WO2018018490A1 - 访问分发方法、装置及系统 - Google Patents

访问分发方法、装置及系统 Download PDF

Info

Publication number
WO2018018490A1
WO2018018490A1 PCT/CN2016/092026 CN2016092026W WO2018018490A1 WO 2018018490 A1 WO2018018490 A1 WO 2018018490A1 CN 2016092026 W CN2016092026 W CN 2016092026W WO 2018018490 A1 WO2018018490 A1 WO 2018018490A1
Authority
WO
WIPO (PCT)
Prior art keywords
request message
access request
data access
indication information
server
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.)
Ceased
Application number
PCT/CN2016/092026
Other languages
English (en)
French (fr)
Inventor
李侠
Current Assignee (The listed assignees may be inaccurate. Google has not performed a legal analysis and makes no representation or warranty as to the accuracy of the list.)
Cloudminds Shenzhen Robotics Systems Co Ltd
Original Assignee
Cloudminds Shenzhen Robotics Systems Co Ltd
Priority date (The priority date 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 date listed.)
Filing date
Publication date
Application filed by Cloudminds Shenzhen Robotics Systems Co Ltd filed Critical Cloudminds Shenzhen Robotics Systems Co Ltd
Priority to PCT/CN2016/092026 priority Critical patent/WO2018018490A1/zh
Priority to CN201680002768.XA priority patent/CN106716974B/zh
Publication of WO2018018490A1 publication Critical patent/WO2018018490A1/zh
Anticipated expiration legal-status Critical
Ceased legal-status Critical Current

Links

Images

Classifications

    • HELECTRICITY
    • H04ELECTRIC COMMUNICATION TECHNIQUE
    • H04LTRANSMISSION OF DIGITAL INFORMATION, e.g. TELEGRAPHIC COMMUNICATION
    • H04L67/00Network arrangements or protocols for supporting network services or applications
    • H04L67/01Protocols
    • H04L67/10Protocols in which an application is distributed across nodes in the network
    • H04L67/1001Protocols in which an application is distributed across nodes in the network for accessing one among a plurality of replicated servers

