JP2005190339A - Apparatus and method for data transfer, communication system, and program - Google Patents

Abstract

<P>PROBLEM TO BE SOLVED: To provide a data transfer apparatus which compresses target data based on reference data to transfer it and can more effectively manage reference data cache. <P>SOLUTION: A cache storage 503 in a server side apparatus 500 caches data which have been transmitted to a client side apparatus in the past as reference data. When a communication part 501 receives data from the server apparatus; the data are differentially compressed based on the reference data stored in the cache storage 503 in a differential compression part 502, and the differentially compressed data are transferred to the client side apparatus from the communication part 501. If total data size cached in the cache storage 503 exceeds a cacheable size, a deletion judge part 508 selects reference data so that more highly compressed data are to be deleted with high priority, and a deletion 504 deletes the selected reference data from the cache storage 503. <P>COPYRIGHT: (C)2005,JPO&NCIPI

Description

  The present invention relates to a data transfer device that performs data transfer, a communication system including a plurality of data transfer devices, a data transfer method, and a program.

  2. Description of the Related Art A client / server type information system is widely spread, which includes a server that provides various services via a network and a client that requests the server for a desired service. In particular, a WORLD WIDE WEB system (or simply referred to as a Web system) composed of a Web server and a client that communicates using the HTTP protocol on the Internet is a client-server type information system that is widely used. Normally, a server program operates on the server, and a predetermined tool (program) such as a browser operates on the client. The contents of services provided on the Internet are also diverse, providing information such as text, still images, moving images, audio (such as homepages, e-mails, digital contents, etc.) and programs via the network. Various services such as distribution or transfer services, electronic store services for selling products, reservation services for seats and rooms, various mediation services, etc. already exist, and new forms of services emerge one after another doing.

  By the way, in a client / server type information system such as the Web, the service is basically provided by transferring data between the client / server regardless of the form of the provided service. Is done. Therefore, the network capacity (bandwidth) used for communication between the client and the server tends to become a performance bottleneck of the entire system. Therefore, a cache technique is usually used to reduce the load on the network.

  When browsing the web, use a caching proxy server between the web browser and the web server or the cache of the web browser itself to reduce the amount of data transferred when browsing the web, and after receiving a request, receive a response to the web browser. There is a technique for increasing the time (response time) required for a series of sequences for displaying content. Squid (http://ring.aist.go.jp/archives/net/www/suid/) is well known as a caching proxy server. By placing squid on a network connecting a web server and a web browser, In response to a request from the browser, in the case of content that has been browsed and cached in the squid, a response is returned from the squid, thereby avoiding network transfer to the Web server and improving the response time. In addition, by using the cache provided in the web browser itself, if the content is already stored in the cache of the web browser when the request is issued from the web browser, the corresponding content is immediately displayed on the web browser, In addition to reducing the amount of data transferred through the network, the response time is improved. These cache technologies are effective means for obtaining a Web data transfer amount reduction effect and a response time improvement effect for a Web content (static content) whose content is uniquely determined with respect to a URL.

  However, with the recent evolution of web technologies, database search results and requests are sent when requests arrive at web servers such as CGI (Common Gateway Interface), ASP (Active Server Pages), JSP (Java Server Pages), etc. Dynamic content that is dynamically generated by user input and the like is increasing. These dynamic contents may differ every time for the same URL, and the contents are not uniquely determined for the URL. Conventional caching proxies and Web browser caches do not operate effectively for such dynamic content, and all content data must be transmitted and received between the network from the Web browser to the Web server for each request. In other words, a large amount of bandwidth is consumed and the response time is delayed.

  Patent Documents 1 to 3 propose a fingerprint cache that caches the content itself in a one-to-one correspondence with the hash value (fingerprint) of the content itself. In these technologies, a server-side proxy having a fingerprint cache as a proxy server is installed in a Web server-side LAN, and a client-side proxy having a fingerprint cache as a proxy server is installed in a client-side LAN. Web data flowing through networks such as the Internet, LAN, and WAN connecting between LANs is reduced. As the method, a fingerprint is calculated for the acquired dynamic content, and the fingerprint and the content are stored in the fingerprint cache. From the next time, if a response is received from the Web server, the server-side proxy calculates the fingerprint, and if it matches, the content itself is determined to match, and only the fingerprint is used instead of the content. The client-side proxy acquires content that matches the sent fingerprint from the fingerprint cache and transfers the content to the Web browser (FP transfer). In addition, when content similar to the response from the Web server is in the fingerprint cache, the server side proxy sends the difference data composed of the fingerprint and the difference part of the base content, and the client side device When the fingerprint content designated by the difference data is acquired from the fingerprint cache, the original content is restored using the difference portion data and transferred to the Web browser (difference transfer). By using these technologies, it is possible to reduce the Web transfer data by using a cache even for dynamic contents and improve the response time.

  The server-side proxy stores the content before differential compression in the fingerprint cache for use as a base from the next time onward, and the client-side proxy also stores the content after differential expansion in the fingerprint cache.

By the way, as a data storage area used for storing cache data, a main storage device or an auxiliary storage device connected to a computer that performs caching directly or via a network is used. With regard to the auxiliary storage device, it is possible to increase the capacity as well as to improve the redundancy by the RAID (Redundant Array of Independent) technique, but it is necessary to delete the cache because the data storage area is limited. Conventionally, an LRU (Least Recently Used) algorithm that deletes a static content cache in order from the oldest referenced date, an LFU (Least Frequently Used) algorithm that deletes content that is used less frequently, etc. is there. When the total amount of cached content exceeds the cacheable size, when an explicit deletion order is issued by the computer administrator, after a certain period of time, or when certain conditions are met When activated, the cache is deleted according to the above LRU or LFU or another algorithm, so that a recently accessed content or a frequently viewed content remains in the cache even after the deletion.
JP 2002-268935 JP 2003-268936 A JP2003-268937

  Conventionally, as in the fingerprint cache, the identifier that uniquely identifies the acquired content is cached as a pair, and the content identifier that becomes the base when the newly acquired content has the same part as the content in the cache The server side device calculates the difference information composed of the difference part and forwards it downstream, and the difference data received by the client side device is acquired from the cache with the specified identifier content and combined with the difference part to obtain the original data There is a system to restore. Such a technique is a very effective means for reducing the amount of transfer data of dynamic content that has become very popular in recent years.

  However, conventionally, an effective method for a dynamic content has not been known as a cache deletion method. Conventionally, all acquired contents are registered in a cache, and a more effective cache method has not been known. Conventionally, if there is a mismatch between the contents cached in the server side device and the client side device, and there is no content with the identifier specified by the difference data when restoring the difference data in the client side device, the difference There may be a case where the extension cannot be performed and the differential transfer fails, but a method for avoiding this has not been known.

  The present invention has been made in view of the above circumstances, and in a data transfer apparatus that compresses and transfers target data based on reference data, data transfer that enables more effective management of the reference data cache It is an object to provide an apparatus, a communication system, a data transfer method, and a program.

  The server apparatus-side data transfer apparatus according to the present invention includes a holding unit for holding at least data compressed in the past and transmitted to a specific client apparatus-side data transfer apparatus as reference data in an uncompressed state. Receiving means for receiving data destined for a certain server device and destined for a certain client device, and compressed data obtained by compressing the received data based on the reference data held in the holding means For each of the generating means for generating, the transmitting means for transmitting the generated compressed data to the data transfer apparatus on the client device side, and the reference data held in the holding means, the data size of the data is indicated. The first information, the data size of the data, and the data size of the compressed data generated from the data Storage means for storing second information indicating the degree of compression determined by: a sum of data sizes indicated by the first information for each of all reference data held in the holding means; and the holding means Is deleted from the holding means based on the second information relating to each reference data stored in the holding means when the relationship with the upper limit value of the data size that can be held satisfies a certain condition And a deletion unit that deletes the reference data selected by the selection unit from the holding unit.

