby principle according to the distribution situation of network in advance and determines good corresponding Edge Server and standby Edge Server for each access server, and the scheduling strategy that provides each access server in advance and adopted, if the corresponding Edge Server of this access server and a standby Edge Server then adopt static policies; If the corresponding one group of Edge Server of this access server and one group of standby Edge Server then adopt dynamic strategy.Here, described static policies is meant according to static configuration and selects optimum Edge Server; Described dynamic strategy is meant determines one group of candidate's Edge Server according to static configuration earlier, then detection of dynamic should the group Edge Server in the network connection state of each candidate's Edge Server between this subscription access server, select optimum Edge Server according to testing result.
As shown in Figure 4, select the optimal service device to be divided into two steps substantially:
A1, user's request scheduling system determine this user's access server according to active user's IP address, and further determine the scheduling strategy to this access server adopted;
If what a2 adopted is static policies, then select optimum Edge Server according to static configuration; If what adopt is dynamic strategy, then earlier determine one group of candidate's Edge Server according to static configuration, then detection of dynamic should the group Edge Server in the network connection state of each candidate's Edge Server between this subscription access server, select optimum Edge Server according to testing result.
Be the concrete steps of static policies in the present embodiment below:
B1, user's request scheduling system determine an Edge Server and standby Edge Server that this access server is bound according to the access server described in the step a1, and herein an Edge Server and a standby Edge Server are that user's request scheduling system is the access server appointment of adopting static policies by nearby principle according to the distribution situation of network in advance;
B2, according to Edge Server whether normally operation judge whether this Edge Server transships, if transship, execution in step b4 then; Otherwise selecting this Edge Server is the optimal edge server;
Whether each network interface of the Edge Server described in b3, the determining step b2 transships, if nonoverload then selects this network interface that the user is served, and process ends; Otherwise, return step b3, next network interface is judged;
B4, according to standby Edge Server whether normally operation judge whether this standby Edge Server transships, if overload, then report an error or select the central store server to come the process user request as standby server, and process ends, otherwise selecting this standby Edge Server is the optimal edge server;
B5, whether each network interface of the standby Edge Server described in the determining step b4 transships successively, if nonoverload then selects this network interface to provide service for the user, and process ends, otherwise, return step b5, next network interface is judged.
Be the concrete steps of dynamic strategy in the present embodiment below:
C1, user's request scheduling system determine the pairing one group of Edge Server of this access server and one group of standby Edge Server according to the access server described in the step c1, and herein one group of Edge Server and one group of standby Edge Server are that user's request scheduling system is the access server appointment of adopting dynamic strategy by nearby principle according to the distribution situation of network in advance;
C2, according to Edge Server whether normally operation judge successively whether each Edge Server in this group Edge Server transships, if overload, then with Edge Server deletion from this group of overload, otherwise, with this Edge Server reservation; Judge whether whole overload of this group Edge Server, if, execution in step c4 then, otherwise, to remaining Edge Server execution in step c3 in this group;
C3, whether each network interface of each remaining Edge Server transships described in the determining step c2 successively, if nonoverload is then selected this network interface, and execution in step c6; Otherwise, return step c3, next network interface is judged;
C4, according to standby Edge Server whether normally operation judge successively whether this each standby Edge Server of organizing in the standby Edge Server transships, if overload, the standby Edge Server that then will transship is deleted from this group, otherwise level should keep by standby Edge Server; Judge that this organizes whether whole overload of standby Edge Server, if, then report an error or select the central store server to come the process user request as standby server, and process ends, otherwise, to remaining standby Edge Server execution in step c5 in this group;
C5, whether each network interface of each remaining standby Edge Server transships described in the determining step c4 successively, if nonoverload is then selected this network interface, and execution in step c6, otherwise, return step c5, next network interface is judged;
C6, with Edge Server remaining in current group as a candidate set;
Whether exist in c7, the described candidate set of determining step c6 and contain the content that the user asks and the Edge Server of type of service, if exist, then with all contain the Edge Server of content that the user asks and type of service as one group of candidate's Edge Server in this candidate set, if there is no, then will from described candidate set, select an Edge Server that load is the lightest to provide service to the user, and perhaps select the central store server to come the process user request as standby server as the optimal edge server.
Each candidate's Edge Server in candidate's Edge Server group described in c8, the indication step c7 of user's request scheduling system, detect the network connection state between this Edge Server self and the above-mentioned subscription access server respectively, for example, send Internet Control Message Protocol (ICMP) message to access server, the test returns time, according to the connection state testing result, selection and subscription access server are apart from the Edge Server of the shortest Edge Server as optimum;
C9, user's request scheduling system use the Edge Server and the selected network interface of this Edge Server of selected optimum among the step c8 to provide service to user side.
