JP2000207270A - Www proxy device - Google Patents

Abstract

PROBLEM TO BE SOLVED: To enable an efficient cache processing and to reduce the load on a network by performing the cache process which is fast, efficient, and reliable and prereading WWW data that a user requests with high reliability. SOLUTION: When a WWW browser 101 sends a request to the WWW proxy device 102, a cache decision means 103 receives the request. The cache decision means 103 passes the URL of requested WWW data to a totalizing and analyzing means 106, which totalizes it as an access frequency. The cache decision means 103 inquires whether the requested WWW data are cached of a cache memory means 104 and a cache disk means 105. When the data are cached, the cache decision means 103 sends the result to the WWW browser 101. Furthermore, the requested WWW data are pre-read.

Description

DETAILED DESCRIPTION OF THE INVENTION

[0001]

[0001] The present invention relates to a WWW (World Wi-Fi).
The present invention relates to a WWW proxy device that performs acquisition of WWW data of a target server on behalf of a user's request in response to a request from a user, and caches the acquired WWW data.

[0002]

2. Description of the Related Art A WWW network system using a WWW proxy device is generally configured as shown in FIG.
The WWW proxy 1402, which has received the request from the WWW client 1401 on the user side, obtains the requested WWW data from the target WWW 1405 server as a proxy, stores it in the cache device 1403, and thereafter stores it in the cache device 1403. When certain WWW data is requested by the client 1401 on the user side, the WWW data in the cache device 1403 is returned to the WWW client on the user side.

At this time, a disk device is usually used as a cache device, but "Norifumi Nishikawa, Takafumi
Hosokawa, Yasuhide Mori, Kenichi Yoshida, Hiroshi
Tsuji, Memory-based architecture for distributed W
WW caching proxy, ComputerNetworks And ISDN System
s (30) 1-7 (1998) pp. 205-214 ", there is an attempt to speed up writing and reading of WWW data using a memory device as a cache device.

[0004] Also, WWW requested by a user.
As for a look-ahead method of acquiring the WWW data of the data reference destination before the user makes a request, all the referenced WWW data is acquired and stored in the cache. When the user makes a request for the data of the reference destination, the WWW data stored in the cache is immediately returned, so that the time required for the reference on the user side is greatly reduced.

[0005]

SUMMARY OF THE INVENTION The above conventional WW
In a case where only a disk device is used as a cache device in a W proxy device, it takes time to perform cache processing and transmit WWW data to a user because the speed of writing and reading WWW data to and from the cache device is low. There's a problem.

In the case where only a memory device is used as a cache device, the speed of writing to and reading from the cache device increases, but the memory itself is more expensive than a disk, so that it is used as a large-capacity cache device. WWW in cache at the time of failure
There is a problem that data is lost.

Further, regarding the prefetching method, prefetching of all the WWW data referred to by the requested WWW data or prefetching of the WWW data further referred to by all the WWW data referred to by the requested WWW data. Therefore, there is a problem that a large amount of WWW data that is unlikely to be requested by the user continuously is read ahead, resulting in an increase in the load on the network.

[0008]

In order to solve the above-mentioned problems of the prior art, a WWW proxy device according to the present invention comprises:
A cache memory unit and a cache disk unit for temporarily storing WWW data, a totaling / analyzing unit for totalizing / analyzing log data of transfer of WWW data on a network, and an instruction from a user are input. Determining whether or not the WWW data is to be stored by the cache memory means or the cache disk means based on the contents of the input means and the contents of the input means and the tallying / analyzing means; A cache determining unit for performing writing and reading by the memory unit or the cache disk unit, or acquiring the WWW data from the upper WWW proxy device or the WWW server device; a database for controlling and managing the WWW proxy device; Control management for controlling and managing the WWW proxy device by using And the stage, was requested WW
A read-ahead means for pre-reading WWW data as a reference destination of the W data.

That is, according to the present invention, WWW data is
In a WWW proxy device that obtains a proxy from a WWW server device in response to a request from a W browser, the WWW proxy device includes a cache memory unit and a cache disk unit for temporarily storing WWW data, and WWW data on a network. Totalizing / analyzing means for totalizing / analyzing transfer log data of the user, inputting means for inputting an instruction from a user, and storing the WWW data on the basis of the contents of the inputting means and the totaling / analyzing means in the cache memory Means for determining whether to store the WWW data by the cache memory means or the cache disk means, and writing and reading of the WWW data by the cache memory means or the cache disk means, or the WWW data from the upper WWW proxy device or WWW server device. With cache determining means for acquiring the A.