  In addition, the client device-side data transfer device according to the present invention includes a holding unit for holding at least compressed data received from a specific server device-side data transfer device in the past as reference data in an uncompressed state. , Receiving means for receiving data from a server device as a transmission source and a client device as a destination from the data transfer device on the server device side, and whether the received data is uncompressed data Or a determination unit that determines whether the compressed data is compressed based on specific information added to the received data, and holds in the holding unit when the received data is compressed data Based on the reference data, the compressed data is decompressed to generate uncompressed data. First information indicating a data size of the data for each of the reference data held in the holding means, sending means for sending the generated data to the certain client device, and the holding data And storage means for storing second information indicating the degree of compression determined by the data size of the data and the data size of the compressed data from which the data is generated, and held by the holding means If the relationship between the sum of the data sizes indicated by the first information relating to each of all reference data and the upper limit value of the data size that can be held by the holding unit satisfies a certain condition, the holding is performed. Selection means for selecting reference data to be deleted from the holding means based on the second information relating to each reference data stored in the means, and selection by the selection means Characterized in that the reference data and a deleting means for deleting from said holding means.

Further, the communication system according to the present invention is a communication system including a server device side data transfer device and a client device side data transfer device, wherein the server device side data transfer device is compressed at least in the past to a specific client device side data. The first holding means for holding the data transmitted to the transfer device as reference data in an uncompressed state, and data destined for a certain client device as a transmission source. First receiving means, first generating means for generating compressed data obtained by compressing the received data based on the reference data held in the first holding means, and the generated compressed data Is held by the first holding means and the first holding means. For each of the reference data, first information indicating the data size of the data, and second information indicating the degree of compression determined by the data size of the data and the data size of the compressed data generated from the data Storage means for storing the information, a sum of data sizes indicated by the first information related to each of all the reference data held in the first holding means, and holdable by the first holding means When the relationship with the upper limit value of the data size satisfies a certain condition, the first and the second information on each reference data stored in the first holding unit is used. A selection unit that selects reference data to be deleted from the second holding unit, a first deletion unit that deletes the reference data selected by the selection unit from the first holding unit, and the selection unit. Information identifying the selected said reference data, and an identification information notification means for notifying the client apparatus side data transfer device,
A second holding unit for holding at least the compressed data received from the server device side data transfer device in the past as reference data in an uncompressed state; or Second receiving means for receiving, from the server apparatus-side data transfer apparatus, data destined for a server apparatus as a transmission source and a client apparatus as a destination, and the received data is uncompressed data Determination means for determining whether the received data is compressed data based on specific information added to the received data, and when the received data is compressed data, the first Based on the reference data held in the holding means, the compressed data is decompressed to generate uncompressed data The second generation means, the second transmission means for transmitting the generated data to the certain client apparatus, and the reference selected by the selection means from the server apparatus-side data transfer apparatus. Specific information notification acquisition means for acquiring notification of information specifying data, and second deletion means for deleting reference data specified by the acquired information from the second holding means And

  According to the present invention, in a data transfer apparatus that compresses and transfers data to be transferred based on reference data, it is possible to manage the reference data cache more effectively (more efficient for differential compression / decompression). A good cache maintenance method).

  For example, the difference between the server-side data transfer device and the client-side data transfer device with respect to dynamic content in which the URL and the content are not uniquely determined (for example, the content differs for each access generated by CGI or the like) Dynamic content by introducing means for selecting content to be deleted from the cache based on the compression rate in differential compression when reducing the amount of data transferred between them by compressing and transferring Content suitable as a base for differential compression can be maintained in the cache. Further, for example, by providing this means only in the server-side data transfer device, the server-side data transfer device notifies the client-side data transfer device of the content to be deleted, and synchronizes each other's cache. In addition, it is possible to eliminate the failure of the differential transfer caused by the cache mismatch between them, and to reduce the processing load of the client-side data transfer device. Further, for example, by introducing a means for determining whether or not to register the content in the cache based on the compression rate in differential compression, registration processing and effective use of the cache area can be performed. Further, for example, this means is provided only in the server-side data transfer device, and the server-side data transfer device notifies the client-side data transfer device of the presence / absence of registration of each content so as to synchronize each other's cache. Thus, it is possible to eliminate the failure of the differential transfer caused by the cache mismatch between them, and to reduce the processing load of the client side data transfer apparatus.

The present invention relating to the apparatus is also established as an invention relating to a method, and the present invention relating to a method is also established as an invention relating to an apparatus.
Further, the present invention relating to an apparatus or a method has a function for causing a computer to execute means corresponding to the invention (or for causing a computer to function as means corresponding to the invention, or for a computer to have a function corresponding to the invention. It can also be realized as a program (for realizing the program), and can also be realized as a computer-readable recording medium on which the program is recorded.

  According to the present invention, it is possible to more effectively manage the cache of reference data in a data transfer apparatus that compresses and transfers data to be transferred based on reference data.

  Hereinafter, embodiments of the present invention will be described with reference to the drawings.

  The Web server, proxy, and Web browser of this embodiment can all be realized by operating software (Web server program, proxy program, Web browser program) on a computer. In this case, an OS or driver software having a function desired by a computer, or hardware such as a communication interface device, an external storage device, or an input / output device is mounted or connected as necessary. In this case, it is preferable to use a graphical user interface (GUI) for information from the user or the administrator, information presentation to the user, or the like.

  In the present embodiment, communication in the World Wide Web system is performed according to the HTTP protocol. The HTTP protocol will be described using HTTP 1.0 shown in RFC1945 or HTTP1.1 shown in RFC2616.

  In the following description, the case where the Internet is a WAN (Wide Area Network), the client is a PC connected to the LAN of the user office, and the HTTP protocol is used will be described as an example. In this case, the present invention can be applied even if the WAN is, for example, the Internet, a wireless network using a PHS, a mobile phone, or the like, or other than the WAN. Further, the present invention can be applied even if the client is installed in a home LAN other than the office, for example, or a protocol other than the HTTP protocol is used.

  FIG. 1 shows a configuration example of a computer network system to which the present invention is applied. In this configuration example, a local area network (LAN) 100 in an ASP (Application Service Provider) server center and a LAN 104 in a user office are connected via a wide area network WAN 103 such as the Internet or a dedicated line. A server 101 connected to the LAN 100 in the ASP server center and a client 106 connected to the LAN 104 in the user office can communicate via the LAN 104, the WAN 103, and the LAN 100. One or more servers are connected to the LAN 100 in the ASP server center, and one or more clients are connected to the LAN 104 in the user office.

  The Web-based ASP provides services based on various application programs from the server 101 installed in the LAN 100 in the server center via the WAN 103, and the user uses the Web browser 107 on the client 106 installed in the LAN 104 in the user office. To access those services.

  In such a usage mode, the effective communication capacity (bandwidth) of the network 104 connecting the LAN 104 of the user office and the LAN 100 of the server center, particularly the wide area network 103 such as the Internet, is larger than that of the LAN 104 of the server center or the LAN 100 of the user office. This is a low performance bottleneck in performance, causing communication delays and reducing application response time.

  Therefore, in this embodiment, as shown in FIG. 1, two modules, a server-side proxy 105 and a client-side proxy 102, are installed at both ends of a wide area network 103 that connects the LAN 104 of the server center and the LAN 100 of the user office. The bottleneck of the wide area network is eliminated by reducing the amount of communication data by performing differential compression described later.

  FIG. 2 shows another configuration example of a computer network system to which the present invention is applied. In this configuration example, a LAN 110 in a business application center in a company and a mobile client 116 such as a notebook PC are connected via a wide area network WAN 113 such as the Internet or a dedicated line and a wireless network 115 such as a mobile phone or PHS. The server 111 connected to the LAN 110 in the business application center and the client PC 116 can communicate with each other via the wireless network 115, the WAN 113, and the LAN 110. One or more servers are connected to the business application LAN 110, and one or more clients are connected via the WAN 113 or the wireless network 115.

  The server 111 installed in the LAN 110 of the Web-based business application center provides various services through various enterprise fixed business application programs and business application programs via the WAN 103, and the user provides a Web browser 118 on the client 116. To access these services.

  In such a usage mode, the bandwidth of the network 110 connecting the business application center LAN 110 and the client 116, particularly the wide area network 113 such as the Internet and the wireless network 115 is lower than the LAN 110 of the business application center, which becomes a bottleneck. This causes a problem that the response time of the application decreases.