The foregoing description is for realizing the present invention program's a more complete execution mode, in concrete selection course, also may increase or delete procedure in step and obtain the server of a near-optimization, such as not doing detection of dynamic, obtain network topology (as prior input) according to other mode.In addition, the factor of considering when selecting is not limited only to the scope of eight listed factors of this paper.
Claims (8)
1. method of selecting server in network is characterized in that this method may further comprise the steps:
The request that user side is received by a1, user's request scheduling system, determine the access server of this user side correspondence according to the IP address of active user's end, and further according to be provided with in advance to the scheduling strategy that each access server adopted, determine scheduling strategy to this access server adopted;
Whether a2, judgement are static policies to the scheduling strategy that this access server adopts, if then select the optimal edge server of binding in advance according to static configuration; Otherwise, then determine candidate's Edge Server of one group of appointment earlier according to static configuration, then detection of dynamic should the group Edge Server in the network connection state of each candidate's Edge Server between this subscription access server, select optimum Edge Server according to testing result.
2. the method for claim 1 is characterized in that, dispatching patcher is determined scheduling strategy to each access server adopted according to the distribution situation of network in advance.
3. the method for claim 1 is characterized in that, described static configuration comprises:
At least comprise the content name that the user asks, the user profile of IP address; Whether Edge Server preserves the content of user's request; The type of service whether Edge Server can provide the user to need; Normally whether Edge Server operation; Whether the inner network interface of Edge Server transships; Whether system uses the central store server as standby server.
4. method as claimed in claim 3 is characterized in that, described Edge Server according to static configuration selection optimum comprises:
B1, user's request scheduling system determine an Edge Server and standby Edge Server that this access server is bound according to the access server described in the step a1;
B2, according to Edge Server whether normally operation judge whether this Edge Server transships, if transship, execution in step b4 then; Otherwise selecting this Edge Server is the optimal edge server;
Whether each network interface of the Edge Server described in b3, the determining step b2 transships, if nonoverload then selects this network interface that the user is served, and process ends; Otherwise, return step b3, next network interface is judged;
B4, according to standby Edge Server whether normally operation judge whether this standby Edge Server transships, if overload then reports an error or selects the central store server to provide service for the user, and process ends, otherwise selecting this standby Edge Server is the optimal edge server;
B5, whether each network interface of the standby Edge Server described in the determining step b4 transships successively, if nonoverload then selects this network interface to provide service for the user, and process ends, otherwise, return step b5, next network interface is judged.
5. method as claimed in claim 4 is characterized in that, nearby principle is pressed for adopting an access server Edge Server of binding and a standby Edge Server of static policies according to the distribution situation of network in advance by user's request scheduling system.
6. method as claimed in claim 3 is characterized in that, describedly determines that according to static configuration one group of candidate's Edge Server may further comprise the steps:
C1, user's request scheduling system determine the pairing one group of Edge Server of this access server and one group of standby Edge Server according to the access server described in the step a1;
C2, according to Edge Server whether normally operation judge successively whether each Edge Server in this group Edge Server transships, if overload, then with Edge Server deletion from this group of overload, otherwise, with this Edge Server reservation; Judge whether whole overload of this group Edge Server, if, execution in step c4 then, otherwise, to remaining Edge Server execution in step c3 in this group;
C3, whether each network interface of each remaining Edge Server transships described in the determining step c2 successively, if nonoverload is then selected this network interface, and execution in step c6; Otherwise, return step c3, next network interface is judged;
C4, according to standby Edge Server whether normally operation judge successively whether this each standby Edge Server of organizing in the standby Edge Server transships, if overload, the standby Edge Server that then will transship is deleted from this group, otherwise, should keep by standby Edge Server; Judge that this organizes whether whole overload of standby Edge Server, if, then report an error or select the central store server to provide service for the user, and process ends, otherwise, to remaining standby Edge Server execution in step c5 in this group;
C5, whether each network interface of each remaining standby Edge Server transships described in the determining step c4 successively, if nonoverload is then selected this network interface, and execution in step c6, otherwise, return step c5, next network interface is judged;
C6, with Edge Server remaining in current group as a candidate set;
Whether exist in c7, the described candidate set of determining step c6 and contain the content that the user asks and the Edge Server of type of service, if exist, then with all contain the Edge Server of content that the user asks and type of service as one group of candidate's Edge Server in this candidate set, if there is no, then will from described candidate set, select an Edge Server that load is the lightest to provide service to the user, and perhaps select the central store server to come the process user request as standby server as the optimal edge server.