Further, the cache judging means caches the WWW data to be normally stored in the cache memory means, and satisfies a certain condition with respect to WWW data overflowed by replacement of cache data in the cache memory means. Only the WWW data is cached in the caching disk means.

Further, according to the present invention, the WWW proxy device caches only the WWW data having the highest access frequency of only the log data for the latest fixed period out of the log data totalized by the totaling / analyzing means. Is performed.

Further, according to the present invention, the above-mentioned fixed condition for judging whether or not to cache the WWW data overflowed due to the replacement of the cache data of the cache memory means in the cache disk means,
The access frequency of the WWW data tabulated by the tabulation / analysis means for tabulating / analyzing the log data of the data transfer is higher.

Further, according to the present invention, the predetermined condition for determining whether or not to cache the WWW data overflowed by the replacement of the cache data of the cache memory means in the cache disk means includes inputting an instruction from a user. The input contents of the input means are input.

Further, according to the present invention, the WWW proxy device includes a database for controlling and managing the WWW proxy device, and the WWW proxy device includes a control management means for controlling and managing the WWW proxy device using the database. It is provided internally or externally, and the control management means controls and manages the cache determination means by giving an instruction to stop the processing of the cache determination means as necessary.

Further, according to the present invention, the WWW proxy device further comprises distributed control means for distributing WWW data requested to be obtained by the WWW browser for proxy to a higher-level WWW proxy device for proxy processing, and the distributed control means. Instructs the cache determining means to request the upper-level WWW proxy device to perform proxy processing, thereby controlling the cache determining means.

Further, according to the present invention, the WWW proxy device and the higher-level WWW proxy device include a database for controlling and managing the WWW proxy device, and the WWW proxy device and the higher-level WWW are used by using the database.
A control management unit for controlling and managing the proxy device is provided inside or outside the WWW proxy device, and the control management unit includes:
An instruction to stop the processing of the cache determination unit of the WWW proxy device and the higher-level WWW proxy device, or a dynamic change of the database of the WWW proxy device and the higher-level WWW proxy device as necessary. The cache determining means of the WWW proxy device and the host WWW proxy device is controlled and managed.

Further, according to the present invention, the distribution control means for further distributing WWW data requested to be obtained by the WWW browser to perform proxy processing to a higher-level WWW proxy device and performing proxy processing includes hashing the domain name portion of the URL of the WWW data. Numerical conversion means for converting to a numerical value using a function, and distributed processing instructing means for requesting a higher-level WWW proxy device corresponding to the numerical value to process the WWW data, wherein the distributed processing instructing means performs proxy processing. The cache determining means is controlled by instructing the cache determining means of a request to a higher-level WWW proxy device.

Further, according to the present invention, the WWW proxy device or the host WWW proxy device
A prefetch means for prefetching WWW data as a reference destination of the WW data, wherein the prefetch means controls the cache determination means by instructing the cache determination means to request; .

Further, the present invention provides a method for requesting a WW
The prefetching means for prefetching WWW data as a reference destination of W data includes a prefetching control means for prefetching only the WWW data satisfying a certain condition, and the prefetching control means comprises:
The read-ahead control of the cache determining means is performed by giving a request instruction to the cache determining means.

Further, the present invention provides a method for requesting a WW
The look-ahead control means for controlling the look-ahead of the WWW data referred to by the W data has the highest access frequency counted by the tally / analysis means for counting and analyzing the log data of the transfer of the WWW data on the network. An instruction to request only the WWW data is given to the cache determining means, thereby controlling the prefetching of the cache determining means.

Further, the present invention provides a method for requesting a WW
The prefetching means for prefetching WWW data as a reference destination of the W data includes importance determining means for determining the importance of the user who has requested the WWW data, and the importance determining means describes the importance of each user. And determining the importance of the user using the importance database. The look-ahead control unit requests the cache determination unit using the determination result by the importance determination unit. The prefetch control of the cache determination means is performed by giving the instruction of (1).

Further, according to the present invention, the prefetch control means includes:
The read-ahead control of the cache determining unit is performed by giving a request instruction to the cache determining unit using the determination result by the importance determining unit and the access frequency calculated by the totaling / analyzing unit. Is what you do.