Definitions

  • the present invention relates to the field of Internet technologies, and in particular, to an access distribution method, apparatus, and system.
  • the application client sends a data access request message to the source server to obtain the corresponding content.
  • the video application can send a data access request message to the source server through the computer to obtain the content.
  • the corresponding video content is played and played.
  • the number of corresponding data access requests is also very large. In this way, it is generally necessary to deploy multiple source server servers to process these data access requests separately.
  • the Application Delivery Network (ADN) system can accelerate the application directly to the edge of the network, or even the terminal, to realize the distribution of various applications.
  • ADN Application Delivery Network
  • access distribution based on the ADN system has the following technical problems: in the prior art, data access request messages are generally randomly or evenly distributed to various source station servers. on. However, the actual situation is that each source server has different processing capabilities for data access request messages. In this way, the processing capability of each source station server does not match the number of allocated data access request messages; some source server may not process the allocated data access request message in time due to weak processing capability, affecting the user experience. On the other hand, some of the processing resources of the source server with strong processing power are not fully utilized, resulting in waste of resources.
  • Embodiments of the present invention propose an access distribution method, apparatus, and system such that the processing capabilities of respective source station servers match the number of allocated data access request messages.
  • an embodiment of the present invention provides an access distribution method, including:
  • first indication information corresponding to the application from the central server, where the first indication information is used to indicate processing capability of each source server of the plurality of source server servers corresponding to the application to the data access request message ;
  • the first indication information is carried in the data access request message and sent to the egress server.
  • the determining to initiate the application of the data access request message comprises:
  • the method before sending the data access request message to the egress server, the method further includes:
  • an embodiment of the present invention provides an access distribution method, including:
  • the first application is an application that initiates the data access request message
  • each data access request message initiated by the first application to a corresponding source station server includes:
  • the generating a weight-based mapping relationship according to the first indication information includes:
  • resource usage information of the plurality of source station servers where the resource usage rate information is used to indicate resource usage rate of each source station server;
  • an embodiment of the present invention provides an access distribution method, including:
  • first indication information corresponding to each application of the multiple applications where the first indication information is used to indicate that each source server of the plurality of source station servers corresponding to the application has processing capability for the data access request message ;
  • the first indication information is sent to a plurality of portal servers.
  • an access distribution apparatus including:
  • a first receiving module configured to receive a data access request message
  • a determining module configured to determine an application that initiates the data access request message
  • a first obtaining module configured to acquire, from a central server, first indication information corresponding to the application, where the first indication information is used to indicate each source server pair of the plurality of source station servers corresponding to the application The processing capability of the data access request message;
  • the first sending module is configured to carry the first indication information in the data access request message and send the message to the egress server.
  • the determining module comprises:
  • the first sending module further includes:
  • an access distribution apparatus including:
  • a second receiving module configured to receive a data access request message carrying the first indication information, where the first indication information is used to indicate that each source server of the plurality of source station servers corresponding to the first application has a data access request The processing capability of the message, where the first application is an application that initiates the data access request message;
  • a second sending module configured to send, according to the first indication information, each data access request message initiated by the first application to a corresponding source station server.
  • the second sending module includes:
  • the generating a weight-based mapping relationship according to the first indication information includes:
  • resource usage information of the plurality of source station servers where the resource usage rate information is used to indicate resource usage rate of each source station server;
  • an access distribution apparatus including:
  • a second obtaining module configured to acquire first indication information corresponding to each of the plurality of applications, where the first indication information is used to represent each source server pair of the plurality of source station servers corresponding to the application The processing capability of the data access request message;
  • the third sending module is configured to send the first indication information to the multiple portal servers.
  • an access distribution system including:
  • An ingress server configured to receive a data access request message, determine an application that initiates the data access request message, and obtain first indication information corresponding to the application from a central controller, where the first indication information is used to indicate The processing capability of each of the plurality of source server servers corresponding to the application to the data access request message, and the first indication information being carried in the data access request message and sent to the egress server;
  • a central controller configured to acquire first indication information corresponding to the application, and send the first indication information to the multiple access servers;
  • the egress server is configured to receive a data access request message carrying the first indication information, and send the data access request message initiated by the application to the corresponding source station server according to the first indication information.
  • the first indication information is carried in the data access request message and sent to the egress server for the data access request message initiated by the application; the first indication information is used to indicate that the application corresponds to the Each source server of each of the source station servers processes the data access request message; and enables the egress server to send each data access request message initiated by the application to the corresponding source station server according to the first indication information.
  • the number of data access request messages received by each source server can be matched to the processing capabilities of the source server. Avoid situations where some source server's processing resources are idle and others are unable to process data access request messages in a timely manner.
  • FIG. 1 is a schematic diagram of a principle of an access server side access distribution method according to Embodiment 1 of the present invention
  • FIG. 2 is a schematic diagram of a GRP protocol format of an access server side access distribution method according to Embodiment 1 of the present invention
  • FIG. 3 is a schematic diagram of a principle of an access server side access distribution method according to Embodiment 1 of the present invention.
  • FIG. 4 is a schematic diagram of a principle of a central server side access distribution method according to Embodiment 1 of the present invention.
  • FIG. 5 is a schematic structural diagram of an access server side access distribution device according to Embodiment 2 of the present invention.
  • FIG. 6 is a schematic structural diagram of an access server side access distribution device according to Embodiment 2 of the present invention.
  • FIG. 7 is a schematic structural diagram of a central server side access distribution device according to Embodiment 2 of the present invention.
  • FIG. 8 is a schematic structural diagram of an access distribution system according to Embodiment 3 of the present invention.
  • the access distribution method, device and system provided by the present invention can be applied to a distribution network to effectively improve the response speed of the source station server.
  • a typical distribution network includes an ingress server and an egress server. The ingress server and the egress server are set at the edge of the network, and the network data entered from the ingress server can be quickly directed to the egress server to reach the edge of the network where the destination terminal or the destination terminal is located, thereby realizing rapid distribution of network data.
  • the embodiment of the present invention provides that the portal server inserts the first indication information into the data access request message, so that the egress server sends the data access request message initiated by the first application to the corresponding source station server according to the first indication information, so that each of the The number of data access request messages received by the source station server matches the processing capability of the source station server, thereby implementing load balancing of the source station server, which will be described below.
  • Embodiment 1 is a diagrammatic representation of Embodiment 1:
  • FIG. 1 is a schematic diagram showing the principle of an access server side access distribution method in the first embodiment of the present invention. As shown in FIG. 1 , the method includes:
  • Step 101 Receive a data access request message.
  • Step 102 Determine an application that initiates the data access request message.
  • Step 103 Obtain first indication information corresponding to the application from a central server, where the first indication information is used to indicate a data access request message for each source server of the plurality of source server servers corresponding to the application. Processing power.
  • Step 104 The first indication information is carried in the data access request message and sent to the egress server.
  • step 102 after the portal server receives the data access request message from the user, the application that initiates the data access request message is determined according to the application identifier in the data access request message.
  • the central server sends a plurality of first indication information belonging to each application to the portal server in advance, so that the portal server obtains the determined application from the plurality of the first indication information according to the determined application.
  • Matching first indication information wherein the first indication
  • the information includes at least an identifier of the plurality of source station servers corresponding to the determined application, and weight information of each of the plurality of source station servers.
  • step 104 the ingress server inserts the obtained first indication information that matches the determined application into the data access request message, and sends the data access request message carrying the first indication message to the egress server. .
  • the first indication information is included in the path optimization policy, so that the egress server can send the data access request message to the corresponding source station server according to the first indication information, that is, the source station server based on the load balancing of the source station server.
  • the response speed thus optimizing the global path. It is not difficult to understand that the sending of the data access request message to the corresponding source station server refers to pointing the destination end in the data access request message to the corresponding source station server, so that the network forwarding device can access the data.
  • the request message is sent to the corresponding source server.
  • the step of determining an application for initiating the data access request message comprises:
  • the ingress server parses the data access request message at the transport layer to obtain an application identifier in the data access request message, that is, determines an application that initiates the data access request message, so that the ingress server sends the input to the portal from the central server according to the determined application identifier.
  • a matching application identifier is found in the mapping between the application identifier of the server and the first indication information, and the first indication information corresponding to the application identifier is obtained.
  • the application identifier is an identifier of an application that initiates the data access request message, and the application identifier may be an application identifier, such as a WeChat client identifier, or a website identifier, for example, a uniform resource locator (URL: Uniform Resoure) Locator) logo and so on.
  • an application identifier such as a WeChat client identifier, or a website identifier, for example, a uniform resource locator (URL: Uniform Resoure) Locator) logo and so on.
  • URL Uniform Resoure
  • the method before the step of sending the data access request message to the egress server, the method further includes:
  • the Global Routing Protocol (GRP) set in the first embodiment of the present invention is a binary stream, hereinafter referred to as GRP, and the global path system is at the transport layer, and is used to define an optimal path format between the ingress server and the egress server, that is, Binary stream by IP address and / or port number
  • GRP Global Routing Protocol
  • the definition can effectively shorten the length of the TCP data segment and increase the usage rate of the network bandwidth.
  • the data access request message is captured, the data access request message is parsed by using a preset GRP parsing method to obtain the application identifier in the data access request message.
  • FIG. 2 is a schematic diagram of a GRP protocol format of an access distribution method according to Embodiment 1 of the present invention.
  • the GRP protocol format definition is specifically:
  • the inserted TCP data segment includes header + data, and the header portion includes a strictly defined 8-byte version number, a 2-byte information type, and a 2-byte data stream length, and the data portion includes bytes.
  • Variable data flow As shown in FIG. 2B, the specific format of the data portion in the TCP data segment inserted in FIG. 2A, that is, the data stream format of the optimal path information, the optimal path data stream includes a plurality of node address units.
  • the specific format of the single node address unit in FIG. 2B includes a 4-bit role, a 4-bit protocol family, a 1-byte weight, a 2-byte port number, and an IP address of 4-16 bytes.
  • the binary stream is defined according to the GRP protocol standard, such as the IP address and the port number.
  • GRP protocol standard such as the IP address and the port number.
  • the specific format is:
  • the header part is in order, the version number is GRP1.001, the information type is 0001, and the data stream length is 000f; the first hop node information is in order, the role is 1, the protocol family is 2, and the weight is 04.
  • the port number is 0050 and the IP address is 0d00000a.
  • the second hop node information is in order, the role is 3, the protocol family is 2, the weight is 02, the port number is 0050, and the IP address is 0d12110b.
  • the binary stream is defined according to the GRP protocol standard, such as the IP address and the port number.
  • GRP protocol standard such as the IP address and the port number.
  • the specific format is:
