Electrical computers and digital processing systems: multicomput – Computer network managing
Reexamination Certificate
1999-09-15
2003-03-11
Barot, Bharat (Department: 2154)
Electrical computers and digital processing systems: multicomput
Computer network managing
C709S203000, C709S213000, C709S214000, C709S224000, C709S245000, C711S118000, C711S216000
Reexamination Certificate
active
06532492
ABSTRACT:
FIELD OF THE INVENTION
The present invention relates to communications and more particularly to cache management for supporting communications.
BACKGROUND OF THE INVENTION
The Internet is a worldwide decentralized network of computers having the ability to communicate with each other. The Internet has gained broad recognition as a viable medium for communicating and interacting across multiple networks. The World-Wide Web (Web) was created in the early 1990's, and is comprised of server-hosting computers (Web servers) connected to the Internet that have hypertext documents (referred to as Web pages) stored therewithin. Web pages are accessible by client programs (e.g., Web browsers) utilizing the Hypertext Transfer Protocol (HTTP) via a Transmission Control Protocol/Internet Protocol (TCP/IP) connection between a client-hosting device and a server-hosting device. While HTTP and hypertext documents are the prevalent forms for the Web, the Web itself refers to a wide range of protocols including Secure Hypertext Transfer Protocol (HTTPS), File Transfer Protocol (FTP), and Gopher, and content formats including plain text, Extensible Markup Language (XML), as well as image formats such as Graphics Interchange Format (GIF) and Joint Photographic Experts Group (JPEG).
An intranet is a private computer network conventionally contained within an enterprise and that conventionally includes one or more servers in communication with multiple user computers. An intranet may be comprised of interlinked local area networks and may also use leased-lines in a wide-area network. An intranet may or may not include connections to the outside Internet. Intranets conventionally utilize various Internet protocols and, in general, often look like private versions of the Internet. An intranet user conventionally accesses an intranet server via a Web browser running locally on his/her computer.
As is known to those skilled in this art, a Web page is conventionally formatted via a standard page description language such as HyperText Markup Language (HTML), which typically contains text and can reference graphics, sound, animation, and video data. HTML provides for basic document formatting and allows a Web content provider to specify anchors or hypertext links (typically manifested as highlighted text) to other Web servers and files. When a user selects a particular hypertext link, a Web browser reads and interprets an address, called a Uniform Resource Locator (URL) associated with the link, connects the Web browser with a Web server at that address, and makes a request (e.g., an HTTP request) for the file identified in the link. The Web server then sends the requested file to the Web client which the Web browser interprets and displays to the user.
In communications between a client and a server, such as over the Internet, and in other applications involving transfers of large amounts of information over a communication channel, it is often desireable to improve the performance of the communication channel by reducing the volume of data which must be transferred. One known approach takes advantage of the fact that some communications contain redundant information, in other words, information which was previously communicated between a client and a server which has not changed in the time between the previous request initiating a transfer and a subsequent new request for the same information. For example, a Web browser may request a file with a particular associated URL address on multiple occasions. Various approaches using caches have been proposed which are intended to recognize when requested information is already available to the requesting device in a local cache. The web browser is then provided the locally available portion of the requested information from the local cache and only new information is transmitted to the requesting client over the communication channel.
An example of such a cache management system that provides for management of the contents of the cache to control the amount of memory required to support the cache is described in Charu Aggarwal, et al., “Caching on the World Wide Web”, IEEE Transactions on Knowledge and Data Engineering, January/February 1999. As described in this article, HTTP caches suffer from the fact that a large number of requests they serve are for objects which will never be requested again before they expire. A proposed solution from the article is a cache admittance protocol which uses a second cache containing only the URLs which have been requested. The actual object would then be cached when a certain frequency of reference to the object has been seen by the cache.
One problem with the approach described in Aggarwal is the amount of storage required. The URLs typically take up approximately 0.3% of the size of the actual objects. Furthermore, it has been suggested that the URL cache should have a capacity (in number of objects) twice as large as the cache used for actual objects—or about 0.6% of total cache size. As object caches for modem HTTP caches are on the order of 100 gigabytes, this would require about 600 megabytes of storage for the URL cache. Using this approach and given the present hardware capabilities of typical existing equipment, the resulting URL cache generally has to be stored in a secondary storage device as opposed to main memory storage. This typically imposes a performance penalty as the URL cache is accessed and updated very frequently. It is expected, in a typical application, that approximately 50-70% of requests will not be served from the object cache and will, therefore, require accessing and updating the URL cache. A further problem of the protocol suggested in Aggarwal results from the nature of typical World Wide Web traffic. A significant percentage, generally at least 40%, of the bandwidth of a communication channel used for HTTP traffic is for large objects (100 kilobytes and larger) which are accessed very infrequently compared to smaller objects. As a result of the less frequent access to large objects, the Aggarwall cache admittance protocol will make it less likely for those objects to be cached, further reducing the effectiveness of the cache by not providing cache support at a desired level as larger objects tend to take up a significant amount of communication channel bandwidth for transmission. For a further discussion of cache management systems such as those described in Aggarwall, reference is made to U.S. Pat. No. 5,924,116 entitled “Collaborative Caching of a Requested Object by a Lower Level Node as a Function of the Caching Status of the Object at a Higher Level Node” which is incorporated herein by reference as if set forth in its entirety.
It has also been proposed in U.S. Pat. No. 5,864,852 entitled “Proxy Server Caching Mechanism That Provides a File Directory Structure and a Mapping Mechanism Within the File Directory Structure” (“the '852 patent”) to use hashing to compress URLs to a fraction of their length. The compressed address is then used as a key for retrieving an associated object. However, the proposal of the '852 patent still stores the entire URL to check for collisions in an attempt to avoid errors as the proposal relates to object retrieval rather than admission control for a cache.
In light of the above discussion, a need exists for improvements in cache management protocols.
SUMMARY OF THE INVENTION
In view of the above discussion, it is an object of the present invention to provide methods, systems and computer program products which can provide for cache management with a reduced memory requirement to support the cache management protocol while still providing for improved performance of the supported communication channel.
A further object of the present invention is to provide such methods, systems and computer program products which can provide for a reduced usage of secondary storage devices in support of cache management.
These and other objects of the present invention may be provided by methods, systems and computer program products which take advantag
Barot Bharat
Clay A. Bruce
Myers Bigel & Sibley & Sajovec
LandOfFree
Methods, systems and computer program products for cache... does not yet have a rating. At this time, there are no reviews or comments for this patent.
If you have personal experience with Methods, systems and computer program products for cache..., we encourage you to share that experience with our LandOfFree.com community. Your opinion is very important and Methods, systems and computer program products for cache... will most certainly appreciate the feedback.
Profile ID: LFUS-PAI-O-3056941