Normally, the WWW data to be stored is cached in the cache memory means by the cache determination means, and the contents of the totaling / analyzing means or the input means are provided for the WWW data overflowed by the replacement of the cache data in the cache memory means. WWW corresponding to the input contents by
By caching only data in the caching disk means, there is an effect that high-speed cache processing with a high cache hit rate can be performed.

Further, the prefetching means for prefetching the WWW data referred to by the requested WWW data gives an instruction of prefetching only the WWW data corresponding to the contents of the counting / analyzing means, so that the cache having a high cache hit rate is obtained. There is an effect that the processing and the load on the network can be reduced.

[0025]

DETAILED DESCRIPTION OF THE PREFERRED EMBODIMENTS An embodiment of the present invention will be specifically described below with reference to the drawings.

First, an example of the physical configuration of the WWW proxy device of the present invention will be described with reference to FIG.

In the physical configuration diagram of FIG.
The WW proxy device 1501 includes a processor 1502 that executes all processing programs, and a memory device 150 that temporarily stores data used for executing the processing programs.
3, a disk device 1504 for storing programs, data, and the like, an input device 1505 for taking input from a user into the device, a communication control device 1506 for transmitting and receiving WWW data and control signals to and from the outside, and a W to be cached.
Cache memory device 15 for storing WW data
07, a cache disk device 1508 for storing WWW data to be cached, and a management control database that records the operation status and load status of the WWW proxy device, and the number and address of upper WWW proxy devices. And a control management device 1509 that constantly monitors the WWW proxy device and the upper WWW proxy device,
Devices other than the control management device 1509 are connected to the internal communication bus 15.
10 are connected to each other inside the WWW proxy device 1501.

The communication controller 1506 is connected to an external network such as the Internet via an external communication bus 1511. The control management device 1509 is connected to the communication control device 1506 via the external communication bus 1511.

Next, the principle of the WWW proxy device according to the present invention will be described with reference to FIGS. 1, 2, 3, 4 and 15.
This will be described with reference to FIG.

FIG. 1 is a diagram showing the principle of a WWW proxy device according to the present invention, and FIG. 2 is a diagram showing a simple principle of distributed processing performed by using a WWW proxy device according to the present invention and a plurality of WWW proxy devices according to the present invention. It is.

The WWW browser 101 is a WWW client that acquires and displays WWW data such as hypertext and image data on the WWW.
11 is connected to the communication control device 1506.

The WWW proxy device 102 generally has a WW
A device that makes a request to the Internet or the like 100 on behalf of the W browser 101.

The cache determining unit 103 receives a request from the WWW browser 101, determines whether the requested WWW data is cached by the cache memory unit 104 or the cache disk unit 105, and determines whether the requested WWW data is cached. If it is, the WWW data is returned to the WWW browser 101; if it is not cached, the WWW data is obtained from a WWW server on the Internet or the like 100;
It is determined whether the data is cached by the cache memory means or the WWW data overflowed due to the replacement of the cache data in the cache memory means by the cache disk means. This is a means for instructing the disk means 105 to cache, has a processing program in the disk device 1504, and the processor 1502 performs actual processing.

The cache memory means 104 is means for storing WWW data in the cache memory or transferring WWW data stored in the cache memory in accordance with an instruction from the cache determination means 103. A processing program in the device 1504,
The processor 1502 performs an actual process, and stores the WWW data received via the communication control device 1506 in the memory device 1
503, and once transferring the WWW data overflowed by the replacement of the cache data by the LRU to the memory device 1503, stores the received WWW data in the memory device 1503 in the cache memory device 1507, and
The WWW data in the cache memory device 1507 is transferred to the communication control device 1506.

The cache disk unit 105 sends the WW overflowed by the cache data replacement of the cache memory unit 104 according to the instruction of the cache determination unit 103.
This is a means for storing W data in the cache disk or for passing WWW data stored in the cache disk to the cache determination means 103. The processing program is stored in the disk device 1504, and the processor 1502 is Of the communication control device 1506
Is transferred to the memory device 1503, and finally stored in the cache disk device 1508, and the WWW data in the cache disk device 1508 is transferred to the communication control device 1506.

The tallying / analyzing means 106 is means for receiving log data of WWW data requested from the WWW browser 101 from the cache judging means 103, tallying / analyzing the log data, and holding it as an access frequency. The disk device 1504 has a processing program and access frequency data, and the processor 1502 performs actual processing.

