Abstract: Decentralized orchestration of composite services results in a number of flow topologies which differ in their messaging patterns and the distribution of flow code amongst different partitions of a topology. Different performance metrics result for different topologies, and this is governed by the availability of resources, such as CPU cycles, memory, network bandwidth, and so on. A performance model is used to evaluate the performance of different topologies based on availability of resources, and dynamically direct client requests between different topologies according to prevailing conditions.

Abstract: A technique for orchestrating composite web services in a constrained data flow environment. Data flow constraints represent the restrictions which the domains involved in composition might want to impose on access to the data they provide or the source from which they can accept data. The data flow constraints are overcome by breaking down an input composite web service specification into topologies (a topology is a set of partitions communicatively connected with one another at runtime). The system applies a rule-based filtering mechanism to choose a topology that does not violate any data flow constraints. Each of the partitions belonging to the chosen topology is executed within the same domain as the corresponding web service it invokes, thus having the same access rights as the corresponding web service. The composite web service is orchestrated in a decentralized fashion using these partitions, thereby ensuring that no data flow constraints are violated.

Abstract: Plan construction and selection decision phase is conducted separately from a plan assignment phase. Furthermore, the estimation of runtime variables is separated from the assignment of service instances. Moreover, at each stage, feedback is provided to enable the composition of the plan to be continuously refined. Optimization of runtime metrics can also be modelled for selection and composition of web services, or any other service-oriented architecture technology in which an application is described using a predetermined description language.

Abstract: Disclosed are methods, apparatus and computer programs for improving performance and resource utilization of cluster-based data processing systems. The number of persistent connections between first and second data processing units of a system is adjusted according to monitored performance of communications between the nodes. The monitored performance measure is the queuing delay between the nodes, and the number of connections is increased and decreased as changing load conditions increase and decrease the queuing delay. A method of establishing and closing persistent connections in response to a monitored queuing-delay between a front-end gateway computer and each of a cluster of back-end servers is used to improve the performance of a cluster-based Web server or provision of Web services. Unwanted connections are closed to enable system resources to be allocated to other processes.