  • the header part is in order, the version number is GRP1.001, the information type is 0001, and the data stream length is 001c; the first hop node information is in order, the role is 1, the protocol family is a, and the weight is 01. The port number is 0050, and the IP address is fe800000000000020c29fffe5841bc. Similarly, the second hop node information is in order, the role is 3, the protocol family is 2, the weight is 02, the port number is 0050, and the IP address is 0d12110b.
  • the global path information including the first indication information is inserted into the data access request message according to the GRP protocol, and the specific format is:
  • the client socket + access node + intermediate node + egress node is the inserted global path information including the first indication information defined according to the GRP protocol.
  • the access distribution method in the first embodiment of the present invention is applicable to the IPv4 and IPv6 protocols.
  • IPv4 protocol when the accessed path requires 8 hops to reach the source server, the existing data access request message does not have a standard binary stream definition.
  • the TCP data segment is performed according to the standard protocol specified by the present invention.
  • FIG. 3 is a schematic diagram showing the principle of the access server side access distribution method in the first embodiment of the present invention, as shown in FIG. 3, including:
  • Step 301 Receive a data access request message carrying the first indication information, where the first indication information is used to indicate that each source server of the plurality of source server servers corresponding to the first application has processing capability for the data access request message.
  • the first application is an application that initiates the data access request message.
  • Step 302 Perform, according to the first indication information, each data access initiated by the first application.
  • the request message is sent to the corresponding source server.
  • the egress server receives the data access request message that is sent by the ingress server and carries the first indication information, and parses the first indication information by using the GRP parsing method; performs calculation on the parsed first indication, and determines the data access request.
  • the source server that matches the message and sends a data access request message to the source server.
  • the step of transmitting, according to the first indication information, each data access request message initiated by the first application to a corresponding source station server comprises:
  • the data access request message from the user can generally correspond to multiple source station IP addresses, that is, multiple source station servers, and the processing capabilities of the respective source station servers are different (eg, each source station)
  • the processing capability of the server itself, or the processing capability of the current time, results in different concurrency capabilities of the source servers; when the egress server connects to the source server, a weight-based mapping relationship is generated according to the first indication information, for multiple sources.
  • the station server performs reasonable connection allocation, thereby improving the load balancing capability of multiple source server.
  • the source station server with strong default processing capability has a larger weight value, and vice versa.
  • the specific insertion format of the first indication information is:
  • source station server 1 weight 3
  • the egress server parses the received data access request message, obtains the first indication information, and the user IP address, and uses the weight-based hash (WSH: Weight Source Hash) algorithm to determine the target according to the user IP address (ie, the source address).
  • WSH Weight Source Hash
  • the source station server, the WSH algorithm is a load balancing algorithm for multi-source station server weight information.
  • the egress server establishes multiple hash tables based on the application identifiers, that is, according to multiple source server identifiers and weight information belonging to the same application identifier, and fills each hash table by using a padding algorithm.
  • the Greek table is used to indicate the correspondence between the user IP address and the source server identifier.
  • the user IP address in the corresponding relationship exists in the form of an index key, that is, the hash value calculated by using the mapping relationship is an index key according to the user IP address.
  • the second index key is calculated according to the parsed user IP address, the first index key that is the same as the second index key is searched by using the established hash table, and the second index key is determined according to the first index key.
  • the matching source station server identifier, the determined source station server is used as the target source station server, and the data access request message is sent to the target source station server.
  • the source station server when the same user continues to access the previous application identifier, the source station to which the data access request message is to be accessed
  • the server is the same, regardless of the weight information, ensuring that the data access request message of the same user can access the same source server and maintain the session state. It is only necessary to implement load balancing of data access based on the weight information of the matched source station server when the source server matching the application that the user wants to access changes.
  • the egress server uses the WSH algorithm to determine the target source station server, that is, for all source server servers that match an application identifier, a hashing algorithm is used to generate a hash table, and an index key is calculated for the user IP address in the data access request message to generate a second index key; determining, according to the second index key, the target source server using the generated hash table; and transmitting the data access request message to the determined target source server.
  • the step of generating a weight-based mapping relationship according to the first indication information includes:
  • resource usage information of the plurality of source station servers where the resource usage rate information is used to indicate resource usage rate of each source station server;
  • the specific method for the egress table to fill the hash table by using the padding algorithm is that, based on the weight information of the plurality of source station servers of an application, the source station server with relatively less CPU usage is selected to fill the hash table each time.
  • the weight ratio of each source server is:
  • the source station server 2 is selected to perform the filling of the hash table.
  • the GRP protocol in the first embodiment of the present invention supports a multi-source server, which can effectively solve the problem that only a single source server is supported, and the connection fails, that is, the access fails. After the egress server fails to connect according to the calculated target source server, Ability to select other alternate source server to connect to improve the success rate of data access.
  • FIG. 4 is a schematic diagram showing the principle of the central server side access distribution method in the first embodiment of the present invention. As shown in FIG. 4, the method includes:
  • Step 401 Acquire first indication information corresponding to each application of the multiple applications, where the first indication information is used to indicate a data access request message for each source server of the plurality of source server servers corresponding to the application. Processing capacity;
  • Step 402 Send the first indication information to multiple portal servers.
  • the central server generates a message including the correspondence between the application identifier and the first indication information by sending the first indication information that matches each application, and sends the message to the portal server, so that the portal server can parse after receiving the data access request message.
  • the data access request message acquires an application identifier of the initiating data access request message, and obtains first indication information that matches the application according to the parsed application identifier.
  • Embodiment 2 is a diagrammatic representation of Embodiment 1:
  • an access distribution device is also provided in the embodiment of the present invention. Since the principle of solving the problem of these devices is similar to an access distribution method, the implementation of these devices can refer to the implementation of the method, and the repetition is no longer Narration.
  • FIG. 5 is a schematic structural diagram of an access server side access distribution device according to Embodiment 2 of the present invention.
  • the portal server side access distribution device 500 may include:
  • the first receiving module 501 is configured to receive a data access request message.
  • the determining module 502 is configured to determine an application that initiates the data access request message.
  • the obtaining module 503 is configured to obtain first indication information corresponding to the application from the central server, where the first indication information is used to represent data of each source server of the plurality of source server servers corresponding to the application. The processing power of the access request message.
  • the first sending module 504 is configured to carry the first indication information in the data access request message and send the message to the egress server.
  • the determining module 502 includes:
  • the first sending module 503 further includes:
  • FIG. 6 is a schematic structural diagram of an access server side access distribution device according to Embodiment 2 of the present invention. As shown in FIG. 6, the egress server side access distribution device 600 may include:
  • the second receiving module 601 is configured to receive a data access request message carrying the first indication information, where the first indication information is used to indicate that each source server of the plurality of source station servers corresponding to the first application accesses the data The processing capability of the request message, the first application being an application that initiates the data access request message.
  • the second sending module 602 is configured to send, according to the first indication information, each data access request message initiated by the first application to a corresponding source station server.
  • the second sending module 602 includes:
  • the step of generating a weight-based mapping relationship according to the first indication information includes:
  • Obtaining resource usage information of the plurality of source station servers, where the resource usage information is used Represents the resource usage of each source server;
  • FIG. 7 is a schematic structural diagram of a central server side access distribution device according to a second embodiment of the present invention.
  • the access distribution device of the central server side 700 includes:
  • the second obtaining module 701 is configured to obtain first indication information corresponding to each application of the multiple applications, where the first indication information is used to indicate each of the plurality of source server servers corresponding to the application. The ability to process data access request messages.
  • the third sending module 702 is configured to send the first indication information to multiple ingress servers.
  • Embodiment 3 is a diagrammatic representation of Embodiment 3
  • an access distribution system is also provided in the embodiment of the present invention. Since the principle of solving the problem of these devices is similar to an access distribution method and an access distribution device, the implementation of these devices can refer to the implementation of the method. , the repetition will not be repeated.
  • FIG. 8 is a schematic diagram showing the architecture of the access distribution system in the third embodiment of the present invention. As shown in FIG. 8, the access distribution system 800 may include:
  • the portal server 801 is configured to receive a data access request message, determine an application that initiates the data access request message, and obtain, from the central controller 802, first indication information corresponding to the application, where the first indication information is used by And indicating that the source station server of the application corresponds to the processing capability of the data access request message, and the first indication information is carried in the data access request message and sent to the egress server 803.
  • the central controller 802 is configured to acquire first indication information corresponding to the application, and send the first indication information to the portal server 801.
  • the egress server 803 receives a data access request message carrying the first indication information, according to the The first indication information is sent by the application-initiated data access request message to the corresponding source station server.
  • the manner of obtaining the first indication information includes static acquisition and dynamic acquisition.
  • the method for obtaining the first indication information in the static manner is that the central server 802 enters the first indication information corresponding to the application by using the WEB interface, and generates a correspondence relationship table between the application identifier and the first indication information, and generates the corresponding correspondence table in real time.
  • the portal server 801 is sent to the portal server 801, so that the portal server obtains the first indication information corresponding to the application identifier according to the application identifier in the data access request message.
  • the method for obtaining the first indication information in a dynamic manner is that the central server obtains the CPU usage rate of the source station server in advance through the SNMP protocol, and calculates a new source station server weight according to the CPU usage rate of the source station server, and generates a new one. Applying a correspondence table between the identifier and the first indication information, and generating the generated new correspondence table in real time to the portal server, so that the portal server implements real-time update of the first indication information.
  • the central server receives CPU usage information of two source server servers of an application, and the CPU usage rates of the source station server 1 and the source station server 2 are 40% and 80%, respectively, and the source station server 1 and the source station server
  • the weight ratio of 2 is 2:1, that is, the ratio of the weight of the source station server is inversely proportional to the CPU usage of the source server. If the CPU usage of a source station server is greater than or equal to 95%, the source station server weight may be set to the lowest, or a new data access request message may not be directly allocated to the source station server.
  • embodiments of the present invention can be provided as a method, system, or computer program product. Accordingly, the present invention may take the form of an entirely hardware embodiment, an entirely software embodiment, or a combination of software and hardware. Moreover, the invention can take the form of a computer program product embodied on one or more computer-usable storage media (including but not limited to disk storage, CD-ROM, optical storage, etc.) including computer usable program code.
  • computer-usable storage media including but not limited to disk storage, CD-ROM, optical storage, etc.
  • the computer program instructions can also be stored in a computer readable memory that can direct a computer or other programmable data processing device to operate in a particular manner, such that the instructions stored in the computer readable memory produce an article of manufacture comprising the instruction device.
  • the apparatus implements the functions specified in one or more blocks of a flow or a flow and/or block diagram of the flowchart.
  • These computer program instructions can also be loaded onto a computer or other programmable data processing device such that a series of operational steps are performed on a computer or other programmable device to produce computer-implemented processing for execution on a computer or other programmable device.
  • the instructions provide steps for implementing the functions specified in one or more of the flow or in a block or blocks of a flow diagram.