Further, the tallying / analyzing means 106 also tallies up log data only for a certain period of time in recent times and holds the log data as a recent access frequency.
If the access frequency tabulated by the data is used and the latest access frequency is set in advance as a cache processing condition, the latest access frequency is also used.

The input means 107 is provided with a WW to be cached.
A means for holding a URL (Uniform Resource Locator), a domain name, and the like of W data in accordance with an instruction from a user, and instructing the cache determination means 103 of the content. The processing is performed by the input device 1505, and the input data is processed. It is stored in the disk device 1504.

The distribution control means 108 is the WWW browser 1
01 is a means for instructing the cache determination means 103 to request the processing by further distributing the request from the upper WWW proxy apparatus 200 to a plurality of upper WWW proxy apparatuses 200. The processing program is provided in the disk apparatus 1504, and the processor 1502 Perform the actual processing.

The database 109 stores the load and operation status of the WWW proxy device 101, the upper WWW proxy device 2
This is a management information base in which the number and address of 00 are recorded. The management information base is held in the disk device 1504, and is read and written by the control management device 1509 via the communication control device 1506.

The control management means 110 is provided in the database 10
9, the operation status of the WWW proxy device 101 is constantly monitored, and the processing is stopped or the processing request is stopped for the cache determining unit 103 of the WWW proxy device 102 having a high load or a failure, and the upper-level WWW proxy. This is a means for instructing a change of the device 200, and is in the control management device 1509.

The pre-reading means 111 is provided with the WW acquired by the cache determining means 103 from the Internet 100 or the like.
Receives W data, reads the URL referred to when the WWW data is hypertext, and reads the URL
Is a means for instructing the cache determination means 103 to acquire WWW data indicated by "."
504 has a processing program, temporarily stores the acquired WWW data in the memory device 1503,
02 reads a URL from WWW data and performs a WWW data acquisition instruction process.

Normally, the WWW data to be stored is cached in the cache memory means, and only the WWW data which satisfies certain conditions is cached in the cache disk means with respect to the WWW data overflowed by the replacement of the cache data in the cache memory means. Accordingly, it is possible to realize high-speed cache processing and improve the cache hit rate.

Also, by caching only the WWW data having a higher access frequency of the log data during the last certain period, even if the tendency of the WWW data that the user wants to acquire changes, the change is cached. This can be reflected in the WWW data, and as a result, the cache hit rate can be improved and the load on the network can be reduced.

For the WWW data overflowed due to the replacement of the cache data in the cache memory means, only the WWW data having the highest access frequency is cached in the cache disk means, so that the WWW data which is likely to be requested is held. As a result, the hit rate of the cache can be improved and the load on the network can be reduced.

Further, with respect to the WWW data overflowing due to the replacement of the cache data in the cache memory means, only the WWW data which matches the input contents of the input means for inputting the instruction from the user is cached in the cache disk means, so Can hold WWW data to be acquired, and as a result, it is possible to improve the hit rate of the cache and reduce the load on the network.

Further, the control management means constantly monitors the operation status and the load status of the WWW proxy device, and issues an instruction to appropriately stop the processing when the load of the WWW proxy device is high, so that the failure of the WWW proxy device can be prevented beforehand. Can be prevented.

Further, by pre-reading the WWW data of the reference destination of the requested WWW data, it is possible to return the corresponding WWW data at a high speed when the user subsequently makes a request of the reference destination. Become.

Further, a plurality of upper W
By further requesting the processing to the WWW proxy device, it is possible to distribute the load of the WWW proxy device, improve the cache hit rate, and reduce the load on the network.

Further, the control and management means constantly monitors the operation status and the load status of the WWW proxy device and the upper WWW proxy device, and not only appropriately stops the processing when the load of the WWW proxy device and the upper WWW proxy device is high. By dynamically changing a database for control management and instructing a higher-load WWW proxy apparatus not to request processing in advance, more efficient distributed processing can be performed.

FIG. 3 is a diagram showing the principle of the distributed control means of the WWW proxy device according to the present invention. The numerical value conversion unit 301 receives the request from the WWW browser 101.
A means for receiving a URL from the cache determining means 103, converting the domain name part of the URL into a numerical value by a hash function, and passing the numerical value to the distributed processing instructing means 302, having a processing program in the disk device 1504, Performs the actual processing.

