JP2011118593A - Data transfer server, data transfer system, data transfer method, and program - Google Patents

Abstract

<P>PROBLEM TO BE SOLVED: To provide a configuration and a method for preventing resources of each data transfer server in a data transfer system for relaying a data request and data from being wasted, according to a predetermined network configuration. <P>SOLUTION: Each data transfer server arranged on a network for relaying a data request and data, when receiving a data request from another data transfer server, determines whether to cache the requested data, and when determining that the data is to be cached, causes the data transfer server that caches the data to participate in a logical network, and transfers and caches the data based on the request between the data transfer servers participating in the logical network. <P>COPYRIGHT: (C)2011,JPO&INPIT

Description

  The present invention relates to a data transfer server, a data transfer system, a data transfer method, and a program, and more particularly, to a data transfer server, a data transfer system, a data transfer method, and a program having a cache function.

  Patent Document 1 discloses a server having a cache function and having a configuration for notifying each other of load information, cache information, and the like within a group. According to the document, when these servers receive a first request by URL designation using a common address from a PC, they accept processing according to the cache state and load state in the group. If there is no cache of the latest content data, the corresponding server obtains it from the Web server by the second request and the second response, caches it, and transmits it to the PC as the first response. According to the document, with the above configuration, it is possible to achieve load distribution and balancing among a plurality of servers without requiring a device for managing the whole.

  Patent Document 2 discloses a communication relay device provided between a client terminal and a server device, and based on a request from the client terminal, a method check and a URI check are performed to determine whether the cache is cacheable or not cacheable. A control method is disclosed.

  In addition, as a background art of the present invention, there is a CDN (Contents Delivery Network). The CDN adopts a hierarchical logical tree structure according to the number of users to be covered or a geographical range, etc., and performs appropriate content deployment according to a viewing request from the user, thereby enabling high-speed response and efficient network. Use has been realized.

  FIG. 16 is a diagram showing a schematic configuration of a CDN having a hierarchical configuration. In the example of FIG. 16, a three-tier tier configuration is adopted from the DC (Data Center) in the metropolitan area to the DC in the regional mid-sized city in the scale of the number of users to be covered. is not.

  Each user using the CDN as shown in FIG. 16 accesses the nearest DC and requests content. If there is no desired content in the nearest DC, the content is acquired from another DC. At this time, a logical network is constructed by each DC, and the content is cached in the DC on the way of transfer, so that the content is copied to a plurality of DCs. As a result, a large number of accesses from geographically dispersed users can be processed by a plurality of dispersed DCs, and load distribution and user content acquisition response can be improved.

  There are several methods for acquiring content from other DCs and caching them in the DCs on the transfer path. The simplest method is to cache all DCs on the way. Furthermore, as a method that enables more efficient use of resources, a complicated cache method that considers the cache usage rate according to the content popularity and the load on the server is also conceivable. When performing such a complicated cache method, it is assumed that a DC to be dynamically cached is determined at the time of content transfer. For example, when determining whether or not to cache according to the popularity of content during content transfer, the DC to be cached differs for each content.

  As a simple method of dynamically caching with a DC on an intermediate path for each content transfer, there is a method of determining whether to transfer the content to be transferred through all the DCs on the intermediate path and caching with each DC.

JP 2006-301769 A JP 2009-9592 A

  However, each of the above-described methods has a problem in that resources such as the CPU, memory, and network bandwidth of the server in the DC are wasted. The reason is that the data is routed through all the DCs on the way and the cache is judged after the server in the DC receives the data. For example, even when the above-described method for dynamically determining whether or not to cache on the midway route DC is used for each content transfer, when a certain DC determines that the content is not cached, the content continues as it is. Accordingly, resources such as the CPU, memory, and network bandwidth of the server in the DC required for the “determination” are wasted. In particular, when the content size such as video is large, the influence is large. Note that the above problem is not limited to DC, and applies to data transfer servers in general having a cache function for caching transfer data.

  Patent Document 1 is an example in which caching is performed in consideration of the server load described above. In addition, Patent Document 2 discloses a configuration that includes a layer 7 load distribution device as an example and distributes requests from client terminals to a plurality of Web communication relay devices. They are installed in parallel for load distribution and do not provide the data for each user assigned to them. In addition, since requests from client terminals are distributed by the layer 7 load balancer, an unspecified Web communication relay device does not cache the data.

