A status management scheme of MMDBMS for consistency of duplicated database

In the Internet, although the processing speed of a system is regarded to be important, the problem of a break in service due to a fault of the system, even in expensive and high-performance systems, may be a fatal effect. However, a clustering structure, which is complicated and very expensive, cannot be a clear and satisfactory solution. Therefore, we propose a fault-tolerant duplicated structure that can offer high reliability, high availability, and low expense while maintaining high performance with respect to Tachyon. Tachyon is the memory resident realtime DBMS used for supporting the operation and maintenance function in the ATM switching system required for high reliability, high availability, and low expense. Also proposed is a dual processing method through DBMS status management for reliable control of data consistency between two systems when operating the dual structure. The proposed dual processing method for the Tachyon DBMS offers the attribute that it can perform reliable control of database consistency and coherence in the unit of a transaction, in spite of the frequent exchange of status control between the two systems, by dividing and managing two DBMS with status duplicated and operated to offer service continuity between the two systems up to six states.