WO2010028590A1

WO2010028590A1 - Method for providing address list, peer-to-peer network and scheduling method thereof

Info

Publication number: WO2010028590A1
Application number: PCT/CN2009/073813
Authority: WO
Inventors: 邵济平
Original assignee: 中兴通讯股份有限公司
Priority date: 2008-09-12
Filing date: 2009-09-08
Publication date: 2010-03-18
Also published as: CN101355591A

Abstract

A Peer-to-Peer (P2P) network is provided. It includes index servers and media servers. The index server responds to a media content inquiry request from a client and returns an inquiry result. The media server stores media contents and accepts the management of the index servers. The index server includes a remote inquiry module, which is used for sending the inquiry request to other index servers and receiving the inquiry result when the index server is lack of the inquiry contents of the client. A method for providing an address list and a scheduling method in P2P network are also provided. The inquiry efficiency can be improved and the quantity of cross-domain flow can be reduced by the methods and network.

Description

TECHNICAL FIELD The present invention relates to streaming media technologies, and in particular, to a method for providing an address list, a P2P network, and a scheduling method thereof. BACKGROUND Currently, P2P technology has been widely used in the Internet, and the P2P mode has become the preferred mode for many services, and is widely used in the fields of file sharing, network video, and Internet telephony. P2P technology features distributed resource sharing and parallel transmission, providing users with more resources, higher available bandwidth, and better services. P2P applications can be divided into three models according to their architecture:

(1) Server-based model. The model provides the name index of the connected peer and all the resources on the network. At the same time, the establishment of communication between peers depends on the initialization of the server. The server is dominant in the system, similar to the traditional C/S (ie, Client/Server) model. The disadvantages of this model are: Because of the dependence of the various parts of the network on the central server, it is easy to cause access to "hot spots", single points of failure and legal issues;

(2) Part of the Server-based model (with-Server), usually, preparation. The peer entering the network notifies the index server of the message that the user has logged in. The index server establishes an index of the connected peer name, and the role is only to provide the index for the peer that is being added, and the resources are still distributed and stored on each peer. The process of establishing connection and communication between peers is completed by each peer, and the index server is no longer involved. This mode is mainly for the purpose of providing high-speed transmission and real-time communication of files. For example, early Napster, KaZaA, and recently emerged Edoney, eMule, Skype, etc. belong to the P2P technology architecture based on the Server model. In the prior art, the mode has problems such as inefficiency and difficulty in controlling the content source;