7. method as claimed in claim 6 is characterized in that, user's request scheduling system specifies one group Edge Server and one group standby Edge Server by nearby principle for the access server that adopts dynamic strategy according to the distribution situation of network in advance.
8. method as claimed in claim 6 is characterized in that, the connection state between each Edge Server of described detection of dynamic and the selected user access server is selected the optimal edge server according to testing result, may further comprise the steps:
Each candidate's Edge Server in candidate's Edge Server group described in c8, the indication step c7 of user's request scheduling system, detect the network connection state between this Edge Server self and the above-mentioned subscription access server respectively, according to the testing result of connection state, selection and this subscription access server are apart from the Edge Server of the shortest Edge Server as optimum;
C9, user's request scheduling system use the Edge Server and the selected network interface of this Edge Server of selected optimum among the step c8 to provide service to user side.
Priority Applications (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CNB031536654A CN100391178C (en) | 2003-08-19 | 2003-08-19 | Method for selecting server in network |
Applications Claiming Priority (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CNB031536654A CN100391178C (en) | 2003-08-19 | 2003-08-19 | Method for selecting server in network |
Publications (2)
| Publication Number | Publication Date |
|---|---|
| CN1585357A CN1585357A (en) | 2005-02-23 |
| CN100391178C true CN100391178C (en) | 2008-05-28 |
Family
ID=34597801
Family Applications (1)
| Application Number | Title | Priority Date | Filing Date |
|---|---|---|---|
| CNB031536654A Expired - Fee Related CN100391178C (en) | 2003-08-19 | 2003-08-19 | Method for selecting server in network |
Country Status (1)
| Country | Link |
|---|---|
| CN (1) | CN100391178C (en) |
Cited By (1)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| EP3879796A1 (en) * | 2020-03-13 | 2021-09-15 | Apple Inc. | Selection of edge application server |
Families Citing this family (21)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN100431306C (en) * | 2005-07-15 | 2008-11-05 | 环达电脑(上海)有限公司 | Dynamic-network connecting allocation system and method |
| CN101064729B (en) * | 2006-04-27 | 2010-06-09 | 中国电信股份有限公司 | System and method for realizing FTP download service through CDN network |
| CN101102288B (en) * | 2006-07-06 | 2011-08-24 | 阿里巴巴集团控股有限公司 | A method and system for realizing large-scale instant message |
| CN101132393B (en) * | 2006-08-25 | 2010-11-17 | 郑州威科姆科技股份有限公司 | Stream media service dynamic load method |
| CN101155071B (en) * | 2006-09-28 | 2012-08-08 | 华为技术有限公司 | Method, device and communication system for selecting local domain supervisor |
| CN101351023B (en) * | 2007-07-16 | 2011-08-24 | 中兴通讯股份有限公司 | Method and apparatus for changing service region of moving body |
| CN101394540B (en) * | 2007-09-21 | 2012-05-23 | 中兴通讯股份有限公司 | Stream distributing method for interactive network television system |
| CN101325559B (en) * | 2008-07-28 | 2010-08-18 | 腾讯科技（深圳）有限公司 | Method for commending game room, system and game server |
| CN101616187B (en) * | 2009-07-21 | 2012-01-25 | 中兴通讯股份有限公司 | User access control system, method and device |
| CN103763209A (en) * | 2014-01-03 | 2014-04-30 | 上海聚力传媒技术有限公司 | Scheduling method and device of CDN servers |
| CN103825837B (en) * | 2014-02-19 | 2017-06-06 | 上海视云网络科技有限公司 | A kind of method of the Distributed C DN overall schedulings of node load |
| CN104468777B (en) * | 2014-12-05 | 2018-01-23 | 北京奇虎科技有限公司 | Data manipulation method and device |
| CN105991430B (en) * | 2015-03-05 | 2022-01-14 | 李明 | Data routing across multiple autonomous network systems |
| CN105871985A (en) * | 2015-12-10 | 2016-08-17 | 乐视网信息技术（北京）股份有限公司 | Data access request processing method and apparatus, server, client and system |
| CN106254235B (en) * | 2016-07-27 | 2020-02-14 | 上海华为技术有限公司 | Load sharing method and equipment |
| CN107786620B (en) * | 2016-08-31 | 2019-10-22 | 华为技术有限公司 | A kind of method, terminal and server obtaining resource |
| CN106453123A (en) * | 2016-09-30 | 2017-02-22 | 广州视睿电子科技有限公司 | Cross-region business processing method and apparatus |
