Integrating Web Caching And Web Pre Fetching

Integrating Web Caching and Web Prefetching In Client-Side Proxies

Objective 

The motivation for our project is to design an innovative cache replacement algorithm, not only considering the caching effect in the Web environment, but also evaluating the prefetching rules provided by various prefetching schemes.

Introduction (Cont.) Purpose •

Reduce user perceived latency

Key Issue •

Efficiently predict and fetch the future URLs likely to be visited by the user in advance  Efficient prediction algorithms

Cost •

Increased network and server load due to errors in prediction  Appropriate threshold value

Web caching

Web Caching : Satisfying user Web requests by servers other than original Web servers publishing the requested Web objects 1. 2. 3.

Latency, External traffic, Load on web servers and routers.

Deployed at: Corporate network boundaries, ISPs, Web Servers, etc.

Web Cache

Browser

Browser Cache

Client

Browser

Web Server

Centralized Web Cache Browser Cache

Client Corporate LAN

Internet

Web caching 

Benefits:   

 

Improves web performance (reduces access latency) Increases web capacity Alleviate traffic congestion (reducing network bandwidth consumption) Reducing number of client requests (workload) Possibly improve failure tolerance and robustness of Web (maintaining cached copies of web objects for unreachable networks)

Web Prefetching •

Prefetching- caches web objects in anticipation of users’ future needs Focuses on making cache-related storage capacity decisions (storage capacity limits the number of prefetched web objects)

Therefore allocate cache storage in prefetching

Categories •

DNS Prefetching 

•

Connection Prefetching  

•

Web client initiate the name resolution process in advance of the user’s request Set up TCP connection before the user’s request Timing issue is important

HTTP Prefetching 

Web client issuing HTTP request in advance and cache the response

Web prefetching techniques 

Client-initiated policies  User A is likely to access URL U2 right after URL U1



Server-initiated policies  Anticipate future requests based on server logs and proactively send the corresponding Web objects to participating cache servers or client browsers Hybrid policies  Combine user access patterns from clients and general statistics from servers to improve the quality of prediction



Who Initiates the Prefetching

Server Side Prefetching •

•

Uses information from the web server to predict the documents that will be requested soon  Prefetch the object from the disk to the main memory of the server Provides hints to the client to improve the client’s response time  The client then can initiate a prefetch

Who Initiates the Prefetching (Cont.) Proxy Side Prefetching 



Proxy prefetches documents from web servers and stores in proxy cache Proxy can gather information from multi-client to multiserver, makes it possible to do good prediction

Who Initiates the Prefetching (Cont.) Client Side Prefetching • • •

Works with browser Learns the personal profile and predicts future requests Uses the available bandwidth to retrieve documents when the user is stop to read web contents

Modules •

System environment  

•

Cache replacement algorithm  

•

System model Prefetching rule

Normalized profit function Algorithm IWCP

Performance analysis   

Impact of cache capacity Impact of confidence threshold Execution overhead

module1

•

System environment  

System model Prefetching rule

Log

-

request

request

Response + hint

response Web Client

Prediction Engine

Proxy server

Request Processing module

Web object depository

System model of web prefetching and web caching

Prefetching rule repository

Web server

Prefetching rule 

Rule1: A prefetching rule is an implication of the form o1,..,oi c oi+1 where o1,..,oi , oi+1 Є O and o1,..,oi , oi+1 is an access sequence in D and c is the confidence of the prefetching rule



Rule2: The confidence c of the prefetching ruleo1,..,oi c oi+1 is a conditional probability of p(o1,..,oi , oi+1 )/ p(o1,..,oi )



An object oi+1 is referred to as an implied object if and only if the prefetching rule is triggered by some client who has already referenced the objects in precedent request.

Hybrid Prefetching-System Model System model of prefetching proxy

System Specification: 



Hardware Specification: Processor Type : Pentium -IV Speed : 1.2 GHZ Ram : 128 MB RAM Hard disk : 20 GB HD Software Specification: Operating System : Linux,Win2000 Pgm Package : JAVA / J2EE Protocol : HTTP. Web Server : Apache Tomcat 5.0