On the performance of distributed lock-based synchronization

Abstract

We study the relation between two classical types of distributed locking mechanisms, called token-based locking and permission-based locking, and several distributed data structures which use locking for synchronization. We have proposed, implemented and tested several lock-based distributed data structures, namely, two different types of counters called find&increment and increment&publish, a queue, a stack and a linked list. For each one of them we have determined what is the preferred type of lock to be used as the underling locking mechanism. Furthermore, we have determined which one of the two proposed counters is better to be used either as a stand-alone data structure or when used as a building block for implementing other high level data structures. © 2011 Springer-Verlag Berlin Heidelberg.