The distributed processing instructing means 302 selects a WWW proxy device corresponding to the numerical value received from the numerical value converting means 301, and instructs the cache judging means 103 to perform proxy processing on the selected WWW proxy device. The processor 1502 has a processing program in the disk device 1504, and the processor 1502 performs actual processing.

By controlling and managing the distributed processing by the distributed processing instructing means, the speed of the distributed processing request can be increased and the UR can be improved.
L enables load balancing of the upper WWW proxy device.

FIG. 4 is a diagram showing the principle of the prefetch means according to the present invention. The prefetch control unit 401 reads the WWW data acquired by the cache determination unit 103,
It is determined whether or not the text is a hypertext. If the text is a hypertext, the URL of the WWW data of the reference destination is analyzed. This is a means for giving an instruction to the cache determination means 103, and has a processing program in the disk device 1504, and the processor 1502 performs actual processing.

The importance database 402 is a database that describes the importance of a user who makes a request by the WWW browser 101 and is held in the disk device 1504.

FIG. 12 shows the importance database 40.
FIG. 3 is a diagram showing an example of the second example, and the importance database 402 is
As shown in FIG.
The height of 02 is indicated. As the user information 1201, user information in a request header, user information at the time of proxy authentication, and the like are used. Next, the importance determination means 403
The importance of the user making the request with the WWW browser 101 is determined using the importance database 402,
This is a means for passing the determination result to the prefetch control means 401, and has a processing program in the disk device 1504,
The processor 1502 performs the actual processing.

The look-ahead control means considers the importance and access frequency of the requesting user, and among the WWW data referenced by the requested WWW data of the user determined to be important, the WWW with the highest access frequency. By prefetching only data, WWW is highly likely to be provided and requested according to the importance of the user.
Network load can be reduced by prefetching only data.

Hereinafter, the operation of the WWW proxy device according to the present invention will be described for a case where the WWW proxy device is used alone as shown in FIG. 1 and a case where a plurality of WWW proxy devices are used as shown in FIG.

Thereafter, in all cases, the control management means 110 constantly monitors the load and operation status of the WWW proxy device 102, and if it is determined that the load is increasing, the corresponding means is stopped. , Or change the database 109 so as not to make a processing request to the higher-level WWW proxy device 200 having the corresponding means, and return an error message to the WWW browser 101 as necessary.

When the WWW proxy device is used alone as shown in FIG. 1, the processing flow is as shown in FIG.

First, when the WWW browser 101 makes a request to the WWW proxy device 102, the cache determining means 103 receives the request (step S).
501). The cache determining unit 103 passes the URL of the requested WWW data to the tallying / analyzing unit 106, and makes the URL as the access frequency (step S50).
2). Next, the cache determination unit 103 inquires of the cache memory unit 104 and the cache disk unit 105 whether the requested WWW data is cached by the cache memory unit 104 and the cache disk unit 105 (Step S).
503).

If it is determined in step S503 that the data is cached,
Transmits the corresponding WWW data to the WWW browser 101 (step S506). Further, a pre-read process is performed on the requested WWW data (step S
507), and the processing ends. The prefetching process in step S507 will be described later with reference to FIG.

If it is determined in step S503 that the cache has not been cached,
3 acquires the requested WWW data from the Internet 101 or the like (step S504). Next, the cache determination unit 103 performs a cache process on the obtained WWW data (step S505). Step S
The cache processing at 505 will be described later with reference to FIG.

Further, the cache determining means 103 transmits the obtained WWW data to the WWW browser 101 (step S506), performs a prefetching process on the obtained WWW data (step S507), and ends.

The cache processing procedure (step S505) in FIG. 5 will be described with reference to FIG.

First, the cache judging means 103 inquires of the tallying / analyzing means 106 whether or not to use the access frequency of the latest certain period as a condition for the cache processing (step S601). When used as a processing condition, the acquired W
An inquiry is made to the tallying / analyzing means 106 as to whether or not the URL of the WW data is a URL having a higher access frequency in a recent certain period (step S602). If the URL of the acquired WWW data is not the URL with the highest access frequency recently, the WW
The process ends without performing the W data caching process.

FIG. 11 shows a method for counting access frequencies when the last certain period is seven days. In the example of FIG. 11, the log data 1101 to 1107 for one day from Sunday to Saturday is updated every day (step S11).
08), the log data 1101 to 1107 of each day of the week are totaled (step S1109), and the access frequency of 7 days is 1
110.