Landscapes

  • Engineering & Computer Science (AREA)
  • Computer Networks & Wireless Communication (AREA)
  • Signal Processing (AREA)
  • Computer And Data Communications (AREA)
  • Data Exchanges In Wide-Area Networks (AREA)

Abstract

本发明提供了访问分发方法、装置及系统,所述方法包括:接收数据访问请求消息;确定发起所述数据访问请求消息的应用;从中央服务器中获取所述应用所对应的第一指示信息,所述第一指示信息用于表示所述应用所对应的多个源站服务器中每一个源站服务器对数据访问请求消息的处理能力;将所述第一指示信息携带在所述数据访问请求消息中发送给出口服务器。本发明基于数据访问请求消息中的第一指示信息,快速确定与数据访问请求消息相匹配的源站服务器,实现源站服务器的负载均衡。

Description

访问分发方法、装置及系统 技术领域
本发明涉及互联网技术领域,特别涉及访问分发方法、装置及系统。
背景技术
用户在使用互联网应用时,一般是通过安装在终端上的应用客户端向源站服务器发送数据访问请求消息以获取相应的内容,比如视频应用可以通过电脑向源站服务器发送数据访问请求消息以获取相应的视频内容并播放。对于用户量比较大的应用来说,其对应的数据访问请求的数量也是非常巨大的。这样,一般需要部署多台源站服务器,对这些数据访问请求分别进行处理。随着互联网技术的发展,应用分发网络(ADN:Application Delivery Network)系统能够实现将应用直接加速到网络边缘,甚至是终端,实现各类应用的分发。
在实现本发明的过程中,本发明的发明人发现现有技术中,基于ADN系统的访问分发存在如下技术问题:现有技术中一般是将数据访问请求消息随机或者平均分发到各个源站服务器上。但是实际情况是,各个源站服务器对数据访问请求消息的处理能力不同。这样就造成各个源站服务器的处理能力与所分配的数据访问请求消息的数量不匹配;导致一些源站服务器因为处理能力较弱不能及时的对所分配的数据访问请求消息进行处理,影响用户体验,而另一方面,一些处理能力较强的源站服务器的处理资源又得不到充分的利用,造成了资源浪费。
发明内容
本发明实施例提出了访问分发方法、装置及系统,以使得各个源站服务器的处理能力与所分配的数据访问请求消息的数量相匹配。
在一个方面,本发明实施例提供了一种访问分发方法,包括:
接收数据访问请求消息;
确定发起所述数据访问请求消息的应用;
从中央服务器中获取所述应用所对应的第一指示信息,所述第一指示信息用于表示所述应用所对应的多个源站服务器中每一个源站服务器对数据访问请求消息的处理能力;
将所述第一指示信息携带在所述数据访问请求消息中发送给出口服务器。
优选的,所述确定发起所述数据访问请求消息的应用,包括:
根据所述数据访问请求消息的传输层协议信息中的应用标识确定发起所述数据访问请求消息的应用。
优选的,在将数据访问请求消息发送给出口服务器之前,还包括:
对数据访问请求消息中的IP地址和/或端口号进行二进制整型。
在一个方面,本发明实施例提供了一种访问分发方法,包括:
接收携带有第一指示信息的数据访问请求消息,所述第一指示信息用于表示第一应用所对应的多个源站服务器中每一个源站服务器对数据访问请求消息的处理能力,所述第一应用为发起所述数据访问请求消息的应用;
根据所述第一指示信息将由所述第一应用发起的各个数据访问请求消息发送到对应的源站服务器上。
优选的,所述根据所述第一指示信息将由所述第一应用发起的各个数据访问请求消息发送到对应的源站服务器上,包括:
根据所述第一指示信息生成基于权重的映射关系;
利用生成的映射关系对待发送的数据访问请求消息的源地址进行运算;
将待发送的数据访问请求消息发送到所得到的运算结果所对应的源站服务器。
优选的,所述根据所述第一指示信息生成基于权重的映射关系,包括:
获取所述多个源站服务器的资源使用率信息,所述资源使用率信息用于表示每一个源站服务器的资源使用率;
根据所述第一指示信息以及所述资源使用率信息生成基于权重的映射关 系。
在一个方面,本发明实施例提供了一种访问分发方法,包括:
获取多个应用中的每一个应用对应的第一指示信息,所述第一指示信息用于表示所述应用所对应的多个源站服务器中每一个源站服务器对数据访问请求消息的处理能力;
将第一指示信息发送给多个入口服务器。
在另一个方面,本发明实施例提供了一种访问分发装置,包括:
第一接收模块,用于接收数据访问请求消息;
确定模块,用于确定发起所述数据访问请求消息的应用;
第一获取模块,用于从中央服务器中获取所述应用所对应的第一指示信息,所述第一指示信息用于表示所述应用所对应的多个源站服务器中每一个源站服务器对数据访问请求消息的处理能力;
第一发送模块,用于将所述第一指示信息携带在所述数据访问请求消息中发送给出口服务器。
优选的,所述确定模块包括:
根据所述数据访问请求消息的传输层协议信息中的应用标识确定发起所述数据访问请求消息的应用。
优选的,所述第一发送模块还包括:
对数据访问请求消息中的IP地址和/或端口号进行二进制整型。
在另一个方面,本发明实施例提供了一种访问分发装置,包括:
第二接收模块,用于接收携带有第一指示信息的数据访问请求消息,所述第一指示信息用于表示第一应用所对应的多个源站服务器中每一个源站服务器对数据访问请求消息的处理能力,所述第一应用为发起所述数据访问请求消息的应用;
第二发送模块,用于根据所述第一指示信息将由所述第一应用发起的各个数据访问请求消息发送到对应的源站服务器上。
优选的,所述第二发送模块包括:
根据所述第一指示信息生成基于权重的映射关系;
利用生成的映射关系对待发送的数据访问请求消息的源地址进行运算;
将待发送的数据访问请求消息发送到所得到的运算结果所对应的源站服务器。
优选的,所述根据所述第一指示信息生成基于权重的映射关系,包括:
获取所述多个源站服务器的资源使用率信息,所述资源使用率信息用于表示每一个源站服务器的资源使用率;
根据所述第一指示信息以及所述资源使用率信息生成基于权重的映射关系。
在另一个方面,本发明实施例提供了一种访问分发装置,包括:
第二获取模块,用于获取多个应用中的每一个应用对应的第一指示信息,所述第一指示信息用于表示所述应用所对应的多个源站服务器中每一个源站服务器对数据访问请求消息的处理能力;
第三发送模块,用于将第一指示信息发送给多个入口服务器。
在另一个方面,本发明实施例提供了一种访问分发系统,包括:
入口服务器,用于接收数据访问请求消息,确定发起所述数据访问请求消息的应用,并从中央控制器中获取所述应用所对应的第一指示信息,所述第一指示信息用于表示所述应用所对应的多个源站服务器中每一个源站服务器对数据访问请求消息的处理能力,以及将所述第一指示信息携带在所述数据访问请求消息中发送给出口服务器;
中央控制器,用于获取所述应用对应的第一指示信息,并将第一指示信息发送给多个所述入口服务器;
出口服务器,用于接收携带有第一指示信息的数据访问请求消息,根据所述第一指示信息将由所述应用发起的数据访问请求消息发送到对应的源站服务器上。
本申请的有益效果如下:
本实施例中,针对一个应用发起的数据访问请求消息,将第一指示信息携带在所述数据访问请求消息中发送给出口服务器;所述第一指示信息用于表示所述应用所对应的多个源站服务器中每一个源站服务器对数据访问请求消息的处理能力;使得出口服务器能够根据第一指示信息将由所述应用发起的各个数据访问请求消息发送到对应的源站服务器上。这样,就可以使得每一个源站服务器所接收的数据访问请求消息的数量与该源站服务器的处理能力相匹配。避免一些源站服务器的处理资源闲置而另外一些源站服务器无法及时处理数据访问请求消息的情况。
附图说明
下面将参照附图描述本发明的具体实施例,其中:
图1为本发明实施例一中入口服务器侧访问分发方法的原理示意图;
图2为本发明实施例一中入口服务器侧访问分发方法的GRP协议格式示意图;
图3为本发明实施例一中出口服务器侧访问分发方法的原理示意图;
图4为本发明实施例一中中央服务器侧访问分发方法的原理示意图;
图5为本发明实施例二中入口服务器侧访问分发装置的结构示意图;
图6为本发明实施例二中出口服务器侧访问分发装置的结构示意图;
图7为本发明实施例二中中央服务器侧访问分发装置的结构示意图;
图8为本发明实施例三中访问分发系统架构示意图。
具体实施方式
为了使本发明的技术方案及优点更加清楚明白,以下结合附图对本发明的示例性实施例进行进一步详细的说明,显然,所描述的实施例仅是本发明的一部分实施例,而不是所有实施例的穷举。并且在不冲突的情况下,本发明中的实施例及实施例中的特征可以互相结合。
本发明提供的访问分发方法、装置及系统,可以应用于分发网络中,以有效提高源站服务器的响应速度。一般的分发网络包括入口服务器(ingress)和出口服务器(egress)。入口服务器和出口服务器设置在网络的边缘,从入口服务器进入的网络数据能够快速的被导向出口服务器,到达目的终端或者目的终端所在的网络边缘,实现网络数据的快速分发。