  The present invention has been made in view of the above circumstances, and the object of the present invention is to request data and data according to a network configuration in which each data transfer server is determined in advance as in the above-described example of DC. It is an object of the present invention to provide a configuration and a method capable of avoiding wasteful use of resources of individual data transfer servers in a data transfer system that relays data.

  According to the first aspect of the present invention, whether or not the requested data is cached when a data request is received from another data transfer server, which is arranged in the network that relays the data request and the data. If it is determined that the data is to be cached, a process of participating in the logical network of the data transfer server that caches the data is executed, and the data transfer server participating in the logical network is based on the request. A data transfer server for transferring and caching data is provided.

  According to a second aspect of the present invention, a data transfer system including a plurality of the above-described data transfer servers is provided.

  According to the third aspect of the present invention, when a data transfer server arranged in a network that relays a data request and data receives a data request from another data transfer server, the requested data is Determining whether or not to cache, and determining that the data is to be cached, executing a process of participating in the logical network of the data transfer server that caches the data; and a data transfer server participating in the logical network A data transfer method is provided that includes transferring data based on the request and caching. Note that. The method is tied to a specific machine, a data transfer server located in a network that relays data requests and data.

  According to a fourth aspect of the present invention, when a data request is received from another data transfer server, the program is executed by a data transfer server arranged in a network that relays the data request and data. Determining whether to cache the requested data, and, if it is determined to cache the data, a process of executing a process of participating in a logical network of a data transfer server that caches the data; and the logical network When the data based on the request is received from the data transfer server that participated in the data transfer, the data transfer includes the data cache and the process of transferring the data to another data transfer server participating in the logical network. A program to be executed by a server is provided. This program can be recorded on a computer-readable storage medium. That is, the present invention can be embodied as a computer program product.

  According to the present invention, useless use of resources of a data transfer server that does not cache data corresponding to a request can be avoided. The reason for this is to determine in advance whether or not to cache when making a request, and to construct a logical network dedicated to data transfer, and to transfer data without using a non-cached server by using that logical network. It is.

It is a figure for demonstrating the outline | summary of this invention. It is a block diagram which shows the structure of the data transfer server which concerns on the 1st Embodiment of this invention. It is a figure which shows the structure of the request which the data transfer server which concerns on the 1st Embodiment of this invention transmits. It is a figure which shows the structure of the data which the data transfer server which concerns on the 1st Embodiment of this invention transmits. It is a figure which shows the structure of the data transfer table which the data transfer server which concerns on the 1st Embodiment of this invention hold | maintains. It is a flowchart which shows the operation | movement at the time of the data transfer server which concerns on the 1st Embodiment of this invention transmits a request. It is a flowchart which shows operation | movement when the data transfer server which concerns on the 1st Embodiment of this invention receives a request. It is a flowchart which shows the operation | movement at the time of the data transfer server which concerns on the 1st Embodiment of this invention transfers data. It is a block diagram which shows the structure of the data transfer server which concerns on the 2nd Embodiment of this invention. It is a figure which shows the structure of the request which the data transfer server which concerns on the 2nd Embodiment of this invention transmits. It is a figure which shows the structure of the data which the data transfer server which concerns on the 2nd Embodiment of this invention transmits. It is a flowchart which shows operation | movement when the data transfer server which concerns on the 1st Embodiment of this invention receives a request. It is a flowchart which shows the operation | movement at the time of the data transfer server which concerns on the 1st Embodiment of this invention transfers data. It is a figure which shows the network structure for demonstrating the specific operation | movement of this invention. It is a figure for demonstrating the specific operation | movement of the network structure of FIG. It is a structural example of a content delivery network (CDN) having a hierarchical structure.

  First, the outline of the present invention will be described with reference to FIG. A solid line between the data transfer server and the router in the upper part of FIG. 1 indicates a physical network connection, and a broken line indicates a logical network that relays data requests and data. The lower part of FIG. 1 shows an example of an operation sequence when a data request is sent from the data transfer server 101a to the data transfer server 101d.

  For example, consider a case where the data transfer server 101a receives a request for data held by the data transfer server 101d from a client (not shown). The data transfer servers 101b and 101c that have received the request from the data transfer server 101a determine whether to cache the data requested in the request when relaying the request to the next node. If it is determined that the data is to be cached, a process of participating in the logical network of the data transfer server that caches the data (for example, adding itself to the server list / address list constituting the logical network) is executed. The example of FIG. 1 is an example in which the data transfer server 101b determines to cache and the data transfer server 101c determines not to cache. Thereafter, the data based on the request is transferred from the data transfer server 101d to the data transfer server 101a via the data transfer server 101b participating in the logical network. Thereby, at the time of data transfer, it is possible to prevent the data transfer server 101c that does not cache from consuming resources.

