Method of commitment in a distributed database transaction

作者: Gerald K. Bortvedt , Robert H. Gerber

DOI:

关键词:

摘要: A method for committing a distributed transaction in database system. The system includes an interval coordinator, plurality of server programs, called coservers, and at least one log. More than coserver can operate on single computer or node, the coservers could share coordinator sends each succession messages, flushes its associated log to non-volatile storage response. After flushing log, transmits closure message coordinator. maintain state which identifies most recently received message. Each owner non-owner, helper. For transaction, request helper identifying operation execute. Upon execution operation, completion with tag receiving said message, eligibility Then writes commit stable storage. transaction.

参考文章(31)
Nathan Goodman, Philip A. Bernstein, Vassco Hadzilacos, Concurrency Control and Recovery in Database Systems ,(1987)
Robert J. Grafe, Melvin R. Zimowski, Bruce G. Lindsay, Patricia G. Selinger, John G. Adair, Robert L. Sunday, Richard R. Sanders, Dusan Ecimovic, Robert D. Jackson, Michael E. Murphy, Richard A. Demers, Roger A. Reinsch, Robert P. Resch, Query language execution on heterogeneous database servers using a bind-file bridge between application and database languages ,(1990)
Robert W. Griffin, James P. Emmond, Exactly-once semantics in a tp queuing system ,(1989)
Yuri J. Breitbart, Glenn R. Thompson, Method for consistent multidatabase transaction processing ,(1987)
Mary E. Vendryes, Edmond August Pruul, Dietrich Oliver Wood, Mark Zielenski, Coordinated handling of error codes and information describing errors in a commit procedure ,(1990)