In this paper, we investigate the issues of performance management and Quality of Service in large scale, autonomous and federated Digital Library systems consisting of many hundreds of servers performing the tasks of indexing, searching and retrieval of information objects. Our contributions are the following: we present a distributed performance management framework for measurement, monitoring and QoS based searching in Digital Library Systems.We designed and implemented our performance monitoring framework and measurement mechanisms over a testbed of NCSTRL based Dienst servers. We de ned relevant parameters for investigating the performance of the servers. Using these parameters and the framework, we developed novel load balancing algorithms for distributed searching over a network of Index servers. In addition, we propose mechanisms for dynamic timeout adaptation for the distributed searches. We studied the performance of one of the load balancing algorithms via simulations.