Survey on Deadlocks in Distributed Database Systems

Abstract

In Distributed Database System (DBS) and multitasking system, the occurrence of deadlocks is one of the most serious problems. If a site request for a resource that is already in the another site which is waiting for another resource then the scenario is called as distributed deadlock. Different distributed environments require a suitable deadlock detection algorithm to detect deadlocks. Different distributed environments needs to maintain their platforms by avoiding deadlocks. To achieve this environment, it is required to fed with optimized deadlock detection and avoidance algorithms. In this article, different deadlock detection algorithms that uses Wait For Graph and resolution algorithms to trace out deadlocks were discussed. An optimization technique is used for resolving deadlock in an efficient manner. A comparison between different deadlock detection algorithms based on different parameters like, delay time, message size, number of messages and whether the algorithm detects false deadlocks or not were performed. Based on the comparisons, a few deadlock detection algorithms were suggested for the distributed environment.