Exploring Planning and Scheduling for Web Services, Grid, and Autonomic Computing
Papers from the AAAI Workshop
Biplav Srivastava and Jim Blythe, Cochairs
There has been increasing interest in service composition for management of tasks and resources. Creating and managing workflows of tasks on the web and on computational grids has simplified data integration and business process integration. Now widely distributed services can be discovered and composed into workflows either manually or automatically, revolutionizing integration approaches to distributed services.
Management of resources, including software and hardware that may be located centrally or distributed across a network, has emerged as a significant obstacle to reducing the cost of information technology while increasing business productivity. Self-management of systems for configuration, protection, recovery and optimization is promoted as autonomic computing in industry. Here, workflows have been adopted as the underlying representation to connect interrelated tasks required for self-management.
AI planning and scheduling techniques will play an essential role in managing workflows of task ranging from their generation, storage and retrieval, analysis, composition, allocation of resources, execution and repair. However, many issues remain to be resolved. These include (1) determining how to bridge the gap between existing plan and workflow languages, whether existing domain and workflow languages are appropriate, or whether new features or languages are needed; (2) forming precise characterizations of service life cycle including composition, execution, and workflow management, (3) identifying the most appropriate ways to formalize service discovery and composition problems; (4) highlighting important challenges for planning and scheduling systems to be maximally effective in these areas; and (5) supporting operational robustness in 24/7 environment.