本发明实施例提出了入口服务器将第一指示信息插入到数据访问请求消息中,使得出口服务器根据第一指示信息将由第一应用发起的数据访问请求消息发送给对应的源站服务器,使得每一个源站服务器所接收的数据访问请求消息的数量与该源站服务器的处理能力相匹配,从而实现源站服务器的负载均衡,下面进行说明。
为了便于本发明的实施,下面以实例进行说明。
实施例一:
图1示出了本发明实施例一中入口服务器侧访问分发方法的原理示意图,如图1所示,包括:
步骤101、接收数据访问请求消息。
步骤102、确定发起所述数据访问请求消息的应用。
步骤103、从中央服务器中获取所述应用所对应的第一指示信息,所述第一指示信息用于表示所述应用所对应的多个源站服务器中每一个源站服务器对数据访问请求消息的处理能力。
步骤104、将所述第一指示信息携带在所述数据访问请求消息中发送给出口服务器。
在步骤102中,当入口服务器接收到来自用户的数据访问请求消息后,根据数据访问请求消息中的应用标识确定发起所述数据访问请求消息的应用。
在步骤103中,中央服务器会预先将多个属于每一个应用的第一指示信息发送给入口服务器,以便入口服务器根据所确定的应用从多个所述第一指示信息中获取与所确定的应用相匹配的第一指示信息。其中,所述第一指示 信息至少包括所确定的应用所对应的多个源站服务器的标识,以及该多个源站服务器中每个源站服务器的权重信息。
在步骤104中,入口服务器将获取到的与所确定的应用相匹配的第一指示信息插入到数据访问请求消息中,并将携带有所述第一指示消息的数据访问请求消息发送给出口服务器。
本实施例通过将第一指示信息纳入到路径优化策略中,使得出口服务器能够根据第一指示信息将数据访问请求消息发送给对应的源站服务器,即基于源站服务器的负载均衡提高源站服务器的响应速度,从而实现全局路径的优化。不难理解的是,这里所指的将数据访问请求消息发送给对应的源站服务器可以是指将数据访问请求消息中的目的端指向对应的源站服务器,使得网络转发设备能够将该数据访问请求消息发送到对应的源站服务器。
优选的,确定发起所述数据访问请求消息的应用的步骤包括:
根据所述数据访问请求消息的传输层协议信息中的应用标识确定发起所述数据访问请求消息的应用。
入口服务器在传输层对数据访问请求消息进行解析,得到数据访问请求消息中的应用标识,即确定发起所述数据访问请求消息的应用,以便入口服务器根据所确定的应用标识从中央服务器发送给入口服务器的应用标识与第一指示信息的对应关系表中找到一致的应用标识,并获取与该应用标识对应的第一指示信息。其中,应用标识为发起所述数据访问请求消息的应用的标识,应用标识可以是应用程序标识,如,微信客户端标识等,也可以是网站标识,如,统一资源定位器(URL:Uniform Resoure Locator)标识等。
优选的,在将数据访问请求消息发送给出口服务器的步骤之前,还包括:
对数据访问请求消息中的IP地址和/或端口号进行二进制整型。
本发明实施例一设定的全局路径协议(GRP:Global Routing Protocol)为二进制流,以下简称为GRP,全局路径系统在传输层,用于定义入口服务器与出口服务器间最优路径的格式,即通过对IP地址和/或端口号进行二进制流 定义,能够有效缩短TCP数据段长度,提高网络带宽的使用率。此外,在对数据访问请求消息进行抓包时,利用预设的GRP解析方法对数据访问请求消息进行解析才能够获取数据访问请求消息中的应用标识。
图2示出了本发明实施例一中访问分发方法的GRP协议格式示意图,GRP协议格式定义具体为:
如图2A所示,所插入的TCP数据段包括头部+数据,头部部分包括严格定义的8字节版本号、2字节信息类型和2字节数据流长度,数据部分包括字节可变的数据流。如图2B所示,为图2A所插入的TCP数据段中数据部分的具体格式,即最优路径信息的数据流格式,最优路径数据流包括多个节点地址单元。如图2C所示,为图2B中单个节点地址单元的具体格式,包括4位角色、4位协议族、1字节权重、2字节端口号和4-16字节的IP地址。
对于IPv4协议,当全局路径信息包括2个节点地址单元时,对IP地址、端口号等依照GRP协议的标准进行二进制流定义,具体格式为:
版本号+信息类型+数据流长度+角色+协议族+权重+端口号+IP地址
为便于说明,以下为16进制示例:
GRP1.001|0001|000f|1204|0050|0d00000a|3202|0050|0d12110b|…
其中,头部部分按照顺序依次为,版本号为GRP1.001,信息类型为0001,数据流长度为000f;第一跳节点信息按照顺序依次为,角色为1,协议族为2,权重为04,端口号为0050,IP地址为0d00000a,同样的,第二跳节点信息按照顺序依次为,角色为3,协议族为2,权重为02,端口号为0050,IP地址为0d12110b。
对于IPv6协议,当全局路径信息包括2个节点地址单元时,对IP地址、端口号等依照GRP协议的标准进行二进制流定义,具体格式为:
版本号+信息类型+数据流长度+角色+协议族+权重+端口号+IP地址
为便于说明,以下为16进制示例说明:
GRP1.001|0001|001c|1a01|0050|fe800000000000020c29fffe5841bc|3202|005 0|0d12110b|…
其中,头部部分按照顺序依次为,版本号为GRP1.001,信息类型为0001,数据流长度为001c;第一跳节点信息按照顺序依次为,角色为1,协议族为a,权重为01,端口号为0050,IP地址为fe800000000000020c29fffe5841bc,同样的,第二跳节点信息按照顺序依次为,角色为3,协议族为2,权重为02,端口号为0050,IP地址为0d12110b。
当入口服务器获取到与数据访问请求消息相匹配的第一指示信息时,依照GRP协议将包含第一指示信息的全局路径信息插入到数据访问请求消息中,具体格式为:
MAC头部+IP头部+TCP头部+SSL/TLS+客户端套接字+接入节点+中间节点+出口节点+源站套接字
其中,客户端套接字+接入节点+中间节点+出口节点即为所插入的依照GRP协议定义的包含第一指示信息的全局路径信息。
本发明实施例一中的访问分发方法适用于IPv4和IPv6协议,以IPv4协议为例,当访问的路径需要8跳到达源站服务器时,现有的数据访问请求消息并没有标准的二进制流定义,插入数据访问请求消息中的TCP数据段(即路径信息)长度为3+22*8+3+1=183字节,带宽占有率较高,依照本发明规定的标准协议对TCP数据段进行二进制流格式定义,使得插入数据访问请求消息中的TCP数据段长度可缩短至12+8*8=76字节,有效降低带宽占有率。
图3示出了本发明实施例一中出口服务器侧访问分发方法的原理示意图,如图3所示,包括:
步骤301、接收携带有第一指示信息的数据访问请求消息,所述第一指示信息用于表示第一应用所对应的多个源站服务器中每一个源站服务器对数据访问请求消息的处理能力,所述第一应用为发起所述数据访问请求消息的应用。
步骤302、根据所述第一指示信息将由所述第一应用发起的各个数据访问 请求消息发送到对应的源站服务器上。
在步骤302中,出口服务器接收入口服务器发送的携带有第一指示信息的数据访问请求消息,通过GRP解析方法解析得到第一指示信息;对解析得到的第一指示进行计算,确定与数据访问请求消息相匹配的源站服务器,并将数据访问请求消息发送给该源站服务器。
优选的,根据所述第一指示信息将由所述第一应用发起的各个数据访问请求消息发送到对应的源站服务器上的步骤包括:
根据所述第一指示信息生成基于权重的映射关系;
利用生成的映射关系对待发送的数据访问请求消息的源地址进行运算;
将待发送的数据访问请求消息发送到所得到的运算结果所对应的源站服务器。
在进行互联网数据访问时,来自用户的数据访问请求消息通常能够对应多个源站IP地址,即多个源站服务器,而各个源站服务器的处理能力存在一定的差异性(如,各个源站服务器本身的处理能力,或当前时刻的处理能力),从而导致各个源站服务器的并发能力不同;当出口服务器连接源站服务器时,根据第一指示信息生成基于权重的映射关系,针对多个源站服务器进行合理的连接分配,从而提升多个源站服务器的负载均衡能力。其中,默认设置处理能力较强的源站服务器的权重值较大,反之则较小。例如,在数据访问请求消息中,第一指示信息的具体插入格式为:
|源站服务器1,权重3||源站服务器2,权重2||源站服务器3,权重1|
出口服务器对接收到的数据访问请求消息进行解析,得到第一指示信息,以及用户IP地址,根据用户IP地址(即源地址),利用基于权重的哈希(WSH:Weight Source Hash)算法确定目标源站服务器,WSH算法为针对多源站服务器权重信息的负载均衡算法。
出口服务器基于应用标识建立多个哈希表,即根据属于同一应用标识的多个源站服务器标识及权重信息,利用填充算法完成每个哈希表的填充,哈 希表用于表示用户IP地址与源站服务器标识的对应关系。其中,对应关系中的用户IP地址以索引键的形式存在,即根据用户IP地址,利用映射关系计算得到的哈希值即为索引键。