(3) Serverless model (non-Server). In this model, all members are peers, there is no concept of a server at all, the peer dynamically discovers the peers connected to it, and the communication between the peers is completely completed by itself without any help from the server, breaking the C/S. The traditional communication method of the model is also called pure P2P model. The model can completely share and quickly provide all the resources in the network under certain rules, and support the fast joining and exiting of each peer. This pure P2P mode The model allows the user to set communication rules and establish a unique network environment, which can maximize the utilization of network resources and achieve the goal of optimizing system performance. Such pure P2P systems include Gnutella and Freenet, Myjxta and the domestic SNS social network, Nexter and others. The pure P2P model can provide approximate plug-and-play characteristics, but there is no effective solution in the prior art for the problem of how peers on the network discover each other and establish communication with each other. SUMMARY OF THE INVENTION In view of this, the present invention provides a method for providing an address list with high query efficiency, a P2P network, and a scheduling method thereof. According to an aspect of the present invention, a P2P network is provided. The P2P network includes an index server and a media server, and the index server is configured to respond to a media content query request of the client and return a query result, where the media server is used to store the media. The content and the management of the index server, the index server includes a remote query module, configured to send a query request to other servers and receive the query result when the client query content is missing. The remote query module includes a DHT (Distributed Hash Table) transceiver module for querying other index servers through the DHT network. The index server further includes a return control module, configured to control the media server or other client address list returned by the query, so that other clients or media servers in the same network in the list are located in front of other media servers in the same area or Client. The cable 1 server further includes a missing counting module and a copy triggering module, where the missing counting module is configured to count the number of times the index server is missing the same query content, and when the number exceeds a preset threshold, the copy triggering module The media server that triggers the index server management issues a content request to the other media server for the same query content and copies it. The cable 1 server further includes a query counting module and a load balancing module, and the query counting module accumulates the number of times the same media content is requested to be queried. When the number exceeds a preset threshold, the load balancing module sets the same content in the The plurality of media servers managed by the index server are distributed, and subsequent multiple query requests for the same content will specify different media server responses for obtaining the distributed content. According to another aspect of the present invention, a scheduling method for a network is provided, the method comprising the steps of: Α, client login, returning an index server address; Β, client to index service The server queries the media server where the request content is located; c. The index server returns a list of addresses of the media server and other clients that have the requested content; D. The client sends a request content message to the media server or other client that has the requested content; The media server or other client returns the requested media content; the step C further includes the following steps: Cl. If the index server has a media server or other client address list that the client requests content, directly return the list, and then Step D, otherwise performing the following steps; C2, the index server sends a query request to other index servers; C3, the index server receives a list of addresses returned by other index servers and returns them to the client. The query request in the step C2 is sent through the DHT network. The address list of the step C will be in the order of other clients or media servers in the same network before other media servers or clients in the same area. The step C3 further includes the following steps: C31: accumulating the number of missing requests of the requested query content; C32. When the number of missing times exceeds a preset threshold, the index server triggers the media server managed by the index server to perform data to the media server returned by the other index servers. copy. The step C1 further includes the following steps: C11. Accumulating the number of requests for the requested content; C12. When the number of requests exceeds a preset threshold, the index server will request to query multiple media managed by the index server. The servers are distributed between the servers, and subsequent multiple query requests for the same content will specify different media server responses to obtain the distributed content. The step A returns a plurality of index server addresses that are geographically close, and the default index server is selected by default, and the step B is to send a request to the default index server. According to still another aspect of the present invention, a method for providing an address list is provided, the method comprising the following steps: A. The index server checks whether there is a media server or other client address list capable of providing content requested by the client, if If yes, go to step B, otherwise, go to step C; B. Return the address list of the media server or the other client directly to the client; C. The index server sends a query request to other index servers. The index server receives the address list returned by the other server 1 and returns it to the client. The beneficial effects of the present invention compared with the prior art are as follows: (1) The query between the index servers through the DHT network improves the query efficiency of the P2P network; (2) For the control of returning the result list order, the resources closer to the client are in front of the list, which is advantageous for controlling the cross-region traffic;