If the URL of the obtained WWW data is higher in the access frequency for the latest certain period, and if the latest access frequency is not used as a condition for the cache processing in step S601, the cache memory unit 104 stores the LRU The oldest WWW data since the last reference based on the cache data replacement algorithm is transferred to the cache determining means 103 and simultaneously deleted from the cache memory, and the cache determining means 103 stores the acquired WWW data in the cache memory means 104. Handover and cache in the memory device (step S60).
3).

Next, the cache judging unit 103 determines whether the certain condition for caching the WWW data overflowed by the replacement of the cache data in the cache memory unit 104 in the cache disk unit is the access frequency by the tallying / analyzing unit 106. Or each unit is inquired as to whether or not the content is the input content of the input unit 107 (step S604).

If the certain condition is based on the access frequency of the tallying / analyzing means 106, the cache judging means 103 determines that the URL of the WWW data overflowed by the replacement of the cache data in the cache memory means 104 has a higher access frequency. Tallying / analyzing means 1
06 (step S605), if the URL has a higher access frequency, the WWW data overflowing due to the replacement of the cache data in the cache memory unit 104 is passed to the cache disk unit 105 and cached in the disk device. If the URL is not the one with the highest access frequency, the process ends.

When the above-mentioned certain condition is the input content of the input means 107,
Asks the input means 107 whether the URL of the WWW data overflowed by the replacement of the cache data in the cache memory means 104 corresponds to the input content (step S606). If the URL corresponds to the input content, the cache memory means The WWW data overflowing due to the replacement of the cache data in 104 is passed to the cache disk means 105 and cached in the disk device. If it does not correspond to the input contents, the processing ends.

The pre-reading processing procedure in FIG. 5 (step S
507) will be described with reference to FIG.

First, the prefetch control means 401 reads the WWW data acquired by the cache determination means 103,
The information of the user who made the request is determined by the importance determination means 40.
Pass to 3. The importance determining unit 403 checks the received user information in the importance database 402 (Step S).
701), the importance of the user is returned to the prefetch control means. If the importance of the user is low, the prefetch control unit 401 ends the processing without performing the prefetching, and if the importance of the user is high, determines whether the acquired WWW data is a hypertext (step S702).

If the acquired WWW data is not a hypertext, the process is terminated. If the acquired WWW data is a hypertext, the URL of the WWW data referred to by the WWW data.
Are sequentially read (step S703). Next, the prefetch control unit 401 determines that the read URL is
Inquiry is made to the tallying / analyzing means 106 as to whether the URL is the one with the highest access frequency (step S704). If it is determined that the URL is not the one with the highest access frequency, the process ends.

If it is determined in step S 704 that the URL has the highest access frequency, the prefetch control unit 401 determines whether the WWW data to be acquired is the cache memory unit 104 or the cache disk unit 10.
5 is inquired via the cache determination means 103 as to whether or not it is cached (step S503).

In step S503, W to be acquired
If the WW data is not cached, the prefetch control unit 401 instructs the cache determination unit 103 to acquire the WWW data of the URL (step S70).
5), the cache determination unit 103 acquires WWW data from a server such as the Internet 100 (step S706), and caches the acquired WWW data (step S505).

In step S503, W to be acquired
When the WWW data is cached, and when the cache processing of the obtained WWW data is completed in step S505, the prefetch control unit 401 sequentially checks the URL of the reference destination of the obtained WWW data and performs prefetch. Of WWW data that has not been determined
It is checked whether there is a URL (step S707).

In step S707, all the reference W
If it is determined whether the WW data is to be prefetched, the prefetching process ends. If there is reference WWW data that has not been determined whether to perform prefetching, step S703 is performed.
And the subsequent processing is repeated.

FIG. 8 shows a processing procedure when a plurality of WWW proxy devices 102 are used hierarchically as shown in FIG. 2 or FIG.

When a request is made to the WWW proxy device 102 by the WWW browser 101, the request is received by the cache determining means 103 (step S501). The cache determination unit 103 counts and analyzes the URL of the requested WWW data
To be counted as an access frequency (step S5).
02). Next, the cache determination unit 103 stores the requested WWW data in the cache memory unit 104.
Then, an inquiry is made as to whether the data is cached by the cache disk means 105 (step S503).

