Coping with value dependency for failure recovery in multidatabase systems

Sun, Yongmei (1997) Coping with value dependency for failure recovery in multidatabase systems. Masters thesis, Memorial University of Newfoundland.

[img] [English] PDF - Accepted Version
Available under License - The author retains copyright ownership and moral rights in this thesis. Neither the thesis nor substantial extracts from it may be printed or otherwise reproduced without the author's permission.

Download (2MB)

Abstract

Local autonomy is the main impediment to achieving failure atomicity in a multidatabase system since it allows a local database to unilaterally commit or abort a subtransaction. Compensating a committed subtransaction is in general hard to realize due to the complication arising from the propagation of the committed effects. Resubmitting an aborted subtransaction is more realistic since the problems arising from inter-subtransaction dependencies are more predictable than those from propagation of committed effects. However, if such a dependency is cyclic or if it not only involves values but also data items, then the problem becomes more complicated. In this thesis, a failure recovery scheme¹ using resubmission is proposed. The scheme is based on distinguishing the subtransactions into two different types, and employing different strategies for them. As a result, the scheme allows an aborted subtransaction to be restarted. Compared with other failure recovery schemes which also do not rely on compensation, the scheme compromises local autonomy to a lesser extent. In this thesis, different kinds of dependencies are also studied, their impact on the correctness of resubmission method discussed and solutions proposed.

Item Type: Thesis (Masters)
URI: http://research.library.mun.ca/id/eprint/9429
Item ID: 9429
Additional Information: Bibliography: leaves 78-83.
Department(s): Science, Faculty of > Computer Science
Date: 1997
Date Type: Submission
Library of Congress Subject Heading: Computer system failures; Distributed databases; Electronic data processing--Distributed processing

Actions (login required)

View Item View Item

Downloads

Downloads per month over the past year

View more statistics