[First Embodiment]
Next, a first embodiment of the present invention will be described in detail with reference to the drawings. FIG. 2 is a block diagram showing the configuration of the data transfer server according to the first embodiment of the present invention.

  Referring to FIG. 2, a data transfer server (hereinafter, abbreviated as “server” unless otherwise specified) 101 according to the first embodiment of the present invention is connected to another server via a transmission line 100, and data Holding, sending and receiving and transferring data and requests. FIG. 2 shows one server, but there is no limit on the number of servers. Further, the connection configuration between servers may be as shown in FIG. 14, or may have a hierarchical configuration like the CDN shown in FIG.

  2 includes a transmission / reception unit 10, a request processing unit 11, a data transfer unit 12, a logical network information management unit 13, a cache determination unit 14, a data holding unit 15, and a data transfer table 16. Including. Each of these operates as follows.

  The transmission / reception unit 10 performs transmission / reception of requests and data for requesting transfer of requested data.

  FIG. 3 is a diagram illustrating a configuration of a request transmitted by the transmission / reception unit 10 of each server 101. In the example of FIG. 3, the identifier of the server that holds the data, the identifier of the data that requests transfer, and the physical address of the server that is the transfer destination of the data are included. The server identifier is uniquely identified on the logical network.

  FIG. 4 is a diagram illustrating a configuration of data transmitted by the transmission / reception unit 10 of each server 101. In the example of FIG. 4, the data identifier is associated with the data body. For example, the server that has received the request can search for data having a corresponding identifier from the data holding unit 15.

  The request processing unit 11 analyzes the received request and creates a request to be transmitted. For example, the request processing unit 11 of the server that has received the request as shown in FIG. 3 refers to the data identifier in the request to determine whether or not it holds data. The cache determination unit 14 is requested to determine whether or not to perform the request, and the logical network information stored in the logical network information management unit 13 is referred to transfer the request to the next destination server. When the request is transferred, if the determination result that the cache is performed is received from the cache determination unit 14, the request processing unit 11 sets the value of the “transfer destination server physical address” field of the request to its own physical address. The operation to rewrite is performed.

  The data transfer unit 12 transmits and transfers the requested data to the requesting server. Specifically, the data transfer unit 12 refers to the data transfer table 16 and transfers the data to a server determined to cache the data.

  When the data transfer table 16 receives a request and determines to cache the data, the data transfer table 16 holds an identifier of the data and a physical address of the next server to be transferred until the data is transferred.

  FIG. 5 is a diagram showing the configuration of the data transfer table 16. For example, when data is received from a certain server, the data transfer unit 12 transmits the data from the data transfer table 16 to the physical address of the server corresponding to the received data identifier.

  The logical network information management unit 13 manages logical network and overlay network information between the servers 101.

  The cache determination unit 14 determines whether to cache the requested data. Various methods can be used for determining the cache. For example, a method of determining by probability, determining according to a load, and determining a use fee when a server is rented from a cloud service or the like can be considered. In addition, there is no particular limitation such as determining whether or not it should be cached using metadata included in the data.

  The data holding unit 15 holds data having the configuration illustrated in FIG. As the data holding unit 15, block access, file access storage, or the like can be used, and there is no particular limitation.

  Next, the overall operation of this embodiment will be described in detail. Hereinafter, “when request is transmitted” when the server is a request transmission source, “when request is received” when the server receives a request, and “when data is transferred” when the server transfers data, FIG. 6, FIG. This will be described with reference to FIG.

[When sending a request]
FIG. 6 is a flowchart showing an operation when the server 101 does not hold data requested by a client and requests data from another server (hereinafter referred to as “other server”). .

  First, the request processing unit 11 stores the identifier of the requested data and the identifier of the server holding the data in the request (step S100). It is assumed that the identifier of the data and the identifier of the server holding the data are obtained from any means (such as a directory service).

  Next, the request processing unit 11 stores its own physical address in the request transfer destination server physical address field (see FIG. 3) (step S101). The physical address is an address that physically represents the position of the server on the network. For example, an IP address is a good example. When it is necessary to notify the port number, the port number is also stored in the request.