  Therefore, in the present embodiment, as shown in FIG. 2, the WAN 113 that connects the LAN 110 of the business application center and the client 116 and two modules of the server side proxy 112 and the client side module 117 on the client 116 are provided at both ends of the wireless network 115. The bottleneck is eliminated by installing and reducing the amount of communication data by performing differential compression described later between them.

  However, in the configuration example of FIG. 1, the server-side device is realized as a server-side proxy, and the client-side device is realized as a client-side proxy. However, the configuration is not limited to this. In the configuration example of FIG. 2, the server-side device is realized as a server-side proxy and the client-side device is realized as an application. However, the configuration is not limited to this. In addition, as long as the server-side device and the client-side device are realized as a component constituting the server-side system and a component constituting the client-side system, respectively, the realization form is not limited.

  Next, the differential compression processing and differential expansion processing will be described with reference to FIG.

  1 and FIG. 2, the server-side proxy 105 in FIG. 1 and the server-side proxy 112 in FIG. 2 correspond to the server-side device 200 in FIG. 3, respectively, and the client-side proxy 102 in FIG. Each client-side module 117 corresponds to the client-side device 207 in FIG.

  A cache 205 connected directly or via a network is installed in the server side apparatus 200, and a cache 212 connected directly or via a network is installed in the client side apparatus 207. Assume that the cache 205 of the server-side device 200 stores content A 203 and content B 204 that have been transferred as HTTP responses by the Web server in response to an HTTP request such as a GET method from a client having a Web browser. An identifier A that can uniquely identify the content A is assigned to the content A203, and an identifier B that can uniquely identify the content B is assigned to the content B204.

  This identifier may be any identifier as long as it uniquely identifies the content itself. For example, it may be numbered in ascending order in the acquired order, or may be calculated using a hash function represented by MD5, SHA-1, or the like.

  In the cache 205 of the server side device 200, the content A is stored in a one-to-one relationship with the identifier A, and similarly, the content B is stored in a one-to-one relationship with the identifier B. On the other hand, similarly, identifier A and content A 210 are stored in a one-to-one relationship and identifier B and content B 211 are stored in a one-to-one relationship in cache 212 of client-side device 207. The client-side device 207 may acquire an identifier for each content with the same algorithm as that of the server-side device 200. Alternatively, the client-side device 207 obtains an identifier from the identifier added to the header of the HTTP response of the content A transferred from the server-side device 200. You may get it.

  An example of the HTTP response when added to the header of the HTTP response is shown in FIG. In FIG. 4, the server-side device 200 adds the X-Identifier header to the HTTP response header and transmits it to the client-side device 207. The client-side device 207 receives the response and acquires the value of the X-Identifier header. The identifier A of the content A is acquired. It is desirable that the client side device 207 deletes the X-Identifier header when transmitting the content A downstream.

  By following the above procedure, it is possible for the client side device 207 to associate the content A 210 and the identifier A on a one-to-one basis and store the content B 211 and the identifier B on a one-to-one basis in the cache 212. It becomes.

  The outline of the differential compression processing and differential expansion processing will be described below with reference to the processing flow shown in FIG. 3 on the premise of the server side device 200 and the client side device 207 placed in the above-described state.

  (1) The server-side device 200 receives new content C201 from upstream. The identifier of the content C201 is C.

  (2) The content C201 is subjected to differential compression processing. At this time, it is assumed that the content A203 portion 2033 and the content B204 portion 2041 cached in the cache 205 match the content C portion 2023 and the portion 2021, respectively. Further, it is assumed that the portion 2022 of the content C202 does not have the same portion as any content stored in the cache 205.

  (3) As a result of differential compression, differential data 206 is generated. The contents of the difference data 206 are the offset byte and length from the beginning of the identifier B equivalent to the portion 2021, the content (original) of the portion 2022 as it is, the offset and length from the beginning of the identifier A equivalent to the portion 2023 Consists of three elements. The generated difference data 206 is transferred to the client side device 207.

  (4) Also, before or after (2) and (3), the content C201 is cached in the cache 205 for use in the subsequent differential compression together with the identifier C.

  (5) The client side device 207 receives the difference data 206.

  (6) The differential data 206 is subjected to differential expansion processing. At this time, the content 211 of the identifier B embedded in the difference data 206 is acquired from the cache 212, and the portion 2111 is extracted from the offset and length information. Further, the part as it is embedded in the difference data 206 is taken out. Further, the content 210 with the identifier A embedded in the difference data 206 is acquired from the cache 212, and the portion 2111 is extracted from the offset and length information. Finally, as the difference data 206, a content 208 is generated by combining a portion 2081 corresponding to the portion 2111 of the content B 211, a portion 2082 as it is, and a portion 2083 corresponding to the portion 2103 of the content A 210.

  (7) The content 208 is given the identifier C specified by the server side device 200 in the HTTP header, and becomes the content C 209 of the identifier C. Content C209 is transmitted downstream.

  (8) The content C209 is cached in the cache 212 together with the identifier C for use in the next and subsequent differential expansion.

  When exactly the same content as the content in the cache is received, the identifier and offset of the cached content are represented by differential data that is 0 and the length is the size of the content in the differential compression processing.

  FIG. 5 shows the response of the difference data transmitted from the server side device 200 in the process (3) according to the HTTP protocol.

  In FIG. 5, the Content-Type: octet-stream / x-diff-compression header indicates that the body is differentially compressed data. The client side device 207 identifies that the body of the response is differential data by looking at this header.

  The X-Identifier: identifier C header is an identifier given to the content after differential restoration, but is unnecessary when the client side device 207 can acquire the identifier by the same algorithm as the server side device 200.

  The Pragma: no-cache, x-diff header is a header attached by the server side device according to the HTTP protocol so that differential data is not cached by a caching proxy between the server side device 200 and the client side device 207. . The same applies to the Cache-Control: no-cache, x-diff header.

  X-Pragma: no-cache and X-Cache-Control: private are Pragma and Cache-Control headers to be added to the restored content. These are obtained by saving the HTTP header added to the content before differential compression in the server side device 200. The client-side device 207 corrects the value of Content-Type header (MIME / TYPE) to the original content value, deletes the X-Identifier header, deletes the Pragma header and Cache-Control header, and deletes the X-Identifier header. -The process of replacing the value of the Pragma header with the value of the Pragma header and replacing the value of the X-Cache-Control header with the value of the Cache-Control header is performed.

  FIGS. 6 and 7 show the server-side device 200 and the client-side device 207 shown in FIG. 3 in functional block diagrams, respectively.

  As illustrated in FIG. 6, the server-side device 300 includes a communication unit 301 that transmits and receives data, a differential compression unit 302 that performs differential compression processing, and server-side registration that stores content before differential compression in a server-side cache holding unit. The unit 303 is configured.

  On the other hand, as illustrated in FIG. 7, the client side device 305 includes a communication unit 306 that transmits and receives data, a differential restoration unit 307 that performs differential decompression processing, and a client that stores the restored content in the client side cache holding unit 309. The side registration unit 308 is configured.

  Here, the server side device 300, the server side cache holding unit 304 in FIG. 6, the client side device 305, and the client side cache holding unit 309 in FIG. 7 are respectively the server side device 200, the cache 205, and the client side device in FIG. 207 corresponds to the cache 212.

  The outline of the differential compression process and the differential decompression process has been described above. However, the HTTP header configuration and the differential data format are not limited as long as they can realize the same as described above. good.

  By the way, conventionally, an appropriate algorithm has not been known as a method for deleting a cache of a server side apparatus and a client side apparatus that take a difference with respect to dynamic content. Accordingly, some specific examples of the cache deletion method will be described below.

  First, an example of a cache deletion method will be described with reference to FIG.

The histogram (400) shown in FIG. 8 is obtained by taking the size of the cached content on the vertical axis and the compression ratio of the differential data acquired at the time of differential compression on the horizontal axis. It is assumed that it is placed in a storage device (memory). For example, when the content size before compression is S and the size of the difference data after compression is ΔS, the compression rate is expressed by equation (1).
Compression rate = 100 × (1−ΔS / S) (1)
(In this example, the smaller the difference data size ΔS after compression is, the larger the compression rate is).

