A multi-agent approach to immediate view maintenance for an operational data store

An operational data store provides current, consolidated data derived from distributed data sources. A data store combined with OLAP applications can provide a decision support or operation-monitoring tool for advanced enterprise resource planning and supply chain management systems. In order to keep the data store up-to-date, changes at the sources must be correctly propagated to the data store. Commercial systems typically designate a maintenance time when the data store is disabled for data updates. The maintenance time window could range from several hours to several days, depending on the size of the data sources. In this research, we apply a multi-agent approach to enable updating of a data store as changes are made at the sources. This approach reduces down time for the data store-a crucial requirement for Internet-based application. A fuzzy priority scheduling system is applied to prioritize tasks. Results from this research showed that the proposed multi-agent system drastically increases the availability of the data store while preserving a high level of data consistency.