If it is determined in step S503 that the data is cached, the corresponding WWW data is
The data is transmitted to the WW browser 101 (step S506).
Further, a pre-read process is performed on the requested WWW data (step S507), and the process ends.

If it is determined in step S503 that the cache has not been cached,
3 is for obtaining requested WWW data in the upper WW
Request one of the W proxy devices 200 (step S8
01). Step S801 will be described later.

Next, the cache determining means 103
Upper WWW proxy device 30 that has requested acquisition of W data
The WWW data is acquired from the third cache determination unit 103 (step S802), and the acquired WWW data is cached (step S505). Hereinafter, FIG.
In the same manner as described above, the acquired WWW data is
1 (step S506), perform prefetching processing (step S507), and end.

Next, step S801 in FIG. 8 will be described in detail with reference to FIG. First, the cache determination unit 103 passes the URL of the requested WWW data to the numerical value conversion unit 301 in the distribution control unit 108 (step S901). The numerical value conversion means 301 receives the received URL
A hash function is applied to the domain name part of (a) to convert it into a numerical value (step S902).

Here, there are several methods for converting into a numerical value. As an example, here, it is assumed that the value of the character code of the URL is simply added. The numerical value conversion means 301 passes the converted numerical value to the distributed processing instruction means 302 (Step S903). The distributed processing instruction unit 302 selects one of the higher-level WWW proxy devices 200 corresponding to the received numerical value (step S904). Here, several methods of selecting the upper WWW proxy device 200 are conceivable. For example, here, the received numerical value is divided by the number of upper WWW proxy devices 200, and the upper WWW proxy device is made to correspond to the remaining value. .

For example, as shown in FIG. 13, 0, 1, 2, and 3 are assigned to upper WWW proxy apparatuses 1301 to 1304, respectively.
And the numeral are assigned, and the numerical value conversion unit 301 receives the URL from the cache determination unit 103 (step S13).
05), if the received URL is converted into the value 13 (step S1306), the distributed processing instruction means 302
Is passed (step S1307), and the distributed processing instruction means 302 calculates the remainder value 1 (step S130).
8) The cache determining unit 10 makes a request to the upper WWW proxy device to which 1 has been assigned.
3 (step S1309), the upper WWW proxy device is selected.

Next, the distributed processing instructing means 302 makes the cache judging means 1 request the selected upper WWW proxy device 303 to acquire WWW data.
03, and the cache determining unit 103 requests the cache determining unit 103 of the selected higher-level WWW proxy device 303 for the WWW data to be acquired (step S905).

Next, the selected upper WWW proxy device 303 acquires the requested WWW data (step S906). Step S906 will be described later.

Further, the upper WWW proxy device 303
Transmits the obtained WWW data to the cache determination unit 103 of the lower-level WWW proxy device 102 (step S907), and ends.

Next, step S906 in FIG. 9 will be described in detail with reference to FIG. WWW proxy device 10
In response to the request made by the second cache determination unit 103, the selected upper WWW proxy device 303
Cache determination unit 103 of WWW proxy device 303
(Step S1001).

[0091] The cache determination means 103 of the WWW proxy device 303 stores the URL of the requested WWW data.
Is passed to the tallying / analyzing means 106 of the WWW proxy device 303 and is counted as an access frequency (step S50).
2). Next, the cache determination unit 103 of the WWW proxy device 303 determines that the requested WWW data is
Cache memory means 104 of W proxy device 303
Then, an inquiry is made as to whether the data is cached by the cache disk means 105 (step S503).

If it is determined in step S503 that the WWW data is cached, the corresponding WWW data is stored in the cache determination unit 1 of the lower-level WWW proxy device 102.
03 (step S1002). Further, a pre-read process is performed on the requested WWW data (step S507), and the process ends.

If it is determined in step S503 that the cache is not cached, the WWW proxy device 30
3 is the cache determination unit 103
The WW data is obtained from the WWW server of the Internet 100 or the like (step S504). Next, the cache determination unit 103 of the WWW proxy device 303 performs a cache process on the obtained WWW data (step S50).
5).

Further, the cache determining means 103 of the WWW proxy device 303 converts the acquired WWW data into the cache determining means 10 of the lower-order WWW proxy device 102.
3 (step S1002), and performs a prefetching process on the obtained WWW data (step S507).
finish.

[0095]

According to the WWW proxy device of the present invention,
By using a memory device and a disk device together as a cache device, high-speed, efficient, and highly reliable cache processing can be performed.