  FIG. 9 shows an example of a registration processing procedure for the histogram (400). First, the content identifier, size, and compression rate are acquired (step S1). When the total size of the contents in the list does not exceed the cacheable size after addition (step S2), CR is calculated (normalized) by CR = (int) ((compression rate / 10) × 10) (step S4). . On the other hand, when the total size of the contents in the list exceeds the cacheable size when added (step S2), after the deletion process is executed (step S3), CR is calculated (normalized) (step S4). And it registers in the grid corresponding to CR% to (CR + 10)% (step S5).

  FIG. 10 shows an example of a procedure for deleting the histogram (400) and the cache. First, all the contents belonging to the grid with the highest compression ratio among the grids that have not been deleted are deleted (step S11). If the cache size is 80% or less (step S12), the process is terminated. On the other hand, if the cache size exceeds 80% (step S12), all the contents belonging to the grid with the highest compression ratio among the undeleted grids are deleted (step S11). The above processing is repeatedly executed until the cache size becomes 80% or less.

  In the specific example of FIG. 8, it is assumed that the contents acquired in time series are A, B, C, D,..., M, N, O. A, B, C, etc. are identifiers of the contents. The compression rate is calculated by the equation (1) from the size of each content before differential compression and the size after differential compression. At that time, the compression ratio of each content is set on the horizontal axis (in the example of FIG. 8, the grid width on the horizontal axis is 10%), and the size of each content is set on the vertical axis. It is assumed that registration in the histogram (400) by the registration procedure shown in FIG. 9 results in the state illustrated in FIG. That is, the compression rate of content A is between 0 and 10%, the compression rate of content B is between 20 and 30%, the compression rate of content C is between 60 and 70%, and so on.

  Therefore, in the state illustrated in FIG. 8, when the content P is newly acquired and the compression ratio is p, it is assumed that the total size of the content cached in the processing flow of FIG. 9 exceeds the cache capacity. In this case, the deletion process (step S3) in FIG. 9 is executed, and according to the deletion process flow shown in FIG. 10, until the total size of the already cached content becomes equal to or smaller than a predetermined reference value, The higher one is deleted preferentially (in FIG. 10, as an example, this reference value is 80% of the cache capacity).

  The predetermined reference value is, for example, 80% of the cache capacity, but is not limited to this, and may be other ratio to the cache capacity or determined regardless of the cache capacity. It may be a value. However, if the cached content is deleted too much, the content that can be used for differential compression will be reduced and the differential compression effect will be reduced, and even if it is left too much, this deletion processing will occur frequently and cause a performance problem, so it is necessary to use a reasonable value .

  Taking an example according to the processing flow with respect to FIG. 9, for example, when the newly acquired content P is content with a compression rate of 32.4%, CR = (int) ((32. 4/10) × 10) = (int) (3 × 10) = 30, which is registered in the grid corresponding to the compression rate of 30% to 40%.

  In the case of the above specific example, if the contents C having a compression ratio of 60 to 70% and the contents O, L, and I having a compression ratio of 50 to 60% are deleted, the total size of the cached contents is 80. % Or less. The total size of the cache is calculated every time the contents included in the grid are deleted in order from the grid with the highest compression rate. In this specific example, the total size of the contents still remaining in the cache after deleting the contents O, L, and I included in the grid (401) corresponding to the compression ratio of 50% to 60% is 80% or less of the cache capacity. Therefore, the grid (402) corresponding to the compression rate of 40% to 50% was not deleted. Further, the content P is registered in an appropriate grid of the histogram and cached according to the value and size of the compression rate p.

  In this specific example, the grid width is set to 10% using a histogram, and the total size of contents remaining in the cache after deletion is set to 80% of the cache capacity with respect to the cache capacity at the time of deletion. The present invention is not limited, and any form may be used as long as a high compression ratio is deleted preferentially.

  Next, another example of a cache deletion method for dynamic content will be described with reference to FIG.

  In the specific example of FIG. 11, when contents acquired in time series are arranged, it is assumed that they are A, B, C, D,..., M, N, O (similar to the previous example). FIG. 11 shows a method of deleting the cached contents in descending order of the compression ratio by using the bi-directional linked list (450) having the head and tail installed on the memory. In the list head (451) of FIG. 11, contents with a low compression rate are registered in the linked list (450) as identifiers, sizes and compression rates in order. For example, the content of the identifier A is registered in the element 452 as an identifier A, a size 10025 (Bytes), and a compression rate of 0.0%. Similarly, all cached contents are registered between the list head (451) and the list tail (462) in ascending order of compression rate.

  FIG. 12 shows an example of a processing procedure when registering contents to be cached in the linked list (450). First, an identifier, a size, and a compression rate are acquired (step S21). If the total size of the contents in the list does not exceed the cacheable size after addition (step S2), the process proceeds to step S24. On the other hand, if the total size of the contents in the list exceeds the cacheable size when added (step S22), the deletion process is executed (step S23), and then the process proceeds to step S24. If the compression ratio is 50% or more in step S24, the search is started from the list tail (step S25), the list is searched for an element smaller than its own size (step S26), and is registered as an element immediately after the corresponding element (step S26). Step S27). On the other hand, if the compression rate exceeds 50% in step S24, the search is started from the list head (step S28), the list is searched for an element larger than its own size (step S29), and the element immediately before the corresponding element is searched. Register as an element (step S30).

  FIG. 13 shows an example of the procedure for deleting the histogram (400) and the cache. First, starting from the list tail (step S31), the previous element is traced (step S32), and the element is deleted (step S33). Here, if the total size of the contents remaining in the cache is 80% or less (step S34), the process is terminated. On the other hand, if the total size of the content remaining in the cache exceeds 80% (step S34), further elements are deleted (step S33). The above processing is repeatedly executed until the total size of the contents remaining in the cache becomes 80% or less.

  Now, it is assumed that the state shown in FIG. 11 is achieved by registering the content according to the registration procedure shown in FIG. However, in the state of FIG. 11, it is assumed that the deletion process (step S23) shown in FIG. 12 has not been performed yet.

  When newly acquiring content in FIG. 11 and registering it in the linked list, according to the procedure in FIG. 12, if the content identifier P, size 7634, and compression factor of 18.5 are registered, cached Since the total size of the contents exceeds the cacheable size, the deletion process (step S23) is executed. In the deletion process, elements are deleted in order from the list tail, that is, in descending order of the compression ratio, and the deletion process ends when the total size of the contents remaining in the cache falls below a predetermined reference value. . Subsequently, the remaining part of FIG. 12 is executed, and the series of processing ends (in FIG. 13, as an example, this reference value is 80% of the cache capacity). FIG. 11 shows a case where identifiers C, I, O, and L are deleted as a result of the above processing.

  In this specific example, using the linked list, the total size of the contents remaining in the cache after deletion is 80% of the cache capacity with respect to the cache capacity at the time of deletion. However, the present invention is not limited to this. However, any form may be used as long as the one with a high compression rate is deleted preferentially.

  Two examples have been shown above, but if such a high compression rate is deleted preferentially, redundant content is deleted from the cache when a difference is taken for dynamic content. become. Because, if one or more contents are already in the cache enough to differentially compress the content α when it is acquired, the newly acquired content α is compressed by the content already in the cache And the compression rate becomes high. Furthermore, even if the content α ′ very similar to the content α is acquired, the content α and α ′ can be sufficiently compressed by one or a plurality of contents already in the cache as in the case of the content α. This is because there is no need to leave it in the cache.

  Next, FIGS. 14 and 15 show functional blocks necessary for realizing the above-described deletion method. 14 and 15 show functional blocks of the server side device 500 and the client side device 520, respectively.

  The server-side device 500 includes a communication unit 501 that receives an HTTP request from the downstream and transmits the HTTP request upstream, or receives an HTTP response from the upstream and transmits an HTTP response such as differential data downstream. ing.

  Upon receiving the HTTP response from the upstream side, the server side device 500 performs differential compression processing (refer to the part indicated by the circled numeral 3 in FIG. 3) by the differential compression unit 502 and downstream through the communication unit 501, the differential compressed data , The server side registration unit 505 caches the content before differential compression in the server side cache holding unit 503. At the same time, the server side compression rate calculation unit 506 calculates the compression rate using the equation (1) from the size of the difference data acquired from the difference compression unit 502 and the content size before the difference compression acquired from the communication unit 501. The server-side size acquisition unit 509 acquires the size of the content before differential compression from the communication unit 501 and registers it in the server-side size / compression rate holding unit 507.

