JP2011154652A - Cache system of dynamically generated data and control method - Google Patents

Abstract

<P>PROBLEM TO BE SOLVED: To cache dynamically generated data solving the problem wherein it is difficult to use a cache since the content of the dynamically generated data changes depending on an application in a method that uses the cache to improve performance in a Web service, etc. <P>SOLUTION: The dynamically generated data is stored in the cache by using a function that returns the same value when dynamically generated data is identical, wherein the result of the function is used as a key. When data is requested, it is checked whether data having the result of function implementation as the key is present in the cache, and if the data is present, the data of the past processing result is reused. <P>COPYRIGHT: (C)2011,JPO&INPIT

Description

  The present invention relates to a cache of data dynamically generated by a program or the like.

  A method of improving efficiency by storing in a cache data that may be reused once used has been used in various fields.

  In order to provide services efficiently in the WWW, a method of caching content has been frequently used in the past, and static content is widely used in common.

  However, it is difficult to use a cache for content that is dynamically generated by a program or the like because the content changes depending on various parameters such as HTTP request parameters, databases, external files, and session status. The use is limited.

  However, even in the process of generating dynamic content, different contents are not always generated every time it is executed, and there are many situations in which the same contents are generated, so if data can be reused using a cache, Efficiency can be improved.

  As a conventional cache technology related to dynamic data, there is a method in which cached data is used for a portion in which content does not change, and only a portion in which the content changes is dynamically generated and inserted into a corresponding portion. Using this method, which focuses on the fact that the whole does not change dynamically, the number of dynamically generated parts can be reduced, but the process of dynamically generating part is still executed.

  In addition, as a conventional technique for caching dynamically generated data with a proxy server, transfer data and fingerprints obtained by applying a hash function to the data are registered in the cache in association with each other, and transfer is attempted. If the data to be stored exists in the cache, there is a method of transferring a fingerprint instead of the data. If this method is used, the amount of communication can be reduced, so that the network load can be reduced, but the processing amount in the server cannot be reduced.

  If not only the amount of communication but also the amount of processing can be reduced, it will lead to improvement in performance and labor saving. Therefore, it is desirable to realize a cache system that reduces the dynamic data generation processing itself.

  The technology for caching dynamically generated data can be applied not only to WWW services, but also to all fields where processing requiring time and power is repeated, improving efficiency in various fields of networked computer systems. It becomes possible.

  A large-capacity storage device is considered necessary to realize a cache that stores dynamically generated data. However, due to advances in hardware technology, a large-capacity storage device can be used at low cost. The implementation and use of the present invention is considered realistic.

  In addition, even if the usage of the storage device increases, the usage of the CPU that consumes a large amount of power can be reduced. Therefore, the power consumption of the information device can be reduced by adjusting the power consumption of the entire system. Can also contribute.

  Considering these factors, even if there are many patterns of dynamically generated data, it is considered effective to use a cache to avoid repeated processing and reuse the generated data efficiently. .

JP 2002-108671 A JP 2002-268935 A JP 2003-242017 A JP 2005-107662 A

Edge Side Includes (ESI) Overview, [online], March 4, 2008, [Search January 18, 2010], Internet <URL: http://otndnld.oracle.co.jp/products/app_server/ cache / pdf / esi_overview.pdf> Virnish ESI Features, [online], [Search January 18, 2010], Internet <URL: http://varnish.projects.linpro.no/wiki/ESIfeatures> ESI Language Specification 1.0, [online], [searched on January 18, 2010], Internet <URL: http://www.w3.org/TR/esi-lang> Dynacache Basics: Part 1 “What is Dynacache?” [Online], [searched on January 19, 2010], Internet <URL: http://www.ibm.com/developerworks/jp/websphere/ library / was / was51_dynacache / 1.html>

  The problem to be solved is to realize a data cache system in which dynamically generated contents change.

  The present invention uses a function that returns the same value when dynamically generated data is the same, stores the execution result in the cache using the result of this function as a key, and then the data is requested. Sometimes the main feature is to use data stored in the cache.

  When the cache system according to the present invention is used, the previously generated data is reused under the condition of generating the same data, so that the performance is improved and the processing amount can be reduced, so that the power consumption can be reduced. is there.

FIG. 1 is a block diagram of a dynamically generated data cache system. FIG. 2 is a conceptual diagram of a WWW server using a dynamic data cache. (Example 1) FIG. 3 is a conceptual diagram of a proxy server using a dynamic data cache. (Example 2)

  In the dynamic data cache system according to the present invention, a data identity determination identifier (Identical ID) acquisition function is required together with a data generator for generating data. This function checks the input data, context, etc., and returns the same value when the data generated by the data generator is the same.

  When dynamically generated data is required, the system calls the data identity identifier (Identical ID) acquisition function, and the identifier obtained thereby and the data generator assigned to uniquely identify the data generator Check if the data with the identifier (Data Generator ID) as a key exists in the cache.

