Shrinking the warehouse update Window
SIGMOD '99 Proceedings of the 1999 ACM SIGMOD international conference on Management of data
| Hi-index | 0.00 |
As data warehouses are presently designed mostly as centralized system, the majority of the research in this area has concentrated on developing updating algorithms for this specific model. Aproaches have been proposed either under the assumption of a single view data warehouse, a multi-view centralized system, or a multi-view distributed system with strict synchronization restrictions. However, we project that a distributed, decentralized system will soon prove to be a good model for the generalization of the data warehouse system. To our knowledge, there is no proposal for algorithms in this context, to incorporate individually and asynchronously updates to different views. Existing protocols either solve the synchronicity problem by grouping out of order updates in batches or by creating multiple versions of the data. In this paper we propose a different method for updating views in a distributed data warehouse, and address issues related to a decentralized system were the sytes are autonomous. Although the procedure of updating views becomes more comples, it does not involve any additional messages with respect to a method that synchronizes communication through the system, and some additional storage is needed only for the control structures.