  The server-side size / compression rate holding unit 507 corresponds to the histogram (400) in FIG. 8, and corresponds to the bidirectional linked list (450) in FIG.

  When the total size of the cached content held by the server-side size / compression rate holding unit 507 exceeds the cacheable size (usually set by the user) acquired from the server-side cacheable size holding unit 510, the server side The deletion determination unit 508 detects one or a plurality of deletion target contents, and the server side deletion determination unit 504 deletes the corresponding deletion target contents from the server side cache holding unit 503.

  The server-side deletion determination unit 508 corresponds to the processing flow shown in FIG. 10 in the form described in FIG. 8, and corresponds to the processing flow shown in FIG. 13 in the form described in FIG.

  The client-side device 520 includes a communication unit 521 that receives an HTTP request from the downstream and transmits the HTTP request upstream, or receives an HTTP response from the upstream and transmits an HTTP response such as difference data to the downstream. ing.

  Upon receiving an HTTP response including differentially compressed data from upstream, the client side device 520 performs a differential decompression process (refer to a part indicated by a circled number 6 in FIG. 3) by the differential decompression unit 522 and via the communication unit 521. When the expanded content is transmitted downstream, the client side registration unit 525 caches the content after the differential expansion in the client side cache holding unit 523. At the same time, the client side compression rate calculation unit 526 calculates the compression rate using the equation (1) from the size of the difference data acquired from the communication unit 521 and the size after the difference expansion acquired from the difference expansion unit 522, and The client-side size acquisition unit 529 acquires the size of the content after differential expansion acquired from the differential expansion unit 522 and registers it in the client-side size / compression rate holding unit 527.

  The client-side size / compression rate holding unit 527 corresponds to the histogram (400) in FIG. 8, and corresponds to the bidirectional linked list (450) in FIG.

  When the total size of the cached content held by the client side size / compression rate holding unit 527 exceeds the cacheable size (usually set by the user) acquired from the client side cacheable size holding unit 530, the client side The deletion determination unit 528 detects one or a plurality of deletion target contents, and the client side deletion determination unit 524 deletes the corresponding deletion target contents from the client side cache holding unit 523.

  The client-side deletion determination unit 528 corresponds to the processing flow shown in FIG. 10 in the form described in FIG. 8, and corresponds to the processing flow shown in FIG. 13 in the form described in FIG.

  In the configuration example described above, the cache deletion method is performed independently by the server side device and the client side device. However, in some cases, the client-side device is a small-sized terminal such as a PDA (Personal Digital Assistants) or a mobile phone, and a cache area is allocated on the memory. In this case, both the client-side cache holding unit 523 and the client-side size / compression rate holding unit 527 shown in FIG. 15 need to be installed in the memory. The size of the client-side cache holding unit 523 and the size shown in FIG. The size of the server side cache holding unit 503 is extremely different. In addition, since it is necessary to determine the deletion target by the client-side deletion determination unit 527 in a small portable terminal with low system performance, there is a case where the client-side device does not want to perform the above determination from the viewpoint of processing amount.

  Therefore, in the following, an example of a method for realizing a difference in the cache area size and a reduction in processing amount of the client side device in cooperation between the server side device and the client device will be described.

  16 and 17 are functional block diagrams of the server side device 600 and the client side device 620, respectively. The client side device 620 is a small portable terminal such as a PDA or a portable terminal.

  The client-side device 620 omits the client-side size / compression rate holding unit 527, the client-side compression rate calculation unit 526, and the client-side deletion determination unit 528 shown in FIG. Since the client side size / compression rate holding unit 527 is omitted, the memory usage is reduced, and the client side compression rate calculation unit 526 and the client side deletion determination unit 528 are omitted, so that the processing amount is reduced. On the other hand, as functions newly added in FIG. 17, there are a client-side cacheable size notification unit 627 and a client-side deletion notification acquisition unit 626. Further, in the server side device 600, a cacheable size holding unit 610, a client side cacheable size acquisition unit 612, and a server side deletion notification unit 611 are added instead of the server side cacheable size holding unit 510.

  First, the client-side device 620 transmits the client-side cacheable size to the server-side device 600 via the communication unit 621 by the client-side cacheable size notification unit 627 in advance.

  The server-side apparatus 600 acquires the client-side cacheable size received by the communication unit 601 by the client-side cacheable size acquisition unit 612. The server-side cacheable size is compared with the client-side cacheable size, and a smaller value is registered in the cacheable size holding unit 610. Here, the cacheable size of the client side device 620 is obviously small, and the client side cacheable size is registered in the cacheable size holding unit 610.

  The client-side cacheable size notified from the client-side cacheable size notification unit 627 of the client-side device 620 to the client-side cacheable size acquisition unit 612 via the communication unit 621 and the communication unit 601 of the server-side device 600 is, for example, As shown in FIG. 18, it can be realized by using the OPTIONS method of HTTP 1.1. 18, the server side device 600 (host name: hogehoge.xxx.co.jp) uses the X-Cache-Size header to indicate that the cacheable size of the client side device 620 is 32 (MBytes). ) To send notifications. The response header is transmitted to the client side device 620 in order to notify that the server side device 600 has normally received the client side cacheable size from the corresponding request. However, instead of using HTTP as described above, it may be set by another protocol or a human hand, and the form is not limited.

  As described above, the server side device 600 acquires the cacheable size of the client side device 620 in advance and determines the value of the cacheable size holding unit 610.

  The server-side device 600 includes a communication unit 601 that receives an HTTP request from the downstream and transmits the corresponding HTTP request upstream, or receives an HTTP response from the upstream and transmits an HTTP response such as difference data downstream. ing.

  Upon receiving the HTTP response from the upstream side, the server side device 600 performs differential compression processing (refer to the part indicated by the circled numeral 3 in FIG. 3) by the differential compression unit 602, and downstream the differential compressed data via the communication unit 601. Is transmitted, the server-side registration unit 605 caches the content before differential compression in the server-side cache holding unit 603. At the same time, the server-side compression rate calculation unit 606 calculates the compression rate using Equation (1) from the size of the difference data acquired from the difference compression unit 602 and the content size before differential compression acquired from the communication unit 601. The server-side size acquisition unit 609 acquires the size of the content before differential compression from the communication unit 601 and registers it in the server-side size / compression rate holding unit 607.

  The server-side size / compression rate holding unit 607 corresponds to the histogram (400) in FIG. 8, and corresponds to the bidirectional linked list (450) in FIG.

  When the total size of the cached content held by the server-side size / compression rate holding unit 607 exceeds the cacheable size acquired from the cacheable size holding unit 610, the server-side deletion determination unit 608 performs one or more The deletion target content is detected, the server side deletion determination unit 604 deletes the deletion target content from the server side cache holding unit 603, and the server side deletion notification unit 611 sends the deletion target content to the client side device via the communication unit 601. 600. However, the deletion target content is notified as a part of the response header when the difference data is transmitted to the client side device 620.

  The server-side deletion determination unit 608 corresponds to the processing flow shown in FIG. 10 in the form described in FIG. 8, and corresponds to the processing flow shown in FIG. 13 in the form described in FIG.

  The client side device 620 receives the difference data transmitted from the server side device 600 and the response including the content to be deleted by the communication unit 621. The differential decompression unit 622 determines that the corresponding response is differential data, and performs differential decompression processing. The restored content after the decompression process is transmitted downstream via the communication unit 621 and is stored in the client side cache holding unit 623 by the client side registration unit 625. In addition, when there is content to be deleted in the response header, the client side deletion notification acquisition unit 626 deletes the content with the specified identifier from the client side cache holding unit 623 by the client side deletion unit 624.

  FIG. 19 shows an example of an HTTP response specifying the difference data and the content to be deleted. In FIG. 19, the identifier of the content to be deleted is the case of C, O, L, I shown in the example of FIG. 8 or FIG. 11, and by specifying the identifier in the X-Purge-Identifier header, The client side device 620 determines that the content to be deleted is C, O, L, I by the client side deletion notification acquisition unit 626 and deletes it from the client side cache holding unit 623 by the client side deletion unit 624.

