Issues and Evaluations of Caching Solutions for Web Application Acceleration

A simulation study of the effects of multi-path approaches in e-commerce applications

“…The edge servers host the business logic for executing transactions against the data centers, which are responsible for guaranteeing the availability and consistency of the application data. The edge serves can perform caching of (dynamic) Web objects, and can also perform caching of application data residing at the data centers via secondary DBMSs residing at edge server locations [12]. This allows the edge servers to reply to read only requests from clients without the need for interaction with the back-end data centers.…”

“…[12]), they still rely on direct access to the origin (primary) DBMS in case of client requests altering the application state, such as product ordering. Therefore, increased proximity to the clients cannot address the level of service seen by the users issuing update requests.…”

“…To limit such a phenomenon, which is actually detrimental to the business process supported by the e-Commerce site, a spectrum of solutions have been proposed in order to ensure application availability and timely delivery of contents to the end users [9,10,12,13,19]. A key approach for most of these solutions is the employment of both Web object caching techniques and also DBMS replication techniques, which can provide the benefits of overcoming network overloads (or failures) by increasing the proximity between clients and contents, thus allowing for enhanced response time and application availability.…”

Constraint-Based Model Weaving

“…For example, keeping the product information in the cache that is most frequently requested will be most beneficial. -The architecture of the back-end database servers providing product, account, and order data to the application determines the cost of a query [31]. For example, in a simple Pet Store deployment where the back-end database is co-located with the Pet Store's application server, queries will be less expensive than in an arrangement where queries must be sent across a network to the database server.…”

WOD – Proxy-Based Web Object Delivery Service

“…Even if the proxy caches some of such web objects, they will most likely be replaced by the cache replacement mechanism before the user has a chance to download them later. Several new caching strategies [9,10,11,12] have been proposed, where a fragment-level caching strategy is applied, to solve the problems. A recent work [13] focuses on integrating a Web content adaptation algorithm and a caching mechanism to serve dynamic content in a mobile computing environment.…”