具体地,根据解析得到的用户IP地址计算得到第二索引键,利用所建立的哈希表查找与第二索引键相同的第一索引键,再根据第一索引键确定与第二索引键相匹配的源站服务器标识,将所确定的源站服务器作为目标源站服务器,并将数据访问请求消息发送给该目标源站服务器。
在源站服务器数量,以及源站服务器权重信息相对不变,且源站服务器未做会话(session)同步的情况下,同一用户继续访问在前应用标识时,数据访问请求消息所要访问的源站服务器是相同的,与权重信息无关,保证了同一用户的数据访问请求消息能够访问到同一个源站服务器,保持会话状态。只有当与用户所要访问应用相匹配的源站服务器发生改变时,才需要基于所匹配的源站服务器的权重信息实现数据访问的负载均衡。
出口服务器利用WSH算法确定目标源站服务器,即针对与某一应用标识相匹配的所有源站服务器,利用填充算法生成哈希表,对数据访问请求消息中的用户IP地址进行索引键计算,生成第二索引键;根据第二索引键,利用生成的哈希表确定目标源站服务器;将数据访问请求消息发送给所确定的目标源站服务器。
优选的,根据所述第一指示信息生成基于权重的映射关系的步骤包括:
获取所述多个源站服务器的资源使用率信息,所述资源使用率信息用于表示每一个源站服务器的资源使用率;
根据所述第一指示信息以及所述资源使用率信息生成基于权重的映射关系。
对于出口服务器利用填充算法进行哈希表填充的具体方法为,基于某一应用的多个源站服务器的权重信息,每次选择CPU使用率相对较少的源站服务器进行哈希表的填充,例如,各源站服务器的权重比为:
源站服务器1∶源站服务器2∶源站服务器3=3∶2∶1
若源站服务器1的CPU使用率与源站服务器1权重的比值大于源站服务器2的CPU使用率与源站服务器2权重的比值,则选择源站服务器2进行哈希表的填充。
本发明实施例一中的GRP协议支持多源站服务器,能够有效地解决仅支持单一源站服务器,连接失败即访问失败的问题,当出口服务器根据计算得到的目标源站服务器连接失败后,仍然能够选择其他备用源站服务器进行连接,提高数据访问的成功率。
图4示出了本发明实施例一中中央服务器侧访问分发方法的原理示意图,如图4所示,包括:
步骤401、获取多个应用中的每一个应用对应的第一指示信息,所述第一指示信息用于表示所述应用所对应的多个源站服务器中每一个源站服务器对数据访问请求消息的处理能力;
步骤402、将第一指示信息发送给多个入口服务器。
中央服务器通过获取与每一个应用相匹配的第一指示信息,生成包含应用标识与第一指示信息对应关系的消息并发送给入口服务器,以便入口服务器在接收到数据访问请求消息后,能够通过解析数据访问请求消息获取发起数据访问请求消息的应用标识,并根据解析得到的应用标识获取与应用相匹配的第一指示信息。
实施例二:
基于同一发明构思,本发明实施例中还提供了一种访问分发装置,由于这些设备解决问题的原理与一种访问分发方法相似,因此这些设备的实施可以参见方法的实施,重复之处不再赘述。
图5示出了本发明实施例二中入口服务器侧访问分发装置的结构示意图,如图5所示,入口服务器侧访问分发装置500可以包括:
第一接收模块501,用于接收数据访问请求消息。
确定模块502,用于确定发起所述数据访问请求消息的应用。
获取模块503,用于从中央服务器中获取所述应用所对应的第一指示信息;所述第一指示信息用于表示所述应用所对应的多个源站服务器中每一个源站服务器对数据访问请求消息的处理能力。
第一发送模块504,用于将所述第一指示信息携带在所述数据访问请求消息中发送给出口服务器。
优选的,确定模块502包括:
根据所述数据访问请求消息的传输层协议信息中的应用标识确定发起所述数据访问请求消息的应用。
优选的,第一发送模块503还包括:
对数据访问请求消息中的IP地址和/或端口号进行二进制整型。
图6示出了本发明实施例二中出口服务器侧访问分发装置的结构示意图,如图6所示,出口服务器侧访问分发装置600可以包括:
第二接收模块601,用于接收携带有第一指示信息的数据访问请求消息,所述第一指示信息用于表示第一应用所对应的多个源站服务器中每一个源站服务器对数据访问请求消息的处理能力,所述第一应用为发起所述数据访问请求消息的应用。
第二发送模块602,用于根据所述第一指示信息将由所述第一应用发起的各个数据访问请求消息发送到对应的源站服务器上。
优选的,第二发送模块602包括:
根据所述第一指示信息生成基于权重的映射关系;
利用生成的映射关系对待发送的数据访问请求消息的源地址进行运算;
将待发送的数据访问请求消息发送到所得到的运算结果所对应的源站服务器。
优选的,根据所述第一指示信息生成基于权重的映射关系的步骤包括:
获取所述多个源站服务器的资源使用率信息,所述资源使用率信息用于 表示每一个源站服务器的资源使用率;
根据所述第一指示信息以及所述资源使用率信息生成基于权重的映射关系。
本发明实施例二中,图7为本发明实施例二中中央服务器侧访问分发装置的结构示意图,中央服务器侧700的访问分发装置包括:
第二获取模块701,用于获取多个应用中的每一个应用对应的第一指示信息,所述第一指示信息用于表示所述应用所对应的多个源站服务器中每一个源站服务器对数据访问请求消息的处理能力。
第三发送模块702,用于将第一指示信息发送给多个入口服务器。
为了描述的方便,以上所述装置的各部分以功能分为各种模块分别描述。当然,在实施本发明时可以把各模块或单元的功能在同一个或多个软件或硬件中实现。
实施例三:
基于同一发明构思,本发明实施例中还提供了一种访问分发系统,由于这些设备解决问题的原理与一种访问分发方法、一种访问分发装置相似,因此这些设备的实施可以参见方法的实施,重复之处不再赘述。
图8示出了本发明实施例三中访问分发系统架构示意图,如图8所示,所述访问分发系统800可以包括:
入口服务器801,用于接收数据访问请求消息,确定发起所述数据访问请求消息的应用,并从中央控制器802中获取所述应用所对应的第一指示信息,所述第一指示信息用于表示所述应用所对应的多个源站服务器中每一个源站服务器对数据访问请求消息的处理能力,以及将所述第一指示信息携带在所述数据访问请求消息中发送给出口服务器803。
中央控制器802,用于获取所述应用对应的第一指示信息,并将第一指示信息发送给所述入口服务器801。
出口服务器803,接收携带有第一指示信息的数据访问请求消息,根据所 述第一指示信息将由所述应用发起的数据访问请求消息发送到对应的源站服务器上。
获取第一指示信息的方式包括静态获取和动态获取。基于静态方式获取第一指示信息的方法具体为,中央服务器802通过WEB接口录入应用对应的第一指示信息,并生成应用标识与第一指示信息的对应关系表,并将生成的对应关系表实时地下发给入口服务器801,以便入口服务器根据数据访问请求消息中的应用标识获取到与应用标识相对应的第一指示信息。
基于动态方式获取第一指示信息的方法具体为,中央服务器预先通过SNMP协议获取源站服务器的CPU使用率,并根据源站服务器的CPU使用率,计算新的源站服务器权重,并生成新的应用标识与第一指示信息的对应关系表,并将生成的新的对应关系表实时地下发给入口服务器,以便入口服务器实现第一指示信息的实时更新。
例如,中央服务器接收某一应用的两个源站服务器的CPU使用率信息,源站服务器1和源站服务器2的CPU使用率分别为40%和80%,则源站服务器1与源站服务器2的权重比为2∶1,即源站服务器权重之比为源站服务器CPU使用率的反比。若某一源站服务器的CPU使用率大于等于95%,则该源站服务器权重可以设置为最低,或直接不再分配新的数据访问请求消息给该源站服务器。
本领域内的技术人员应明白,本发明的实施例可提供为方法、系统、或计算机程序产品。因此,本发明可采用完全硬件实施例、完全软件实施例、或结合软件和硬件方面的实施例的形式。而且,本发明可采用在一个或多个其中包含有计算机可用程序代码的计算机可用存储介质(包括但不限于磁盘存储器、CD-ROM、光学存储器等)上实施的计算机程序产品的形式。
本发明是参照根据本发明实施例的方法、设备(系统)、和计算机程序产品的流程图和/或方框图来描述的。应理解可由计算机程序指令实现流程图和/或方框图中的每一流程和/或方框、以及流程图和/或方框图中的流程 和/或方框的结合。可提供这些计算机程序指令到通用计算机、专用计算机、嵌入式处理机或其他可编程数据处理设备的处理器以产生一个机器,使得通过计算机或其他可编程数据处理设备的处理器执行的指令产生用于实现在流程图一个流程或多个流程和/或方框图一个方框或多个方框中指定的功能的装置。
这些计算机程序指令也可存储在能引导计算机或其他可编程数据处理设备以特定方式工作的计算机可读存储器中,使得存储在该计算机可读存储器中的指令产生包括指令装置的制造品,该指令装置实现在流程图一个流程或多个流程和/或方框图一个方框或多个方框中指定的功能。
这些计算机程序指令也可装载到计算机或其他可编程数据处理设备上,使得在计算机或其他可编程设备上执行一系列操作步骤以产生计算机实现的处理,从而在计算机或其他可编程设备上执行的指令提供用于实现在流程图一个流程或多个流程和/或方框图一个方框或多个方框中指定的功能的步骤。
尽管已描述了本发明的优选实施例,但本领域内的技术人员一旦得知了基本创造性概念,则可对这些实施例作出另外的变更和修改。所以,所附权利要求意欲解释为包括优选实施例以及落入本发明范围的所有变更和修改。

