Electrical computers and digital processing systems: multicomput – Remote data accessing – Accessing a remote server
Reexamination Certificate
2002-11-04
2004-11-23
Lin, Wen-Tai (Department: 2154)
Electrical computers and digital processing systems: multicomput
Remote data accessing
Accessing a remote server
C709S217000, C711S135000
Reexamination Certificate
active
06823374
ABSTRACT:
FIELD
This patent relates to accelerating the serving of dynamically generated web documents through caching, without significantly compromising freshness of the content.
BACKGROUND
It is commonly required to serve fresh content to web users, for example in the form of web pages that are dynamically generated by formatting the latest information retrieved from a database.
1
Examples of such web pages might include pages of stock quotes, election results and auction prices. While such dynamic delivery offers the user the freshest content, dynamic delivery often exhibits poor performance characteristics under load. When many users visit the web-site at the same time, the server's computational resources can be inadequate to individually generate the page for each user. As a result, users see poor performance, or sometimes no performance owing to server crashes. Some examples of such situations include: (a) an accounting or tax related web site at the close of an accounting period; (b) a news site after the occurrence of a significant news event; and (c) a sports information site during a major sports event.
1
The terms “content,” “web page,” “information” and the like should not be restricted to non-functional subject matter, since computer code and other functional subject matter is also often provided by content servers.
In order to improve performance, caches are often installed in tandem with the content server. A cache is a server or other mechanism that stores and reuses content for re-use. By combining a server and a cache, static content that does not change with time can be offloaded to the cache, freeing up the content server to focus primarily on dynamic content.
However, since dynamic content (such as transactional and/or personalized information) changes frequently and across users, it cannot readily be cached and re-used. Thus, in the simplest configuration, dynamic content is not offloaded to the cache. Nevertheless, some system administrators do choose to cache dynamic content by specifying a (relatively short, fixed) time interval so that the content is never out of date by more than the time interval. But it is often infeasible to significantly increase the caching interval (e.g., beyond a few seconds) without significantly compromising freshness (for at least some content). Also, such dynamically cached content is typically out of date by a time equal to the specified time interval, whether or not the content server is experiencing a heavy load. For example, for web sites involved in auctions, the load on the server tends to peak at the close of an auction, with the content also changing frequently during that time period. This is true whether or not the load is nearing the capacity of the system. Thus a simple, fixed caching time interval cannot simultaneously satisfy the competing constraints of heavy loads and rapidly changing content.
Some systems (for example, the 9iAS Web Cache from Oracle) operate by adjusting the caching interval as a function of the number of connections to the content server. However, not all connections make equal demands on the server, and the number of connections is rarely proportional to the ability of the server to respond. For example, even a small number of high demand connections (e.g., downloading multiple files of hundreds of megabytes each, especially over a slow connection) will have a much greater impact than a large number of low demand connections (e.g., checking simple pages with only one changing field). Furthermore, for similar reasons, it is difficult (e.g., for system, administrators) to establish, in a meaningful way, a baseline against which to compare a metric such as the number of connections. That is, comparisons based on numbers of connections are likely to be arbitrary rather than truly indicative of the ability of the content server to respond to requests for content.
REFERENCES:
patent: 5664106 (1997-09-01), Caccavale
patent: 5892937 (1999-04-01), Caccavale
patent: 5907681 (1999-05-01), Bates et al.
patent: 6189080 (2001-02-01), Ofer
patent: 6266742 (2001-07-01), Challenger et al.
patent: 6343065 (2002-01-01), Serbest
patent: 6351767 (2002-02-01), Batchelder et al.
patent: 02001168921 (2001-06-01), None
Oracle9iAS Web Cache, A Technical White Paper, Jun. 2001, Oracle Corporation.
Jawahar Janardhanan
Kausik Balas Natarajan
Race Daren A.
FineGround Networks
Law Offices of Rowana H. Yang
Lin Wen-Tai
LandOfFree
Adjusting the cacheability of web documents according to the... does not yet have a rating. At this time, there are no reviews or comments for this patent.
If you have personal experience with Adjusting the cacheability of web documents according to the..., we encourage you to share that experience with our LandOfFree.com community. Your opinion is very important and Adjusting the cacheability of web documents according to the... will most certainly appreciate the feedback.
Profile ID: LFUS-PAI-O-3291082