  If data exists in the cache, the data stored in the cache is returned as an execution result.

  If the data does not exist in the cache, the data is created by the application data generator (Data Generator), the execution result is returned, and the contents of the data are stored in the dynamic data cache.

  FIG. 2 shows an embodiment of the present invention, which is a WWW server using a dynamic data cache.

  The WWW server stores in advance a program (Data Generator) for generating content created by the developer of the Web application and a data identity determination identifier acquisition function for the content generated thereby.

  The data identity determination identifier acquisition function can be created as one function in the Web application program, and this function returns the same value when the generated content is the same.

For example, if the content to be generated depends on the session variable id, the contents of the file / tmp / updatetime, and the GET parameter cmd, all of which have the same value, the same content is returned. As a function, a function that returns a character string obtained by concatenating these three values can be considered. When the function is created by a PHP script, it can be described as follows.

  Since the data identity discrimination identifier acquisition function is repeatedly executed and greatly affects the efficiency of the system, it is important to make this function efficient. As a method for improving the efficiency, for example, if a file is used instead of a database with high input / output costs, or if the value used to obtain the data identity identifier is known before the data request, the value is obtained in advance. It may be possible to reduce the cost of inputting information and processing necessary to execute this function.

  It is thought that the performance can be improved by designing the system so that the developer of the Web application can efficiently determine the condition that the content contents are the same and creating a data identity determination identifier acquisition function. It can also be automatically generated by software that analyzes the data generation program.

  In this embodiment, a URL is used as a data generator identifier (Data Generator ID) for uniquely identifying a program, and a program that generates content using the URL and a data identity determination identifier (Identical ID) as a key. A cache is implemented by storing the generated data in a table.

  When receiving the HTTP request from the client, the WWW server of this embodiment calls a data identity determination identifier acquisition function corresponding to the Web application program specified by the URL, and acquires a data identity determination identifier (Identical ID).

  Using the acquired data identity identifier (Identical ID) and the URL requested by the client as a key, check whether the content data generated in the past is stored in the dynamic content cache. Return data to the client as a response.

  If there is no data in the dynamic data cache, execute the program that generates the content to generate the data, return it as a response to the client, and store the data in the dynamic data cache for future use .

  In the response sent to the client, the ETag is included in the response header, and the value of the execution result of the data identity determination identifier acquisition function is sent as the value, so that the client uses the If-None-Match header, It is possible to save the bandwidth required for communication.

  The embodiment of FIG. 3 is a proxy server using a dynamic data cache according to the present invention.

  In this embodiment, as a WWW server that generates content to be used, a server programmed to send a value of an execution result of a data identity determination identifier acquisition function as an ETag value in a response header when a response is returned is used. .

  When the proxy server receives the HTTP request from the client, it transmits a HEAD request to the WWW server that provides the content in order to acquire the data identity determination identifier for the URL.

  The WWW server that has received the request from the proxy server stores the value of the execution result of the data identity determination identifier acquisition function corresponding to the requested URL in the ETag response header, and returns a response to the proxy server.

  The proxy server confirms whether data using the requested URL and the obtained data identity determination identifier as a key exists in the dynamic data cache, and if it exists, generates a response from the cached data and returns it to the client.

  When there is no data in the cache, the content is acquired from the WWW server, the obtained content is returned to the client as a response, and the content content is stored in the dynamic data cache.

  Normally, the ETag header returns a character string representing the component version. When a new version of content is returned, the proxy server cache data is generally discarded, but the proxy server of this embodiment receives it. Since the value of the ETag header is a character string representing the data identity determination identifier, even when a response having a different ETag header value for one URL is received, it is stored in the cache as another valid data. It will be.

DESCRIPTION OF SYMBOLS 1 Dynamic data cache 2 Dynamic data storage table 3 Data identity determination identifier acquisition function 4 Data generator 5 Dynamic data generation application 6 Dynamic data cache management program

Claims (5)

  A dynamic data caching method characterized by using a function that returns the same value when the same data is generated as a processing result for data dynamically generated by processing information, and this method A device that caches data.   A WWW server having a function of caching content generated by a Web application using the method according to claim 1.   3. The WWW server according to claim 2, wherein a function created by a web application developer or a person who understands the operation of the web application is used as a function for determining whether dynamically generated data has the same contents. A WWW server characterized by the above.   4. The WWW server according to claim 3, wherein a function created by a web application developer or a person who understands the operation of the web application is used as a function for determining whether dynamically generated data has the same contents. Instead, a WWW server using a function generated by analyzing a data generator by a machine.   A proxy server for controlling a cache using the method according to claim 1, wherein an origin server that generates data executes a function that returns the same value when the same data is generated, and the value is used as the proxy server. A proxy server that caches data generated in the origin server by using it on the server.

Images