Web services is a relatively recent initiative that aims to promote program-toprogram interaction across the Internet, but while web services is based on a set of XML standards, new standards continue to emerge and existing standards to evolve. Also, web services relies on Remote Procedure Call (RPC) for communication and is thus influenced by the semantics of RPC. In this research, we investigated the juxtaposition of RPC with Generative Communications (GC). GC is a communication paradigm where messages exist independently of the sender and receiver and are stored in a network accessible buffer called a &quotspace": this leads to interactions which are inherently decoupled (in time and space). Also, messages are addressed to recipients by their content, rather than by network addresses, opening up the possibility for one-to-many interactions. These aspects are a marked departure from the RPC paradigm and introduce two main implications: 1) GC messages can be intercepted when in-transit between participants thus introducing the opportunity for mediation and 2) GC can be used as the basis for the aggregation of simple services into more complex ensembles. In this research, we explored these possibilities by creating proof-of-concept prototypes in three areas. 1) Mediation - GC based mediation was used to intercede between clients and services to allow a client using one protocol to interact with a service using a different protocol. For example, a GC based client interacting with a SOAP service (leading to backward compatibility). 2) Location services - a location service is a GC based service that performs a similar function to a UDDI registry but can be treated as just another service rather than part of an architecture. 3) Aggregation - a workflow design was used as the basis of an aggregated service using GC as the means by which the aggregation elements interact. We concluded that GC provides a natural platform for mediation, location services and aggregation and that these aspects could be combined to produce a holistic service environment.

Scalable on-demand content delivery systems, designed to effectively handle increasing request rates, typically use service aggregation or content replication techniques. Service aggregation relies on one-to-many communication techniques, such as multicast, to efficiently deliver content from a single sender to multiple receivers. With replication, multiple geographically distributed replicas of the service or content share the load of processing client requests and enable delivery from a nearby server.<p>Previous scalable protocols for downloading large, popular files from a single server include batching and cyclic multicast. Analytic lower bounds developed in this thesis show that neither of these protocols consistently yields performance close to optimal. New hybrid protocols are proposed that achieve within 20% of the optimal delay in homogeneous systems, as well as within 25% of the optimal maximum client delay in all heterogeneous scenarios considered.<p>In systems utilizing both service aggregation and replication, well-designed policies determining which replica serves each request must balance the objectives of achieving high locality of service, and high efficiency of service aggregation. By comparing classes of policies, using both analysis and simulations, this thesis shows that there are significant performance advantages in using current system state information (rather than only proximities and average loads) and in deferring selection decisions when possible. Most of these performance gains can be achieved using only local (rather than global) request information.<p>Finally, this thesis proposes adaptations of already proposed peer-assisted download techniques to support a streaming (rather than download) service, enabling playback to begin well before the entire media file is received. These protocols split each file into pieces, which can be downloaded from multiple sources, including other clients downloading the same file. Using simulations, a candidate protocol is presented and evaluated. The protocol includes both a piece selection technique that effectively mediates the conflict between achieving high piece diversity and the in-order requirements of media file playback, as well as a simple on-line rule for deciding when playback can safely commence.

Scalable on-demand content delivery systems, designed to effectively handle increasing request rates, typically use service aggregation or content replication techniques. Service aggregation relies on one-to-many communication techniques, such as multicast, to efficiently deliver content from a single sender to multiple receivers. With replication, multiple geographically distributed replicas of the service or content share the load of processing client requests and enable delivery from a nearby server.<p>Previous scalable protocols for downloading large, popular files from a single server include batching and cyclic multicast. Analytic lower bounds developed in this thesis show that neither of these protocols consistently yields performance close to optimal. New hybrid protocols are proposed that achieve within 20% of the optimal delay in homogeneous systems, as well as within 25% of the optimal maximum client delay in all heterogeneous scenarios considered.<p>In systems utilizing both service aggregation and replication, well-designed policies determining which replica serves each request must balance the objectives of achieving high locality of service, and high efficiency of service aggregation. By comparing classes of policies, using both analysis and simulations, this thesis shows that there are significant performance advantages in using current system state information (rather than only proximities and average loads) and in deferring selection decisions when possible. Most of these performance gains can be achieved using only local (rather than global) request information.<p>Finally, this thesis proposes adaptations of already proposed peer-assisted download techniques to support a streaming (rather than download) service, enabling playback to begin well before the entire media file is received. These protocols split each file into pieces, which can be downloaded from multiple sources, including other clients downloading the same file. Using simulations, a candidate protocol is presented and evaluated. The protocol includes both a piece selection technique that effectively mediates the conflict between achieving high piece diversity and the in-order requirements of media file playback, as well as a simple on-line rule for deciding when playback can safely commence.