A proposed algorithm for solving deadlock detection in distributed database systems

Deadlock is one of the most serious problems in multitasking concurrent programming systems. The deadlock problem becomes further complicated when the underlying system is distributed and when tasks have timing constraints. Distributed deadlock detection has been studied to some extent in distributed database systems and distributed timesharing operating systems but has not been widely used in real-time systems. This paper introduces a brief overview of the most recent algorithm for deadlock detection and resolution. The main objective of this paper is to investigate an integrated algorithm for detecting and resolving the deadlock within the distributed database system (DDBS) environment. Examining how to represent the deadlock problem and translation from direct graph into a matrix to improve the deadlock process is discussed. A new phase for resolving the deadlock is introduced. Integration between two phases is also introduced.