  In this way, by coordinating the server-side device and the client-side device with clearly different cacheable sizes, the processing amount of the client-side device is reduced, and the differential decompression processing failure due to a mismatch of cached contents is prevented. The amount of data transferred between the server side device and the client side device can be further reduced.

  In the above description, the content to be deleted is specified as a part of the response header of the difference data. Of course, the client side apparatus may be notified separately using another protocol.

  Incidentally, the configuration example shown in FIG. 14 and FIG. 15 shows an effective deletion method for deleting the cache of dynamic content, but in this case, the content that is subsequently subjected to differential compression even after the cache is deleted Are newly cached regardless of the compression ratio. If a new content that is sufficiently compressed by one or more contents already in the cache is acquired, the compression rate increases. In this case, it is possible to cache dynamic content more efficiently by determining whether to register in the cache based on the compression ratio before registering in the cache.

  This determination requires a threshold value of a certain compression ratio, and this threshold value can be acquired by the configuration of FIG. 8 or FIG. 11, for example. In FIG. 8, for example, the minimum value of the grid when the cache size becomes 80% or less in the deletion processing flow of FIG. In the specific example of FIG. 8, a value of T = 50 can be acquired (threshold value T (403) of FIG. 8). In FIG. 11, for example, the compression rate of the content of the last element deleted when the cache size becomes 80% or less in the deletion processing flow of FIG. In the case of the specific example of FIG. 8, a value of T = 54.5 (threshold value T (464) of FIG. 11) can be acquired.

  This threshold value T is updated by the above method every time deletion processing is performed. The initial value of T before the first deletion process is desirably 100%, but is not limited to this.

  Hereinafter, an example of a dynamic content registration determination method will be described.

  FIG. 20 shows an example of a processing procedure for determining whether to register in the cache using the threshold value T acquired as described above. First, a compression rate is acquired (step S41). If the compression rate is T or less (step S42), registration is performed (step S43). On the other hand, if the compression rate exceeds T (step S42), it is not registered (step S44).

  FIG. 21 shows a functional block of the server side apparatus when the determination unit based on the compression rate threshold shown in FIG. 20 is included. Compared to FIG. 14, a server-side registration determination unit 711 is newly added.

  A series of processing flow in FIG. 21 is as follows.

  The server side device 700 is provided with a communication unit 701 that receives an HTTP request from the downstream and transmits the corresponding HTTP request upstream, or receives an HTTP response from the upstream and transmits an HTTP response such as differential data downstream. ing.

  When the server side device 700 receives the HTTP response from the upstream side, the server side device 700 performs the differential compression processing (refer to the portion indicated by the circled numeral 3 in FIG. 3) in the differential compression unit 702, and differentially compressed data downstream via the communication unit 701. Send. Subsequently, the server-side registration determination unit 711 acquires the compression rate from the server-side compression rate calculation unit 706 according to the equation (1), acquires the threshold T at that time from the server-side deletion determination unit 708, and FIG. It is determined whether or not the content before differential compression is cached by the indicated determination process. If the server-side registration determination unit 711 determines not to register, the processing ends there. If it is determined to be registered, the server-side registration unit 705 caches the content before differential compression in the server-side cache holding unit 703. At the same time, the compression rate and size acquired from the server side compression rate calculation unit 706 and the server side size acquisition unit 709 are registered in the server side size / compression rate holding unit 707.

  The server-side size / compression rate holding unit 707 corresponds to the histogram (400) in FIG. 8, and corresponds to the bidirectional linked list (450) in FIG.

  When the total size of the cached content held by the server-side size / compression rate holding unit 707 exceeds the cacheable size acquired from the server-side cacheable size holding unit 710, the server-side deletion determination unit 708 performs one or A plurality of deletion target contents are detected, and the server side deletion determination unit 704 deletes the corresponding deletion target contents from the server side cache holding unit 703.

  The server-side deletion determination unit 708 corresponds to the processing flow shown in FIG. 10 in the form described in FIG. 8, and corresponds to the processing flow shown in FIG. 13 in the form described in FIG. In addition, the server-side deletion determination unit 508 recalculates the threshold T upon deletion.

  FIG. 22 shows functional blocks of the client side apparatus when the determination unit based on the compression rate threshold shown in FIG. 20 is inserted. Compared to FIG. 15, a client side registration determination unit 731 is newly added.

  A series of processing flow in FIG. 22 is as follows.

  The client side device 720 includes a communication unit 721 that receives an HTTP request from the downstream and transmits the corresponding HTTP request upstream, or receives an HTTP response from the upstream and transmits an HTTP response such as difference data downstream. ing.

  When the client side device 720 receives an HTTP response including differentially compressed data from the upstream side, the client side device 720 performs a differential decompression process (refer to a part indicated by a circled number 6 in FIG. 3) by the differential decompression unit 722, via the communication unit 721. Then, the expanded content is transmitted downstream. Subsequently, the client-side registration determination unit 731 acquires the compression rate from the client-side compression rate calculation unit 726 according to the equation (1), acquires the threshold T at that time from the client-side deletion determination unit 728, and FIG. It is determined whether or not the content after differential expansion is cached by the indicated determination process. If the client side registration determination unit 731 determines not to register, the processing ends there. If it is determined to be registered, the client side registration unit 725 caches the content after differential expansion in the client side cache holding unit 723. At the same time, the compression rate and content size are acquired from the client side compression rate calculation unit 726 and the client side size acquisition unit 729 and registered in the client side size / compression rate holding unit 727.

  The client-side size / compression rate holding unit 727 corresponds to the histogram (400) in FIG. 8, and corresponds to the bidirectional linked list (450) in FIG.

  When the total size of the cached content held by the client side size / compression rate holding unit 727 exceeds the cacheable size acquired from the client side cacheable size holding unit 730, the client side deletion determination unit 728 selects one or A plurality of deletion target contents are detected, and the client side deletion determination unit 724 deletes the corresponding deletion target contents from the client side cache holding unit 723.

  The client-side deletion determination unit 728 corresponds to the processing flow shown in FIG. 10 in the form described in FIG. 8, and corresponds to the processing flow shown in FIG. 13 in the form described in FIG. In addition, the client-side deletion determination unit 708 recalculates the threshold T upon deletion.

  FIG. 23 and FIG. 24 show an application of the cache registration method based on the compression rate described in FIGS. 21 and 22 in addition to the cache deletion method by the server side device based on the compression rate described in FIGS. It is the example of a structure shown.

  Hereinafter, the server-side apparatus 800 performs both determination of content to be deleted from the cache and determination of content to be registered in the cache based on the compression ratio, and notifies the client-side apparatus 820 of the determination, thereby the server-side apparatus 800. And a method of synchronizing the cache of the client side device 820.

  A server-side device 800 in FIG. 23 newly has a server-side registration determination notification unit 813 in addition to the server-side device 600 in FIG. The client side device in FIG. 24 newly has a client side registration determination notification acquisition unit 828 in addition to the client side device 620 in FIG.

  When the server-side device 800 newly receives content from the upstream side, the registration determination based on the threshold value T described in the configuration example of FIGS. 20 and 21 is performed by the server-side registration determination unit 814, and it is determined that the content is registered in the cache. The server-side registration unit 805 stores the information in the server-side cache holding unit 803, the server-side registration determination unit 814 notifies the server-side registration determination notification unit 813 of the determination, and the server-side registration determination notification unit 813 When transmitting the difference data, an HTTP header as shown in FIG. 25 is added to the response, and the response is transmitted downstream via the communication unit 801.

  When the client side device 820 receives the response via the communication unit 821, whether or not the client side registration determination notification acquisition unit 828 caches the content decompressed by the differential decompression unit 822 is shown in FIG. If it is determined from the header and specified to be cached, the decompressed content is registered in the client-side cache holding unit 823 by the client-side registration determination unit 825.

  Note that the deletion of the cache is the same as the configuration example shown in FIGS.

  In the example shown in FIG. 25, the client-side device 820 is designated to cache the expanded content by “, save” added to the X-Identifier header. If the server-side registration determination unit 814 determines not to cache, the above-mentioned “, save” is deleted from the HTTP header to instruct the client-side device 820 not to cache the expanded content.

  In this way, by coordinating the server side device and the client side device even when registering the cache, the failure of the differential decompression process due to the mismatch of the cached contents can be avoided, and the server side device and the client side device can be more connected. The amount of data transferred can be reduced.