(3) The replication of missing content can establish local backup for frequently missing content, improve query efficiency and reduce cross-regional volume; (4) Frequently requested content distribution can play a role of load balancing. BRIEF DESCRIPTION OF THE DRAWINGS The accompanying drawings, which are set to illustrate,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,, 1 is a structural diagram of a P2P network index server according to an embodiment of the present invention; FIG. 2 is a P2P network architecture diagram according to an embodiment of the present invention; FIG. 3 is a diagram of a method according to an embodiment of the present invention. FIG. 4 is a flow chart of method steps according to an embodiment of the present invention; FIG. 5 is a flowchart of a method for providing an address list according to an embodiment of the present invention. DETAILED DESCRIPTION OF THE PREFERRED EMBODIMENTS Apparatus Embodiments Hereinafter, the present invention will be further described in detail by way of specific embodiments and with reference to the accompanying drawings. The embodiment of the present invention provides a P2P network. FIG. 1 is a structural diagram of a P2P network index server according to an embodiment of the present invention. As shown in FIG. 1, the P2P network includes an index server and a media server. The client issues a content query request to the index server, the index server responds to the query request and returns a list of media servers or other clients having the requested content, and the media server stores the media content. The index server may include a remote query module, a return control module, a missing count module, a copy trigger module, a query count module, and a load balancing module. The remote query module sends a query request to other servers 1 and receives the returned query result when the index server does not have a corresponding content resource list. The remote query module may include a DHT transceiver module for supporting the use of the DHT network for other index servers. Inquire. The return control module is used to control the media service returned by the query. The order of the servers is such that other clients or media servers in the same network in the list are located in front of other media servers in the same area, so that the client can preferentially access the media servers or other clients in the same network to avoid Large cross-regional network traffic. The missing counting module is configured to count the number of times the index server is missing the same query content. When the number exceeds a preset threshold, the replication triggering module triggers the media server to send a content request to the other media server for the same query content, so that the content is made. There is a backup on the local media server, and the request for the media content can no longer cause cross-region traffic problems. The query counting module is configured to count the number of times the same content is requested. When the number of times exceeds a preset threshold, the load balancing module distributes the same content between multiple media servers managed by the index server, and the subsequent occurrence is targeted. Multiple requests for the same content will specify different media server responses to balance the load. The network architecture of the embodiment of the present invention is as shown in FIG. 2, wherein a dotted double-headed arrow indicates a signaling message flow, and a solid double-headed arrow indicates data and a signaling flow. The client η first obtains the media identifier from the EPG, and then sends a query request to the cable server, the index server searches for the server with the content of the media, and returns its address and other client addresses to the requesting client. When returning the list of media servers, the index server preferentially returns other clients or media servers in the same network that request data from the client, and then returns to other media servers in the same area (compensating the server and other clients). This can effectively reduce the generation of cross-regional traffic. The method embodiment uses the foregoing network, and the embodiment of the present invention further provides a scheduling method for the network, as shown in FIG. 3, including the following steps S1 to S10:

S1. The client sends a request message to the index server 1;

52. After receiving the request from the client, the index server 1 locally queries the content required by the client;

53. If the local query is successful, the media server 1 address is immediately returned to the client; S4, the client sends a message to the media server 1, requesting the required media content;

S5. The media server 1 returns the data requested by the client, and the process ends;

S6. If the query fails on the local index server 1, the server queries the other index servers through the DHT network between the index servers; 57. The other index server n returns the found media server address to the local index server 1;

58. The above-mentioned found media server address is returned to the requesting client by the local index server 1; S9. After obtaining the media server address of the request, the local client directly sends a request message to the remote media server.

S10. The remote media server n responds to the media content required by the local client, and the process ends. Figure 4 depicts a flow chart of the method steps for P2P network scheduling, including the process of missing content replication and frequent request content distribution. That is, in the above step S6, the method further includes the following steps: S61. Determine whether the number of times the content of the local index server is missing exceeds a preset threshold;

562, if the threshold is not exceeded, then go to step S7;

563. If the threshold is exceeded, trigger the local media server to obtain media content from the remote media server. When there is no media content required by the client on the local server 1 server, it will query the other servers and feed the result back to the client. If the media content required by the client is found in other index servers and the content is missing multiple times on the local index server, when the accumulated number of misses reaches a preset threshold, the content is relatively local. For popular content, the local index server triggers the local media server to go to the remote media server to download the content. As a result, the local client will only generate local traffic during the request for this content without going to the remote to download data, reducing cross-region traffic on the backbone network. Step S3 further includes the following steps:

S31. Determine whether the number of times the content is requested exceeds a preset threshold.

S32. If the threshold is not exceeded, directly return the address of the media server 1 to the client, and go to the step.

S4: S33. If the threshold is exceeded, triggering content distribution between the local media servers. Usually, a certain media content is more frequently requested by the client, and the source of the media is hotter, resulting in more than The pre-set distribution threshold will trigger the data distribution operation in the local area, so that backups can be stored between multiple local media servers. When the client requests this media, it is respectively set by different media servers. In response, it can play a role in load balancing. FIG. 5 is a flowchart of an address list providing method according to an embodiment of the present invention. As shown in FIG. 5, the address list providing method includes the following steps: Step S501: The index server checks whether there is a medium capable of providing content requested by the client. The server or other client address list, if yes, go to step S502, otherwise, go to step S503; step S502, directly return the address list of the media server or the other client to the client; Step S503: The index server sends a query request to another index server, and the server receives the address list returned by the other server and returns it to the client. According to an embodiment of the present invention, there is also provided a computer readable medium having stored thereon computer executable instructions for causing a computer or processor to perform, for example, when executed by a computer or processor The processing of the steps shown in Figs. 3, 4, and 5, optionally, one or more of the above-described embodiments may be performed. In addition, the implementation of the present invention does not modify the system architecture and the current processing flow, is easy to implement, facilitates promotion in the technical field, and has strong industrial applicability. The above is a further detailed description of the present invention in connection with the specific preferred embodiments, and the specific embodiments of the present invention are not limited to the description. It will be apparent to those skilled in the art that the present invention can be made without departing from the spirit and scope of the invention.