| CN108696549B (en) * | 2017-04-05 | 2021-05-25 | 北京京东尚科信息技术有限公司 | Load balancing method, device and system |
| CN108600165A (en) * | 2018-03-15 | 2018-09-28 | 北京大米科技有限公司 | Communication means, client, Centroid based on TCP and communication system |
| CN108737544B (en) * | 2018-05-22 | 2021-11-26 | 中国联合网络通信集团有限公司 | CDN node scheduling method and device |
| CN114553964A (en) * | 2020-11-20 | 2022-05-27 | 中移动信息技术有限公司 | Control method, device and equipment of simulcast system and simulcast system |
Citations (2)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| WO2000014633A1 (en) * | 1998-09-03 | 2000-03-16 | Sun Microsystems, Inc. | Load balancing in a network environment |
| US20030055971A1 (en) * | 2001-09-19 | 2003-03-20 | Menon Rama R. | Providing load balancing in delivering rich media |
-
2003
- 2003-08-19 CN CNB031536654A patent/CN100391178C/en not_active Expired - Fee Related
Patent Citations (2)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| WO2000014633A1 (en) * | 1998-09-03 | 2000-03-16 | Sun Microsystems, Inc. | Load balancing in a network environment |
| US20030055971A1 (en) * | 2001-09-19 | 2003-03-20 | Menon Rama R. | Providing load balancing in delivering rich media |
Cited By (1)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| EP3879796A1 (en) * | 2020-03-13 | 2021-09-15 | Apple Inc. | Selection of edge application server |
Also Published As
| Publication number | Publication date |
|---|---|
| CN1585357A (en) | 2005-02-23 |
Similar Documents
| Publication | Publication Date | Title |
|---|---|---|
| CN100391178C (en) | Method for selecting server in network | |
| EP3382963B1 (en) | Method and system for self-adaptive bandwidth control for cdn platform | |
| US6449647B1 (en) | Content-aware switching of network packets | |
| CN104885431B (en) | The method and device of traffic engineering based on content in software definition information centre network | |
| US6128644A (en) | Load distribution system for distributing load among plurality of servers on www system | |
| US7120662B2 (en) | Conductor gateway prioritization parameters | |
| US20030172163A1 (en) | Server load balancing system, server load balancing device, and content management device | |
| EP1704686B1 (en) | Directed pppoe session initiation over a switched ethernet | |
| JPH10307783A (en) | Site access control system and recording medium | |
| CN106888270A (en) | Return the method and system of source routing scheduling | |
| WO2004073269A1 (en) | Transmission system, distribution route control device, load information collection device, and distribution route control method | |
| CN109151009A (en) | A kind of CDN node distribution method and system based on MEC | |
| WO2005022825A1 (en) | A system of layered content delivery network and a method of layered delivering content | |
| CN102065142A (en) | File downloading based scheduling method and system for content delivery network (CDN) | |
| JP2004215116A (en) | Ip router, communication system, band setting method to be used therefor, and its program | |
| CN102195882A (en) | Method and device for selecting route according to data stream application type | |
| CN109120953A (en) | Self adaptation stream processing system for video and method based on SDN and EC technology | |
| CN103685609A (en) | Method and device for collecting routing configuration information in domain name resolution | |
| CN106533760A (en) | Method and apparatus for constructing CDN system in CM network | |
| CN100334837C (en) | A method for assigning path bandwidth in bearing control layer | |
| JP3453120B2 (en) | Communication service condition control method and apparatus | |
| CN109195187A (en) | Handover network dicing method based on network slice server log file | |
| US20020122388A1 (en) | Dynamically reconfigurable intelligent content-based network | |
| KR20080046900A (en) | Load balancing system and method load balancing device | |
| JP2001067290A (en) | High quality server access device |
Legal Events
| Date | Code | Title | Description |
|---|---|---|---|
| C06 | Publication | ||
| PB01 | Publication | ||
| C10 | Entry into substantive examination | ||
| SE01 | Entry into force of request for substantive examination | ||
| C14 | Grant of patent or utility model | ||
| GR01 | Patent grant | ||
| C17 | Cessation of patent right | ||
| CF01 | Termination of patent right due to non-payment of annual fee |
Granted publication date: 20080528 Termination date: 20120819 |