  The request processing unit 11 refers to the logical network information management unit 13 to instruct the transmission / reception unit 10 to transmit the request in order to transfer the request to a server on an intermediate path on the logical network (step S102). .

[When receiving a request]
FIG. 7 is a flowchart showing the operation when the server 101 receives a request. First, when a request is received through the transmission / reception unit 10 (step S200), the request processing unit 11 of the server 101 refers to the data holding unit 15 through the data transfer unit 12 and has data identifiers that match the requested data. Is determined (step S201).

  When data having a data identifier that matches the requested data is held (Yes in step S201), the request processing unit 11 of the server 101 reads out the corresponding data through the data transfer unit 12 (step S206). The data is transmitted to the physical address stored in the transfer destination server physical address field (see FIG. 3) of the request (step S207). The server that has received the data executes data cache processing and transfer processing according to the flow shown in FIG.

  On the other hand, when data having a data identifier that matches the requested data is not held (No in step S201), the request processing unit 11 of the server 101 caches the request target data in the cache determination unit 14. Whether or not is determined (step S202).

  When the cache determination unit 14 determines that the request target data is to be cached (Yes in step S202), the request processing unit 11 of the server 101 stores the data identifier stored in the request in the data transfer table 16 and the transfer destination. A set of server physical addresses is stored (step S203). Next, the request processing unit 11 of the server 101 rewrites the value of the transfer destination server physical address field in the request with its own physical address (step S204), and then refers to the logical network information management unit 13 to transmit / receive the unit 10 Is instructed to transmit the rewritten request (step S205).

  On the other hand, when the cache determination unit 14 determines not to cache the request target data (No in step S202), the request processing unit 11 of the server 101 transmits the request as it is to the next server on the logical network (step S205). ). Omitting the processing in steps S203 and S204 in the server that has been determined not to cache means removing itself from the logical network path during data transfer.

[When transferring data]
FIG. 8 is a flowchart showing the operation when the server 101 receives data. First, when data is received through the transmission / reception unit 10 (step S300), the data transfer unit 12 of the server 101 stores (caches) the data in the data holding unit 15. (Step S301).

  Next, the data transfer unit 12 of the server 101 refers to the data transfer table 16 and transfers the data to the server of the transfer destination server physical address corresponding to the data identifier of the data (step S302). Next, the data transfer unit 12 of the server 101 deletes the row corresponding to the data identifier of the data transfer table 16 (step S303).

  Thereafter, the processes of steps S301 to S302 are repeated on the server that has received the data transfer, that is, the server that has determined that the cache determination unit 14 performs caching.

  As described above, in this embodiment, it is possible to avoid wasteful use of resources in a server that does not cache. The reason is that the address of the server to be cached is notified at the time of request, a logical network path dedicated to data transfer is established between the servers that are determined to be cached, and the data is routed through a server that does not cache by using the logical network path. It is because it transfers without.

  Further, as a secondary effect of the present embodiment, there is also an effect that data can be acquired from other servers at high speed by not caching in the server on the way route. The reason is that the time required for communication is reduced because the data is transferred without going through a non-cached server.

[Second Embodiment]
Next, a second embodiment of the present invention will be described in detail with reference to the drawings. In the first embodiment described above, each server holds the data transfer table and stores the physical address of the server determined to be cached, thereby constructing the logical network path at the time of data transfer. The form is such that logical network path information at the time of data transfer is stored in the request and data, respectively.

  FIG. 9 is a block diagram showing the configuration of the data transfer server according to the second embodiment of the present invention. As shown in FIG. 9, the difference in configuration between the data transfer server 201 of the present embodiment and the data transfer server 101 of the first embodiment of the present invention is a configuration without the data transfer table 16. Is a point.

  FIG. 10 is a diagram illustrating a configuration of a request transmitted by the transmission / reception unit 10 of each server 201. In the example of FIG. 10, in addition to the identifier of the server that holds the data and the identifier of the data that requests transfer, the physical address of the server that is the transfer destination of the data can be sequentially added. The physical address of the sequentially added server is used as a logical network path during data transfer.

  FIG. 11 is a diagram illustrating a configuration of data transmitted by the transmission / reception unit 10 of each server 201. In the example of FIG. 11, in addition to the data identifier and the data body, the physical address of the server extracted from the request shown in FIG. 10 is associated. For example, the server that holds the requested data can specify the logical network path at the time of transferring the data together with the data having the corresponding identifier from the data holding unit 15.

