Method and system for efficiently storing web pages for quick downloading at a remote device

Abstract

A web server for efficiently storing web pages for quick downloading at a remote device comprises a fast memory, a disk memory and a storage device. The fast memory stores a first plurality of web pages and receives a first signal from the remote device indicating selection of one of the first plurality of web pages. The disk memory stores a second plurality of web pages linked to the first plurality of web pages and transfers all of the second plurality of web pages that are linked to the selected one of the first plurality of web pages to the fast memory in response to the first signal. The disk memory also transmits a second signal to the storage device in response to the first signal. The storage device stores a third plurality of web pages linked to the first plurality of web pages and transfers all of the third plurality of web pages that are linked to the selected ones of the second plurality of web pages to the disk memory in response to the second signal. In addition, the fast memory deletes each of the non-selected first plurality of web pages in response to the first signal so as to allow for memory reuse and to aid in quickly downloading selected web pages to the remote device.

Description

TECHNICAL FIELD [0001] This invention relates to methods and systems for efficiently storing web pages at a server for quick downloading at a remote device in communication with the server. BACKGROUND ART [0002] Traditional web servers depend on sequential access to HTML (HyperText Markup Language) pages by loading hyperlinked pages as the client requests them while navigating through the current page. Each client initiated hit involves a random access to the memory, even if multiple clients access the same page. Thus, overall latency is high even when there is an overlap in the access. Latency refers to the amount of time it takes a web server to respond to a client beginning from the time the request was sent by the client to the server. To reduce latency in the network, some proposals have been made to have local machines, mostly proxy servers, to prefetch the pages linked to the current page before the client actually “hits” them. However, since all web pages may not be requeste...

AI Technical Summary

Benefits of technology

[0006] It is a general object of the present invention to provide a method and system for efficiently storing web pages at a server for quick downloading at a remote device in communication with the server.

[0007] In carrying out the above object and other objects, features, and advantages of the present invention, a method is provided for efficiently storing web pages for quick downloading at a remote device in a computer network including a server computer having a fast memory for storing a plurality of home web pages, a disk memory for storing a first plurality of web pages linked to the home web pages, and a storage device coupled to the computer fore storing a second plurality of web pages linked to the home web pages. The method includes the step of receiving a first signal from the remote device at the server computer indicating selection of one of the plurality of home pages. The method also includes the step of transferring at least one of the first plurality of web pages linked to the selected home page from the disk memory to the fast memory in response to the first signal. Still further, the method includes the step of transmitting a second signal from the disk memory to the storage device in response to the first signal. Finally, the method includes the step of transferring at least one of the second plurality of web pages linked to the selected home page from the storage device to the disk memory in response to the second signal wherein anticipated web pages linked to the selected home page are quickly accessible by the remote device.

Problems solved by technology

Thus, overall latency is high even when there is an overlap in the access.

However, since all web pages may not be requested by the client, storage is wasted.

Furthermore, bandwidth is also wasted.

Consequently, if the network is busy, aggressive prefetching can be expensive.

Since the access is usually to the disk, this penalty is only apparent on a small percentage of the accesses.

Thus, the main motivation for using HSM is cost.

However, in some file requests, latency will suffer.

Images