Architecture and Interface of Scalable Distributed Database System SD-SQL Server

We present a scalable distributed database system SD-SQL Server. Its original feature is the scalable distributed partitioning of its relational tables. The system dynamically distributes the tables into segments created each at a different SD-SQL Server node. The partitioning is transparent to the applications. New segments result from splits following overflowing inserts. SD SQL Server avoids the periodic and cumbersome manual reorganizing of scaling tables, characteristic of the current DBMS technology. With the comfort of a single node SQL Server database user, the SD-SQL Server user may dispose of many times larger tables. We present the architecture of our system, and its user/application interface. Related work discusses our implementation and shows that the overhead of our scalable distributed table management should be typically negligible.

• Type: Conference
• Year: 2006
• Where: DBA
• Authors: Witold Litwin, Soror Sahri, Thomas J. E. Schwarz
• Keywords: Database, DBA 2006, Scalable Distributed Database, Scalable Distributed Partitioning, SD-SQL Server