  Next, the overall operation of this embodiment will be described in detail. Hereinafter, with respect to the “request transmission” when the server is a request transmission source, the “request reception” when the server receives a request, and the “data transfer” when the server transfers data, FIG. 2, FIG. This will be described with reference to FIG.

[When sending a request]
The operation at the time of request transmission in this embodiment is the same as that in the first embodiment (see FIG. 2).

[When receiving a request]
FIG. 12 is a flowchart showing the operation when the server 201 receives a request. The difference from the first embodiment shown in FIG. 7 is that when it does not hold data and it is determined that the data is cached (Yes in step S202), step S203 in FIG. Instead of S204, an operation of adding its own physical address to the physical address storage area of the request (see FIG. 10) is performed (step S204A). Note that the addition of the physical address is not limited as long as it is added so that the order of addition can be understood. For example, when adding, there is always a method of adding at the end.

  The second difference from the first embodiment shown in FIG. 7 is that it is included in the request after reading the requested data when it holds the data (Yes in step S201). The physical address group stored in the physical address storage area (see FIG. 11) of the read data (step S207A), the recently added address in the physical address group, the starting point of the logical network path The point is that data is transmitted to the server (step S208A).

[When transferring data]
FIG. 13 is a flowchart showing the operation when the server 201 receives data. The difference from the first embodiment shown in FIG. 8 is that instead of using the data transfer table, the physical address next to the physical address of the server 201 stored in the data is extracted and the data is transferred. Yes (step S302A).

  As described above, according to the second embodiment of the present invention, compared to the first embodiment, in addition to being able to omit the data transfer table 15, there is an effect that it is possible to cope with a server failure at the time of data transfer. Played. The reason is that the information of all servers to be transferred, that is, the logical network path configured by the server that is determined to be cached, is stored in the data itself, and even if a failure occurs in a certain server, This is because it can be transferred to the other server.

  In the above-described embodiment, the server 201 has been described as extracting the physical address next to the physical address of the server 201 stored in the data and transferring the data. However, before each server 201 performs the transfer, It is also possible to add a step of deleting the physical address used for the transfer from the data. In this way, each server can transfer data by referring to a specific portion of data (the beginning or end of a physical address group), and can reduce the number of physical addresses attached to the data. Become.

[Specific operation example]
Next, a specific operation of the present invention will be described using a simple network configuration shown in FIG.

  The solid line in FIG. 14 means a physical network connection. The broken line in FIG. 14 means a logical network between servers. Here, a one-layer straight logical network is assumed. The logical network may be a tree shape, a ring shape, or a mesh shape, and is not particularly limited in the present invention.

  Each server is given an identifier (ID) on the logical network. On the logical network, a route is determined based on this ID. Assuming that an IP address is assumed as the physical address of each server, the server 101a with ID1 is 1.1.1.1, the server 101b with ID2 is 2.2.2.2, and the server 101c with ID3 is 3.3. It is assumed that the IP address of 4.4.4.4 is set in the server 101d of 3.3 and ID4. Note that description of the port number is omitted.

  At this time, it is assumed that the logical network management information unit 13 of each of the servers 101a to 101b manages the correspondence between the ID on the logical network and the physical address. For example, the server 101b with ID2 has information that the IP address of the server 101a with ID1 is 1.1.1.1 and the IP address of the server 101c with ID3 is 3.3.3.3. .

  Here, for example, it is assumed that the server 101a with ID1 requests the data held by the server 101d with ID4 via ID2 and ID3 which are intermediate paths on the logical network. Here, it is determined that the server 101b with ID2 caches and the server 101c with ID3 does not cache.

  First, the server 101a with ID1 receives a request for data with identifier 999 from the client. The ID1 server 101a knows that the data requested by the external directory service or the database held by the ID1 server 101d is held by the ID4 server 101d.

  Then, as shown in FIG. 15, the server 101a with ID1 creates a request based on the acquired information, and stores its own IP address 1.1.1.1 in the request. The server 101a with ID1 transmits the created request to the server 101b with ID2.

  The server 101b of ID2 that has received the request determines that the data of the identifier 999 is cached as shown in FIG. 15, and the IP address stored in the request as the identifier 999 and the corresponding value in the data transfer table 16 1.1.1.1 is stored.

  Then, the server 101b with ID2 updates the request transfer destination server physical address 1.1.1.1 to its own IP address 2.2.2.2, and transmits the request to the server 101c with ID3.

  Since the server 101c with ID3 determines not to cache, the request is directly transferred to the server 101d with ID4.

