Showing papers by "Johann-Christoph Freytag published in 1995"

An annotated bibliography on active databases (short version)

Abstract: This is a bibliography on active databases and active database systems which reflects the various research activities in this field. We compiled this bibliography for our own use, but hopefully it might be useful to other people as well. All papers that appear in the following list, are generally accessible.We do not claim that the bibliography is exhaustive and covers the complete range of literature that deals with activities. We decided to focus on central approaches, concepts, methods, and systems in the area of active databases. It does not contain entries in the area of "pure" real-time, object-oriented, temporal, and deductive databases. But we did include publications related to those approaches, as long as they discuss active databases.We divided the material into various sections following our own personal perception of the field. The sections provide an overview on different projects in the area of active databases, followed by sections on relevant research topics. Each section contains a few remarks followed by a list of cross references into the annotated bibliography. Papers might appear in more than one section in case they discuss different topics relevant to different sections.Additionally, when relevant we also included unpublished, but publicly available material. For those papers we included information how to obtain them from the authors or from the organizations where the were produced.The beauty of our work is the individual annotation to almost all publications. Due to space limitations we are forced to leave out those annotations in the version published here. For a complete annotated bibliography we refer to the entry in our WWW server.The effort to build up such a bibliography is an endless task. Since we believe it now provides a comprehensive overview on the existing literature in the field, we decided to publish it. However, we invite all readers to add remarks, corrections, updates, additions (including further annotations).Part of this work was done while we were associated with the FORWISS Institute of the Technical University of Munich. We would like to thank our student, Markus Blaschka, who compiled many references during his master's thesis.

Modelling the Dynamic Evolution of System Workload During Pipelined Query Execution

Abstract: Database parallelism increases the complexity of query optimization. In particular, the query optimizer must take into account the impact of processor and network workload on competing and communicating processes. During query execution this additional overhead becomes even more important when exploiting pipelined parallelism, since the workload evolves over time dynamically as processes start and complete execution. We propose a model which incorporates the eeect of the system workload on the execution of parallel and pipelined queries. We extend the conventional notion of pipeline in two directions: First, we include pipes with multiple producer nodes into our model. Second, we consider the latency between the start time of the producing process and the start time of the consuming process. Based on this execution model, we derive a cost model that incorporates system workload. In particular, we calculate the eeects of pipelining on processor multitasking and the eeects of multitasking on the total execution time of the query. Additionally, we model the load on the network channels used to connect communicating processors, taking the fact into account that those channels may be changed dynamically by the routing mechanism. We consider the communication delays caused by this load and estimate their impact on pipelined execution. Finally, the model integrates the cost of remote disk accesses to the computation of system workload. It is general enough to cover query processing for both shared-nothing and shared-disk architectures.