  In the example shown in FIG. 25, the content to be cached is notified from the server side device to the client side device using the HTTP response header including the difference data, but of course, it may be notified using another protocol or the like. .

Each of the above functions can be realized even if it is described as software and processed by a computer having an appropriate mechanism.
The present embodiment can also be implemented as a program for causing a computer to execute predetermined means, causing a computer to function as predetermined means, or causing a computer to realize predetermined functions. In addition, the present invention can be implemented as a computer-readable recording medium that records the program.

  Note that the present invention is not limited to the above-described embodiment as it is, and can be embodied by modifying the constituent elements without departing from the scope of the invention in the implementation stage. In addition, various inventions can be formed by appropriately combining a plurality of components disclosed in the embodiment. For example, some components may be deleted from all the components shown in the embodiment. Furthermore, constituent elements over different embodiments may be appropriately combined.

The figure which shows the structural example of the computer network system which concerns on one Embodiment of this invention. The figure which shows the other structural example of the computer network system which concerns on the same embodiment Diagram for explaining differential transfer for dynamic content The figure which shows an example of the HTTP header which a server side apparatus notifies the identifier of a content to a client side apparatus. The figure which shows an example of the HTTP response at the time of the server side apparatus transferring difference data to the client side apparatus The figure which shows the structural example of the server side apparatus which concerns on the embodiment The figure which shows the structural example of the client side apparatus which concerns on the same embodiment The figure for demonstrating an example of the cache deletion method in the embodiment The flowchart which shows an example of the registration process procedure to the histogram of FIG. The flowchart which shows an example of the deletion procedure from the histogram and cache of FIG. The figure for demonstrating the other example of the cache deletion method in the embodiment The flowchart which shows an example of the registration processing procedure to the bidirectional | two-way linked list of FIG. The flowchart which shows an example of the deletion processing procedure from the bidirectional | two-way linked list and cache of FIG. The figure which shows the other structural example of the server side apparatus which concerns on the same embodiment The figure which shows the other structural example of the client side apparatus which concerns on the same embodiment The figure which shows the further another structural example of the server side apparatus which concerns on the same embodiment. The figure which shows the further another structural example of the client side apparatus which concerns on the same embodiment. The figure for demonstrating an example of the notification method of the client side cacheable size in the structural example of FIG.16 and FIG.17. The figure which shows an example of the HTTP response of the difference data which the server side apparatus in the structural example of FIG.16 and FIG.17 transmits to a client side apparatus The flowchart which shows an example of the process sequence of the cache registration determination part by the compression rate which concerns on the embodiment The figure which shows the further another structural example of the server side apparatus which concerns on the same embodiment. The figure which shows the further another structural example of the client side apparatus which concerns on the same embodiment. The figure which shows the further another structural example of the server side apparatus which concerns on the same embodiment. The figure which shows the further another structural example of the client side apparatus which concerns on the same embodiment. The figure which shows an example of the HTTP response of the difference data which the server side apparatus in the structural example of FIG.23 and FIG.24 transmits to a client side apparatus

Explanation of symbols

  DESCRIPTION OF SYMBOLS 100 ... Server center LAN, 101, 111 ... Server, 102, 112 ... Server side proxy, 103, 113 ... WAN, 104 ... Client side office LAN, 105 ... Client side proxy, 106, 116 ... Client, 107, 118 ... Web browser, 110 ... LAN of business application center, 114 ... Access point, 115 ... Wireless network, 117 ... Client side module, 200,300,500,600,700,800 ... Server side device, 205 ... Server side device Cache, 207, 305, 520, 620, 720, 820 ... client side device, 212 ... client side device cache, 301, 501, 601, 701 ... server side device communication unit, 302, 502, 02, 702, 802 ... differential compression unit, 303, 505, 605, 705, 805 ... server side registration unit, 304, 503, 603, 703, 803 ... server side cache holding unit, 306, 521, 621, 721, 821 ... communication unit of client side device, 307,522,622,722,822 ... difference expansion unit, 308,525,625,725,825 ... client side registration unit, 309,523,623,723,823 ... client side cache Holding unit, 504, 604, 704, 804 ... Server side deletion unit, 506, 606, 706, 806 ... Server side compression rate calculation unit, 507, 607, 707, 807 ... Server side size / compression rate holding unit, 508, 608, 708, 808 ... Server-side deletion determination unit, 509, 609, 709, 8 9: Server side size acquisition unit, 510, 710 ... Server side cacheable size holding unit, 524, 624, 724, 824 ... Client side deletion unit, 526, 726 ... Client side compression rate calculation unit, 527, 727 ... Client side Size / compression rate holding unit, 528, 728... Client side deletion determining unit, 529, 729... Client side size obtaining unit, 530, 730... Client side cacheable size holding unit, 610, 810. , 811 ... Server-side deletion notification unit, 612, 812 ... Client-side cacheable size acquisition unit, 626, 826 ... Client-side deletion notification acquisition unit, 627, 827 ... Client-side cacheable size notification unit, 731 ... Client-side registration determination 801, Sir Bar-side device communication unit, 813 ... Server-side registration determination notification unit, 814 ... Server-side registration determination unit, 828 ... Client-side registration determination notification acquisition unit

Claims (21)