Claims (15)

  1. 一种访问分发方法,其特征在于,包括:
    接收数据访问请求消息;
    确定发起所述数据访问请求消息的应用;
    从中央服务器中获取所述应用所对应的第一指示信息,所述第一指示信息用于表示所述应用所对应的多个源站服务器中每一个源站服务器对数据访问请求消息的处理能力;
    将所述第一指示信息携带在所述数据访问请求消息中发送给出口服务器。
  2. 如权利要求1所述的方法,其特征在于,所述确定发起所述数据访问请求消息的应用,包括:
    根据所述数据访问请求消息的传输层协议信息中的应用标识确定发起所述数据访问请求消息的应用。
  3. 如权利要求1所述的方法,其特征在于,在将数据访问请求消息发送给出口服务器之前,还包括:
    对数据访问请求消息中的IP地址和/或端口号进行二进制整型。
  4. 一种访问分发方法,其特征在于,包括:
    接收携带有第一指示信息的数据访问请求消息,所述第一指示信息用于表示第一应用所对应的多个源站服务器中每一个源站服务器对数据访问请求消息的处理能力,所述第一应用为发起所述数据访问请求消息的应用;
    根据所述第一指示信息将由所述第一应用发起的各个数据访问请求消息发送到对应的源站服务器上。
  5. 如权利要求4所述的方法,其特征在于,所述根据所述第一指示信息将由所述第一应用发起的各个数据访问请求消息发送到对应的源站服务器上,包括:
    根据所述第一指示信息生成基于权重的映射关系;
    利用生成的映射关系对待发送的数据访问请求消息的源地址进行运算;
    将待发送的数据访问请求消息发送到所得到的运算结果所对应的源站服务器。
  6. 如权利要求5所述的方法,其特征在于,所述根据所述第一指示信息生成基于权重的映射关系,包括:
    获取所述多个源站服务器的资源使用率信息,所述资源使用率信息用于表示每一个源站服务器的资源使用率;
    根据所述第一指示信息以及所述资源使用率信息生成基于权重的映射关系。
  7. 一种访问分发方法,其特征在于,包括:
    获取多个应用中的每一个应用对应的第一指示信息,所述第一指示信息用于表示所述应用所对应的多个源站服务器中每一个源站服务器对数据访问请求消息的处理能力;
    将第一指示信息发送给多个入口服务器。
  8. 一种访问分发装置,其特征在于,包括:
    第一接收模块,用于接收数据访问请求消息;
    确定模块,用于确定发起所述数据访问请求消息的应用;
    第一获取模块,用于从中央服务器中获取所述应用所对应的第一指示信息,所述第一指示信息用于表示所述应用所对应的多个源站服务器中每一个源站服务器对数据访问请求消息的处理能力;
    第一发送模块,用于将所述第一指示信息携带在所述数据访问请求消息中发送给出口服务器。
  9. 如权利要求8所述的装置,其特征在于,所述确定模块包括:
    根据所述数据访问请求消息的传输层协议信息中的应用标识确定发起所 述数据访问请求消息的应用。
  10. 如权利要求8所述的装置,其特征在于,所述第一发送模块还包括:
    对数据访问请求消息中的IP地址和/或端口号进行二进制整型。
  11. 一种访问分发装置,其特征在于,出口服务器包括:
    第二接收模块,用于接收携带有第一指示信息的数据访问请求消息,所述第一指示信息用于表示第一应用所对应的多个源站服务器中每一个源站服务器对数据访问请求消息的处理能力,所述第一应用为发起所述数据访问请求消息的应用;
    第二发送模块,用于根据所述第一指示信息将由所述第一应用发起的各个数据访问请求消息发送到对应的源站服务器上。
  12. 如权利要求10所述的装置,其特征在于,所述第二发送模块包括:
    根据所述第一指示信息生成基于权重的映射关系;
    利用生成的映射关系对待发送的数据访问请求消息的源地址进行运算;
    将待发送的数据访问请求消息发送到所得到的运算结果所对应的源站服务器。
  13. 如权利要求12所述的装置,其特征在于,所述根据所述第一指示信息生成基于权重的映射关系,包括:
    获取所述多个源站服务器的资源使用率信息,所述资源使用率信息用于表示每一个源站服务器的资源使用率;根据所述第一指示信息以及所述资源使用率信息生成基于权重的映射关系。
  14. 一种访问分发装置,其特征在于,包括:
    第二获取模块,用于获取多个应用中的每一个应用对应的第一指示信息,所述第一指示信息用于表示所述应用所对应的多个源站服务器中每一个源站服务器对数据访问请求消息的处理能力;
    第三发送模块,用于将第一指示信息发送给多个入口服务器。
  15. 一种访问分发系统,其特征在于,包括:
    入口服务器装置,用于接收数据访问请求消息,确定发起所述数据访问请求消息的应用,并从中央控制器中获取所述应用所对应的第一指示信息,所述第一指示信息用于表示所述应用所对应的多个源站服务器中每一个源站服务器对数据访问请求消息的处理能力,以及将所述第一指示信息携带在所述数据访问请求消息中发送给出口服务器;
    中央控制器,用于获取所述应用对应的第一指示信息,并将第一指示信息发送给多个所述入口服务器;
    出口服务器,用于接收携带有第一指示信息的数据访问请求消息,根据所述第一指示信息将由所述应用发起的数据访问请求消息发送到对应的源站服务器上。
PCT/CN2016/092026 2016-07-28 2016-07-28 访问分发方法、装置及系统 Ceased WO2018018490A1 (zh)