Claims

A P2P network includes an index server and a media server, wherein the index server is configured to respond to a media content query request of the client and return a query result, where the media server is configured to store media content and accept management of the index server. The index server includes a remote query module, configured to send a query request to other index servers and receive the query result when the client query content is missing. The P2P network according to claim 1, wherein: the remote query module comprises a DHT transceiver module for querying other cable servers through the DHT network. The P2P network according to claim 1, wherein: the index server further comprises a return control module, configured to control a media server or other client address list returned by the query, so that other clients in the same network in the list Or the location of the media server is preceded by other media servers or clients in the same region. The P2P network according to any one of claims 1 to 3, wherein: the server further comprises a missing counting module and a copy triggering module, wherein the missing counting module is configured to count the index server The number of times the same query content is missing, when the number exceeds a preset threshold, the copy triggering module triggers the media server managed by the index server to issue a content request to the other media server and copy the same query content. The P2P network according to any one of claims 1 to 3, wherein: the index server further comprises a query counting module and a load balancing module, wherein the query counting module accumulates the number of times the same media content is requested to be queried, When the number of times exceeds a preset threshold, the load balancing module distributes the same content between multiple media servers managed by the index server, and subsequent multiple query requests for the same content will be specified. A different media server response to get the distribution. A P2P network scheduling method includes the following steps: A. Client login, returning the index server address; B. The client queries the index server for the media server where the requested content is located; C. The index server returns the media server with the requested content and The address list of the other client; D, the client sends a request content message to the media server or other client that has the content requested; E, the media server or other client returns the requested media content; and the feature C is that the step C further includes the following step: CI, if the index server has a media server or other client address list that the client requests content, directly return the list, and then go to step D, otherwise perform the following steps;

C2. The index server sends a query request to another index server;

C3. The server 1 receives the address list returned by the other server 1 and returns it to the client.

The P2P network scheduling method according to claim 6, wherein the query request in the step C2 is sent through the DHT network.

The P2P network scheduling method according to claim 6, wherein: the address list of the step C is to be preceded by other media servers in the same area according to other clients or media servers in the same network. Or the order of the client.

The P2P network scheduling method according to claim 6, wherein the step C3 further comprises the following steps:

C31. Accumulating the number of missing requests for the requested content;

C32. When the number of deletions exceeds a preset threshold, the index server triggers the media server managed by the index server to perform data replication on the media server returned by the other index servers.

The P2P network scheduling method according to claim 6, wherein the step C1 further comprises the following steps:

C11. The number of requests for accumulating the requested content;

C12. When the number of requests exceeds a preset threshold, the index server requests the query content to be distributed among multiple media servers managed by the index server, and subsequent multiple query requests for the same content will be specified. A different media server response to get the distribution.

The P2P network scheduling method according to any one of claims 7-11, wherein: the step A returns a plurality of index server addresses that are geographically close, and the default cable server is selected by default. Step B is to send a request to the default cable 1 server.

12. A method for providing an address list, comprising the steps of:

A, the index server checks whether there is a media server or other client address list capable of providing the content requested by the client, if yes, step B is performed, otherwise, step C is performed; B. Returning, to the client, the address list of the media server or the other client directly;

C. The index server sends a query request to other index servers, and the index server receives the address list returned by the other server and returns it to the client.