  The server 101d with ID4 that has received the request holds the data with the identifier 999, and therefore reads the data from the data holding unit 15 and starts the transfer. At the time of transfer, data is transmitted to the server 101b of the transfer destination server physical address 2.2.2.2 stored in the request.

  The server 101b with ID2 that has received the data with the identifier 999 refers to the data transfer table 16 and transfers the data to the server 101a with the address 1.1.1.1. Then, the data transfer to the server 101a with ID1 that requested the data is completed.

  As described above, whether to cache at the time of the request is determined, and by not passing through the server that does not cache at the time of data transfer, the server that is not cached (the server 101c in FIG. 15) is not subjected to unnecessary load, and cache The bandwidth between the server that does not use and its router is not wasted, and the route to the server that does not cache is avoided, so that data can be acquired at a higher speed.

  The preferred embodiments of the present invention have been described above. However, the present invention is not limited to the above-described embodiments, and further modifications, replacements, and replacements may be made without departing from the basic technical idea of the present invention. Adjustments can be made. For example, in the above-described second embodiment, it has been described that each server sequentially adds its own physical address to the request when it is determined that the requested data is cached. If it is determined that the requested data is not cached, each server can delete the physical address of the server from the request.

  The present invention can be applied to uses such as data distribution and cache in a CDN, a wide area environment, and a data delivery system using P2P.

100 transmission path 101, 101a to 101d, 201 server (data transfer server)
DESCRIPTION OF SYMBOLS 10 Transmission / reception part 11 Request processing part 12 Data transfer part 13 Logical network information management part 14 Cache judgment part 15 Data holding part 16 Data transfer table

Claims (14)

Located in a network that relays data requests and data,
When a data request is received from another data transfer server, it is determined whether or not the requested data is to be cached. If it is determined that the data is to be cached, the logical network of the data transfer server that caches the data Execute the process to participate in
A data transfer server that transfers and caches data based on the request between data transfer servers participating in the logical network. When receiving a data request from another data transfer server, determine whether to cache the data, and if it is determined to cache the data, save the address information included in the data request, By storing its own address information in the request, it joins the logical network of the server that caches the data,
The data transfer server according to claim 1, wherein data transfer is performed using the address information stored when the request is received. When a request for data is received from another data transfer server, it is determined whether to cache the data. If it is determined that the data is to be cached, the address information is added to the request by adding its own address information. Join a logical network of data transfer servers that cache data,
The data transfer server according to claim 1, wherein data transfer is performed using address information included in the request.   The data transfer server according to claim 2 or 3, wherein when the data requested by the received request is held, the requested data is transmitted using address information stored in the request. . A request processing unit that transmits a request for data by referring to network information;
A cache determination unit that determines whether to cache the requested data;
A data transfer unit for transferring received data;
A data transfer server according to any one of claims 1 to 4.   The data transfer server according to claim 5, wherein the data transfer unit transmits the data to a data transfer server using the address information stored in the request.   The data transfer server according to claim 2, further comprising a data transfer table that associates and holds address information included in the data request and data.   The data transfer server according to claim 7, wherein the data transfer unit transfers received data using address information stored in the data transfer table.   The data transfer server according to any one of claims 1 to 8, wherein when the cache determination unit determines that the requested data is not cached, the request is transmitted as it is to a next data transfer server.   The data transfer server according to any one of claims 1 to 8, wherein when the cache determination unit determines that the requested data is not cached, a process of leaving the logical network of the data transfer server that caches the data is executed.   The data transfer server according to any one of claims 1 to 10, wherein the logical network is a part of a logical tree structure constituting a content delivery network.   A data transfer system comprising a plurality of data transfer servers according to any one of claims 1 to 11. A data transfer server located in a network that relays data requests and data,
When a data request is received from another data transfer server, it is determined whether or not the requested data is to be cached. If it is determined that the data is to be cached, the logical network of the data transfer server that caches the data The step of executing the process of participating in
A data transfer method comprising: transferring data based on the request and caching between data transfer servers participating in the logical network. A program for causing a data transfer server arranged in a network that relays a data request and data,
When a data request is received from another data transfer server, it is determined whether or not the requested data is to be cached. If it is determined that the data is to be cached, the logical network of the data transfer server that caches the data The process of executing the process of participating in
When data based on the request is received from a data transfer server participating in the logical network, the data cache and a process of transferring the data to another data transfer server participating in the logical network, A program to be executed by the data transfer server.

Images