Priority Applications (2)

Application Number Priority Date Filing Date Title
PCT/CN2016/092026 WO2018018490A1 (zh) 2016-07-28 2016-07-28 访问分发方法、装置及系统
CN201680002768.XA CN106716974B (zh) 2016-07-28 2016-07-28 访问分发方法、装置及系统

Applications Claiming Priority (1)

Application Number Priority Date Filing Date Title
PCT/CN2016/092026 WO2018018490A1 (zh) 2016-07-28 2016-07-28 访问分发方法、装置及系统

Publications (1)

Publication Number Publication Date
WO2018018490A1 true WO2018018490A1 (zh) 2018-02-01

Family

ID=58906794

Family Applications (1)

Application Number Title Priority Date Filing Date
PCT/CN2016/092026 Ceased WO2018018490A1 (zh) 2016-07-28 2016-07-28 访问分发方法、装置及系统

Country Status (2)

Country Link
CN (1) CN106716974B (zh)
WO (1) WO2018018490A1 (zh)

Cited By (3)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
US11511053B2 (en) 2017-02-16 2022-11-29 Embecta Corp. Pen needle hub injection depth optimization
CN115801791A (zh) * 2022-10-27 2023-03-14 北京天空卫士网络安全技术有限公司 一种实现负载均衡的方法和负载均衡器
CN120804105A (zh) * 2025-09-15 2025-10-17 厦门渊亭信息科技有限公司 基于lvc仿真的虚拟训练数据管理系统及方法

Families Citing this family (4)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
CN108064443B (zh) * 2017-09-30 2021-08-06 达闼机器人有限公司 一种代理转发方法和装置、代理服务器和多级代理网络
CN108881057B (zh) * 2018-04-20 2022-08-02 网宿科技股份有限公司 回源线路的选择方法和流量分发器
CN110858844A (zh) * 2018-08-22 2020-03-03 阿里巴巴集团控股有限公司 服务请求处理方法、控制方法、装置、系统及电子设备
CN115017259A (zh) * 2022-05-09 2022-09-06 隆昌市高新技术创业服务中心 确定检索服务器的方法及装置

Citations (4)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
US20100228819A1 (en) * 2009-03-05 2010-09-09 Yottaa Inc System and method for performance acceleration, data protection, disaster recovery and on-demand scaling of computer applications
CN102439913A (zh) * 2009-02-27 2012-05-02 雅塔公司 网络业务管理和负载平衡的系统和方法
CN102792292A (zh) * 2009-12-07 2012-11-21 考持·维 站点性能优化和因特网业务处理的系统和方法
US20140280959A1 (en) * 2013-03-15 2014-09-18 Eric J. Bauer Application server instance selection based on protocol latency information

Family Cites Families (3)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
CN101626387B (zh) * 2009-08-12 2013-08-07 中兴通讯股份有限公司 媒体业务处理系统、方法及终端
US8996614B2 (en) * 2011-02-09 2015-03-31 Citrix Systems, Inc. Systems and methods for nTier cache redirection
CN104283948B (zh) * 2014-09-26 2018-12-07 东软集团股份有限公司 服务器集群系统及其负载均衡实现方法

Patent Citations (4)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
CN102439913A (zh) * 2009-02-27 2012-05-02 雅塔公司 网络业务管理和负载平衡的系统和方法
US20100228819A1 (en) * 2009-03-05 2010-09-09 Yottaa Inc System and method for performance acceleration, data protection, disaster recovery and on-demand scaling of computer applications
CN102792292A (zh) * 2009-12-07 2012-11-21 考持·维 站点性能优化和因特网业务处理的系统和方法
US20140280959A1 (en) * 2013-03-15 2014-09-18 Eric J. Bauer Application server instance selection based on protocol latency information

Cited By (4)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
US11511053B2 (en) 2017-02-16 2022-11-29 Embecta Corp. Pen needle hub injection depth optimization
US12023474B2 (en) 2017-02-16 2024-07-02 Embecta Corp. Pen needle hub injection depth optimization
CN115801791A (zh) * 2022-10-27 2023-03-14 北京天空卫士网络安全技术有限公司 一种实现负载均衡的方法和负载均衡器
CN120804105A (zh) * 2025-09-15 2025-10-17 厦门渊亭信息科技有限公司 基于lvc仿真的虚拟训练数据管理系统及方法

Also Published As

Publication number Publication date
CN106716974A (zh) 2017-05-24
CN106716974B (zh) 2020-08-04

Similar Documents

Publication Publication Date Title
CN106716974B (zh) 访问分发方法、装置及系统
US20190124018A1 (en) Cloud resource processing method and physical node
US10069946B2 (en) Hardware-based packet editor
CN103597471B (zh) 用于对计算机网络上的数据通信进行缓存的方法和系统
US20110040892A1 (en) Load balancing apparatus and load balancing method
CN103647853A (zh) 一种在VxLAN中发送ARP报文发送方法、VTEP及VxLAN控制器
CN103428261A (zh) 通过硬件辅助处理http报头的方法
EP3735768B1 (en) Improving qoe for video and web services using cross-layer information
CN114338574B (zh) 一种即时通讯方法、管理节点及系统
CN112087382A (zh) 一种服务路由方法及装置
US10952264B2 (en) Persistent distribution of connectionless traffic supporting virtual connection migration
CN106453460A (zh) 一种文件分发方法、装置和系统
CN103685496A (zh) 一种文件数据的传输方法和设备
JP2017500679A (ja) メディアリソースフィードバック方法、装置、プログラム及び記録媒体
CN111837358A (zh) 分布式网络时间协议
CN116633934A (zh) 负载均衡方法、装置、节点及存储介质
WO2022166593A1 (zh) 点对点网络调度方法和系统
CN106888261A (zh) 一种分布式文件传输方法、服务器及私有云设备
WO2014142258A1 (ja) 通信システム、制御装置、アドレス割当方法及びプログラム
WO2016180188A1 (zh) 分发建链方法、装置和系统
CN106302661B (zh) P2p数据加速方法、装置和系统
CN103746768B (zh) 一种数据包的识别方法及设备
CN107070719B (zh) 一种设备管理方法和装置
CN112968965A (zh) Nfv网络节点的元数据服务方法、服务器及存储介质
CN107612831B (zh) 一种访问源站的数据报文的传输方法及装置

Legal Events

Date Code Title Description
121 Ep: the epo has been informed by wipo that ep was designated in this application

Ref document number: 16910071

Country of ref document: EP

Kind code of ref document: A1

NENP Non-entry into the national phase

Ref country code: DE

32PN Ep: public notification in the ep bulletin as address of the adressee cannot be established

Free format text: NOTING OF LOSS OF RIGHTS PURSUANT TO RULE 112(1) EPC (EPO FORM 1205A DATED 16/05/2019)

122 Ep: pct application non-entry in european phase

Ref document number: 16910071

Country of ref document: EP

Kind code of ref document: A1