Holding means for holding at least data compressed in the past and transmitted to a specific client device side data transfer device as reference data in an uncompressed state;
Receiving means for receiving data destined for a server device as a transmission source and a client device as a destination;
Generating means for generating compressed data obtained by compressing the received data based on the reference data held in the holding means;
Transmitting means for transmitting the generated compressed data to the client device-side data transfer device;
For each of the reference data held in the holding means, the first information indicating the data size of the data, and the data size of the data and the data size of the compressed data generated from the data Storage means for storing second information indicating the determined degree of compression;
The relationship between the sum of the data sizes indicated by the first information relating to each of all reference data held in the holding means and the upper limit value of the data size that can be held by the holding means is a certain condition. A selection unit that selects reference data to be deleted from the holding unit, based on the second information relating to each reference data stored in the holding unit,
A server apparatus-side data transfer apparatus comprising: deletion means for deleting reference data selected by the selection means from the holding means. Holding means for holding at least compressed data received from a specific server device side data transfer device in the past as reference data in an uncompressed state;
Receiving means for receiving, from the server apparatus side data transfer apparatus, data having a certain server apparatus as a transmission source and a certain client apparatus as a destination;
A determination means for determining whether the received data is uncompressed data or compressed compressed data based on specific information added to the received data;
Generating means for decompressing the compressed data and generating uncompressed data based on the reference data held in the holding means when the received data is compressed data;
Transmitting means for transmitting the generated data to the certain client device;
For each of the reference data held in the holding means, the first information indicating the data size of the data, and the data size of the data and the data of the compressed data from which the data is generated Storage means for storing second information indicating the degree of compression determined by the size;
The relationship between the sum of the data sizes indicated by the first information relating to each of all reference data held in the holding means and the upper limit value of the data size that can be held by the holding means is a certain condition. A selection unit that selects reference data to be deleted from the holding unit, based on the second information relating to each reference data stored in the holding unit,
A client apparatus-side data transfer apparatus comprising: deletion means for deleting reference data selected by the selection means from the holding means.   The selecting means is configured such that a sum of the first information related to all reference data held in the holding means after the selected reference data is deleted from the holding means by the deleting means is below a reference value. The data transfer apparatus according to claim 1, wherein the selection is performed until it becomes. Prior to holding certain data as reference data in the holding means, the selecting means is data indicated by the first information relating to each of the certain data and reference data held in the holding means. The selection is performed when the sum of the sizes exceeds the upper limit of the data size that can be held by the holding unit;
4. The holding unit holds the certain data as reference data after the reference data selected by the selection unit is deleted from the holding unit by the deletion unit. The data transfer device described in 1.   5. The data transfer according to claim 1, wherein the selection unit preferentially selects reference data indicated by the second information that the degree of compression is higher. 6. apparatus.   2. The data transfer apparatus according to claim 1, further comprising notification means for notifying the client apparatus-side data transfer apparatus of information specifying the reference data selected by the selection means.   3. A determination unit for determining whether or not to hold certain data as reference data in the holding unit based on the second information relating to the certain data. The data transfer device described in 1. Determining means for determining whether or not to hold certain data to be transmitted to the client apparatus side data transfer apparatus as reference data in the retaining means based on the second information relating to the certain data;
2. The data transfer apparatus according to claim 1, further comprising notification means for notifying the client apparatus-side data transfer apparatus of a determination result by the determination means relating to the certain data.   9. The data transfer apparatus according to claim 7, wherein the holding unit holds the certain data as reference data only when it is determined to be held by the determination unit.   10. The determination unit according to claim 7, wherein when the degree of compression indicated by the second information relating to the certain data is equal to or less than a reference value, the determination unit determines to hold the data in the holding unit. The data transfer device according to any one of the above. In a communication system including a server device side data transfer device and a client device side data transfer device,
The server device side data transfer device is:
First holding means for holding at least data compressed in the past and transmitted to a specific client device side data transfer device as reference data in an uncompressed state;
First receiving means for receiving data destined for a server device as a transmission source and a client device as a destination;
First generation means for generating compressed data obtained by compressing the received data based on the reference data held in the first holding means;
First transmission means for transmitting the generated compressed data to the client device-side data transfer device;
For each of the reference data held in the first holding means, the first information indicating the data size of the data, and the data size of the data and the data of the compressed data generated from the data Storage means for storing second information indicating the degree of compression determined by the size;
A sum of data sizes indicated by the first information relating to each of all reference data held in the first holding means, and an upper limit value of the data size that can be held by the first holding means When the relationship satisfies a certain condition, the relationship should be deleted from the first and second holding means based on the second information relating to each reference data stored in the first holding means A selection means for selecting reference data;
First deletion means for deleting the reference data selected by the selection means from the first holding means;
Specific information notifying means for notifying the client device-side data transfer device of information for specifying the reference data selected by the selecting means;
The client device side data transfer device is:
A second holding means for holding at least the compressed data received from the server device side data transfer device in the past as reference data in an uncompressed state;
Second receiving means for receiving, from the server apparatus side data transfer apparatus, data destined for a certain server apparatus as a transmission source and a certain client apparatus as a destination;
A determination means for determining whether the received data is uncompressed data or compressed compressed data based on specific information added to the received data;
When the received data is compressed data, a second data is generated by decompressing the compressed data and generating uncompressed data based on the reference data held in the first holding means Generating means;
Second transmission means for transmitting the generated data to the certain client device;
Specific information notification acquisition means for acquiring a notification of information for specifying the reference data selected by the selection means from the server device-side data transfer device;
A communication system, comprising: a second deletion unit that deletes reference data specified by the acquired information from the second holding unit.   The selection means includes the first reference data related to all reference data held in the first holding means after the selected reference data is deleted from the first holding means by the first deletion means. The communication system according to claim 11, wherein the selection is performed until a sum of information becomes a reference value or less. The selection unit may store the first data related to each of the certain data and the reference data held in the first holding unit before holding the data as reference data in the first holding unit. When the sum of the data sizes indicated by the information exceeds the upper limit value of the data size that can be held by the first holding means, the selection is performed.
The first holding means and the second holding means, respectively, after the reference data selected by the selecting means are deleted from the first and second holding means by the first and second deleting means. 13. The communication system according to claim 12, wherein the data is stored as reference data.   14. The communication system according to claim 11, wherein the selection unit preferentially selects reference data indicated by the second information that the degree of compression is higher. . The first data transfer device includes:
It is determined based on the second information related to the certain data whether or not certain data to be transmitted to the second data transfer device is retained as reference data in the first and second retaining means. A determination means;
A determination result notifying unit for notifying the second data transfer device of a determination result by the determination unit relating to the certain data;
The second data transfer device includes:
12. The communication system according to claim 11, further comprising determination result notification acquisition means for acquiring notification of a determination result by the determination means relating to the certain data from the first data transfer device.   16. The data transfer apparatus according to claim 15, wherein the first and second holding units hold the certain data as reference data only when it is determined to be held by the determination unit.   The determination means determines that the first and second holding means hold when the degree of compression indicated by the second information relating to the certain data is equal to or less than a reference value. The data transfer device according to claim 15 or 16. A holding step for holding at least the data compressed in the past and transmitted to a specific client device side data transfer device in the holding means as reference data in an uncompressed state;
A receiving step of receiving data destined for a certain server device as a transmission source and a certain client device as a destination;
A generating step of generating compressed data obtained by compressing the received data based on the reference data held in the holding unit;
A transmission step of transmitting the generated compressed data to the client device-side data transfer device;
For each of the reference data held in the holding means, the first information indicating the data size of the data, and the data size of the data and the data size of the compressed data generated from the data A storage step of storing in the storage means second information indicating the degree of compression determined;
The relationship between the sum of the data sizes indicated by the first information relating to each of all reference data held in the holding means and the upper limit value of the data size that can be held by the holding means is a certain condition. A selection step of selecting reference data to be deleted from the holding means based on the second information relating to each reference data stored in the holding means when satisfying
And a deletion step of deleting the reference data selected in the selection step from the holding means. A holding step for holding the compressed data received from at least a specific server device side data transfer device in the past in the holding means as reference data in an uncompressed state;
A receiving step of receiving, from the server apparatus side data transfer apparatus, data destined for a certain server apparatus as a transmission source and a certain client apparatus as a destination;
A determination step of determining whether the received data is uncompressed data or compressed compressed data based on specific information added to the received data;
When the received data is compressed data, based on the reference data held in the holding unit, a generation step of generating the uncompressed data by decompressing the compressed data;
Transmitting the generated data to the certain client device;
For each of the reference data held in the holding means, first information indicating the data size of the data, and the data size of the data and the data of the compressed data from which the data is generated A storage step of storing in the storage means second information indicating the degree of compression determined by the size;
The relationship between the sum of the data sizes indicated by the first information relating to each of all reference data held in the holding means and the upper limit value of the data size that can be held by the holding means is a certain condition. A selection step of selecting reference data to be deleted from the holding means based on the second information relating to each reference data stored in the holding means when satisfying
And a deletion step of deleting the reference data selected in the selection step from the holding means. In a program for causing a computer to function as a server device side data transfer device,
The program is
A holding step for holding at least the data compressed in the past and transmitted to a specific client device side data transfer device in the holding means as reference data in an uncompressed state;
A receiving step of receiving data destined for a certain server device as a transmission source and a certain client device as a destination;
A generating step of generating compressed data obtained by compressing the received data based on the reference data held in the holding unit;
A transmission step of transmitting the generated compressed data to the client device-side data transfer device;
For each of the reference data held in the holding means, the first information indicating the data size of the data, and the data size of the data and the data size of the compressed data generated from the data A storage step of storing in the storage means second information indicating the degree of compression determined;
The relationship between the sum of the data sizes indicated by the first information relating to each of all reference data held in the holding means and the upper limit value of the data size that can be held by the holding means is a certain condition. A selection step of selecting reference data to be deleted from the holding means based on the second information relating to each reference data stored in the holding means when satisfying
A program causing a computer to execute a deletion step of deleting reference data selected in the selection step from the holding means. In a program for causing a computer to function as a data transfer device on the client device side,
The program is
A holding step for holding the compressed data received from at least a specific server device side data transfer device in the past in the holding means as reference data in an uncompressed state;
A receiving step of receiving, from the server apparatus side data transfer apparatus, data destined for a certain server apparatus as a transmission source and a certain client apparatus as a destination;
A determination step of determining whether the received data is uncompressed data or compressed compressed data based on specific information added to the received data;
When the received data is compressed data, based on the reference data held in the holding unit, a generation step of generating the uncompressed data by decompressing the compressed data;
Transmitting the generated data to the certain client device;
For each of the reference data held in the holding means, first information indicating the data size of the data, and the data size of the data and the data of the compressed data from which the data is generated A storage step of storing in the storage means second information indicating the degree of compression determined by the size;
The relationship between the sum of the data sizes indicated by the first information relating to each of all reference data held in the holding means and the upper limit value of the data size that can be held by the holding means is a certain condition. A selection step of selecting reference data to be deleted from the holding means based on the second information relating to each reference data stored in the holding means when satisfying
A program causing a computer to execute a deletion step of deleting reference data selected in the selection step from the holding means.

Images