[0096] In the prefetching process, by prefetching only WWW data that is highly likely to be requested by the user, more efficient cache processing and reduction of network load can be realized.

[Brief description of the drawings]

FIG. 1 is a simple principle diagram of a WWW proxy device according to the present invention.

FIG. 2 is a simple principle diagram of a WWW proxy device and distributed processing according to the present invention.

FIG. 3 is a principle diagram relating to a distributed control means of the WWW proxy device according to the present invention.

FIG. 4 is a diagram illustrating the principle of a prefetching unit according to the present invention.

FIG. 5 is an explanatory diagram of a processing procedure when a WWW proxy device according to the present invention is used alone.

FIG. 6 shows a cache processing procedure (step S) in FIG.
505).

FIG. 7 shows a pre-read processing procedure (step S50) in FIG.
Description of 7).

FIG. 8 illustrates a processing procedure when a plurality of WWW proxy devices according to the present invention are used.

FIG. 9 is a view for explaining a processing request procedure (step S801) to an upper WWW proxy device in FIG. 8;

FIG. 10 shows a procedure for acquiring WWW data by the selected higher-level proxy device in FIG. 9 (step S906).
Description.

FIG. 11 is an example of a method of calculating an access frequency for a recent certain period.

FIG. 12 shows an example of an importance database.

FIG. 13 is a diagram showing an example of a method for selecting a higher-level WWW proxy device.

FIG. 14 is a diagram showing the role of a general WWW proxy device.

FIG. 15 is a diagram showing an example of a physical configuration of a WWW proxy device according to the present invention.

[Explanation of symbols]

100: Internet, etc., 101: WWW browser,
102: WWW proxy device, 103: cache determining means, 104: cache memory means, 105: cache disk means, 106: totaling / analyzing means, 10
7 ... input means, 108 ... distributed control means, 109 ... database, 110 ... control management means, 111 ... look-ahead means,
Reference numeral 200: upper WWW proxy device, 301: numerical value conversion means, 302: distributed processing instruction means, 303: selected upper WWW proxy device, 401: prefetch control means, 40
2 ... importance database, 403 ... importance determination means, 1
301: upper WWW proxy device (0 is assigned), 1
302: Upper WWW proxy device (assigned 1), 1
303: Upper WWW proxy device (2 is assigned), 1
304: Upper WWW proxy device (3 is assigned), 1
401 WWW client, 1402 WWW proxy, 1403 cache device, 1404 Internet, 1405 WWW server, 1501 WWW proxy device, 1502 processor, 1503 memory device, 1504 disk device, 1505 input device, 1
506: Communication control device, 1507: Cache memory device, 1508: Cache disk device, 1509
... Control management device, 1510 ... Internal communication bus, 1511 ...
External communication bus.

Claims (4)

[Claims] 1. A WWW proxy device for obtaining WWW data from a WWW server device in response to a request from a WWW browser, wherein the WWW proxy device temporarily stores WWW data and a cache disk. Means, tallying / analyzing means for tallying / analyzing log data of transfer of WWW data on the network, input means for inputting an instruction from a user, and contents of the input means and the tallying / analyzing means. The WWW
Determining whether data is to be stored by the cache memory means or the cache disk means;
Writing and reading of data by the cache memory means or the cache disk means;
A cache determining unit for obtaining the WWW data from a WWW proxy device or a WWW server device, wherein the cache determining unit caches the WWW data to be normally stored in the cache memory unit, WWW data that satisfies certain conditions for WWW data overflowing due to replacement of cache data in the memory means
A WWW proxy device wherein only data is cached in the caching disk means. 2. The WWW proxy device according to claim 1, wherein the WWW proxy device has an access frequency of only log data for a recent certain period out of the log data totaled by said totaling / analyzing means. A WWW proxy device, which performs a cache process only on WWW data. 3. The WWW proxy device according to claim 1, wherein said predetermined condition for judging whether or not to cache WWW data overflowing due to replacement of cache data in said cache memory means in said cache disk means. Is calculated by a totalizing / analyzing means for totalizing / analyzing log data of transfer of WWW data on the network.
W characterized by being higher in data access frequency
WW proxy device. 4. The WWW proxy device according to claim 1, wherein said predetermined condition for judging whether or not WWW data overflowing due to replacement of cache data in said cache memory means is cached in said cache disk means. Is an input content of the input means for inputting an instruction from a user.