Static and Dynamic Data Management in Networks

We survey strategies for distributing shared objects in large parallel and distributed systems. Examples of such objects are global variables in a parallel program, pages or cache lines in a virtual shared memory system, shared les in a distributed le system, and videos and pictures in a distributed multimedia server. We focus on strategies for distributing, accessing, and (consistently) updating such objects. The strategies are provably eecient with respect to various cost measures. We describe and analyse static, hashing based schemes that minimize the contention at the memory modules in worst case scenarios. Especially, the beneet of redundant placement schemes is discussed. We further take network congestion and bandwidth into account. Here we present schemes that are provably eecient w.r.t. information about access frequencies. Further, dynamic schemes are presented which have good competitive ratio, i.e., are eecient compared to an optimal dynamic distribution that is constructed using full knowledge of the dynamic access pattern.