Policy enabled web caching

Details Policy enabled web caching Policy enabled web caching

1999-09-30

2004-08-03

Wiley, David (Department: 2143)

Electrical computers and digital processing systems: multicomput

Computer-to-computer data routing

Prioritized data routing

C707S793000, C709S219000, C709S229000

Reexamination Certificate

active

06772225

ABSTRACT:

FIELD OF THE INVENTION
The invention relates to the field of computer networks and in particular, to that of caching services for a computer network. More specifically, a caching service for use on the World Wide Web is disclosed which can improve user response times and reduce the amount of data transmitted over the Web.
BACKGROUND OF THE INVENTION
Caching has been used as a technology to improve user response time and to decrease network bandwidth utilization for distributed applications
10
such as web browsing. In the context of the World Wide Web, it is common to deploy caching services to improve the response time to users and to reduce the amount of data that is transmitted over the Web. The user invokes a browser program to retrieve data from the server in the network. A Uniform Resource Locator or URL identifies the address of the source of the data. Without an intervening cache, the request goes directly to the server, and the server returns the desired information.
When a cache is used in a network, a request for information is first sent to the cache. If the cache contains the desired information it is returned to the client. If the requested information is not found in the cache, the catch retrieves the information from the server and returns the information to the client. The cache will also store a copy of the information locally. Since the local storage of the cache is limited only a small portion of all possible information can maintain locally. Caches may implement various techniques to decide which information is maintained and which is discarded. A very common technique is the least recently used scheme, in which the URL information, which has not been accessed for the longest amount of time, is replaced by the new URL information being accessed. An overview of different caching schemes can be found in the document Aggarwal, C., et al.,
Caching on the World Wide Web
, IEEE Transactions on Knowledge and Data Engineering, Vol. 11, No. 1 January/February 1999, pp. 94-107.
There are two common modes for web caching: client proxy caches and transparent caches. In a client proxy cache, the browser is typically configured to send a request for information directly to the cache rather than the server. A transparent cache works like a proxy cache except that the browser need not be configured to send a request to the transparent cache. A transparent cache detects the packets belonging to the web-application by looking at information in the request such as the port number carried in IP packets (Web applications usually carry a port number of 80). The transparent cache then direct packets to the cache.
In some variants of the caching architecture, multiple caches can be deployed in the network. In one example, the proxy uses a static hashing of the URL to determine which cache should receive the request. Different caches can also be arranged in a hierarchy. The browser sends the request to a first cache. If the first cache does not find the URL information locally, the request is sent to a second cache. The second cache can send to a third cache, when the final cache is reached the request is sent to the server. The topology in which caching occurs is usually configurable. An algorithm for static hashing is CARP, or Cache Array Routing Protocol described in Ross, K.,
Hash Routing for Collections of Shared Web Caches
, IEEE, Network, November/December 1997, pp. 37-44.
Although caching in the web has been researched extensively, the effectiveness of the caches has been found to be relatively poor. Usually, the probability that a web page is found in the local cache is less than half, possibly around 35-40%. Thus, more than half of the requests result in a cache miss, i.e., they are not found in the cache.
The cache miss factor is high due to a variety of reasons. Many of the URLs, associated with pages which browsers attempt to access, identify data that is dynamic (e.g., a program to be executed at a server commonly called cgi-bin scripts). Some URLs identify information that is highly specific to the user (e.g., uses a cookie or creates a special identifier for the user). Some URLs identify special programs like a video or audio clip that need special handling or special protocols between the client and the server, and cannot be handled by an intervening cache.
Each cache miss adds extra latency to the packet request, which degrades the performance perceived by the browser. Since more than half of the requests result in a cache miss, traditional caching is more likely to result in degraded user performance than improved response time.
Figure 1
illustrates the different components that interact together to implement a prior art caching system. Within this caching system, a client
101
wishes to access a URL that identifies some information located at server
111
. The client
101
initially contacts a cache
105
. The cache
105
is connected to the client
101
by a network
113
. Typically, the network
113
is a campus network or fast local area network. The cache
105
serves multiple clients that are present on the network
113
, e.g., another client
103
in the network may access the same cache. The cache
105
connects to the server
111
via a network
107
. Typically, network
113
is faster than network
107
, so that response time is improved every time there is a hit in the cache. The cache
105
may coordinate caching with other caches in the network, i.e., cache
109
in the network.
In order to improve the caching behavior, a system of multiple proxying caches may be deployed. In addition, special caching servers, that can provide caching techniques that work with cookies or provide a specialized protocol for caching video and audio clips can be added to the network. While there are several caching architectures for interconnecting a multiple number of caches, most do not perform well due to a poor cache hit ratio. Plus, the number of proxying mechanisms deployed in the network adds additional latency in the caching architecture, and usually degrades the performance of the network, rather than improving it.
SUMMARY OF THE INVENTION
A policy enabled caching system based upon policy rules which define whether a request from a client is directed to a cache or a server. The client is coupled to a plurality of caches and to at least one server. The caches may store a subset of the data stored on the server.
The policy enabled caching system stores policy rules which comprise at least one matching condition, where every request containing a matching condition falls into an associated class. Each class will have an associated routing rule, where a routing rule defines the type of routing for all the requests which fall into that class.
The policy enabled caching system will receive the request from the client and classify the request according to the policy rules. The request is then routed according to the routing rule associated with the class to which the request belongs.


REFERENCES:
patent: 5924116 (1999-07-01), Aggarwal et al.
patent: 6006264 (1999-12-01), Colby et al.
patent: 6026413 (2000-02-01), Challenger et al.
patent: 6052718 (2000-04-01), Gifford
patent: 6112279 (2000-08-01), Wang
patent: 6138162 (2000-10-01), Pistriotto et al.
patent: 6185598 (2001-02-01), Farber et al.
patent: 6212565 (2001-04-01), Gupta
patent: 6330561 (2001-12-01), Cohen et al.
patent: 6345303 (2002-02-01), Knauerhase et al.
patent: 6351775 (2002-02-01), Yu
patent: 6374300 (2002-04-01), Masters
patent: 6449647 (2002-09-01), Colby et al.
C. Aggarwal et al., “Caching on the World Wide Web”, IEEE Transactions on Knowledge and Data Engineering, vol. 11, No. 1, Jan./Feb. 1999, pp. 94-107.

Affiliated with

Also associated with

Rating

Policy enabled web caching does not yet have a rating. At this time, there are no reviews or comments for this patent.

If you have personal experience with Policy enabled web caching, we encourage you to share that experience with our LandOfFree.com community. Your opinion is very important and Policy enabled web caching will most certainly appreciate the feedback.

Rate now

Comments { 0 }

Profile ID: LFUS-PAI-O-3278671