Daniel Moldt, Jan Henrik Röwekamp, and Michael Simon.
A simple prototype of distributed execution of reference nets based
on virtual machines.
In Bergenthum and Kindler (éd.), Algorithms and Tools for Petri Nets Proceedings of the Workshop AWPN 2017, Kgs. Lyngby, Denmark October 19–20, 2017, pages 51-57.
Abrégé: Fast execution of Petri nets is one goal of Petri net tool providers. For complex models with a large set of concurrent simulation tasks in each simulation engine there are multiple possibilities for a speed up. Beside parallel Petri net execution within one simulation engine distributed execution is another option. In our research group a proposal for the distributed simulation of Reference Nets has been made. The approach connects multiple simulations over a network, but is agnostic to the environment in which these are executed. In this contribution the distributed simulation's scalability by the usage of virtual machines is shown. We use an algorithm to calculate prime numbers to illustrate the performance changes. The costs of communication and synchronization for complex bindings are discussed to find a good trade-off for the decision whether to distribute an execution or not.
@InProceedings{Moldt+17, author = {Moldt, Daniel and R{\"o}wekamp, Jan Henrik and Simon, Michael}, title = {A Simple Prototype of Distributed Execution of Reference Nets Based on Virtual Machines}, crossref = {Bergenthum+-e-17}, pages = {51--57}, keywords = {Petri Nets, Reference Nets, Distributed Simulation, Virtual Machines, Synchronous Channels, Renew}, abstract = {Fast execution of Petri nets is one goal of Petri net tool providers. For complex models with a large set of concurrent simulation tasks in each simulation engine there are multiple possibilities for a speed up. Beside parallel Petri net execution within one simulation engine distributed execution is another option. In our research group a proposal for the distributed simulation of Reference Nets has been made. The approach connects multiple simulations over a network, but is agnostic to the environment in which these are executed. In this contribution the distributed simulation's scalability by the usage of virtual machines is shown. We use an algorithm to calculate prime numbers to illustrate the performance changes. The costs of communication and synchronization for complex bindings are discussed to find a good trade-off for the decision whether to distribute an execution or not.}, year = 2017 } @Proceedings{Bergenthum+-e-17, editor = {Bergenthum, Robin and Kindler, Ekkart}, editorplain = {Bergenthum, Robin and Kindler, Ekkart}, title = {{Algorithms and Tools for Petri Nets Proceedings of the Workshop AWPN 2017, Kgs. Lyngby, Denmark October 19–20, 2017}}, booktitle = {{Algorithms and Tools for Petri Nets Proceedings of the Workshop AWPN 2017, Kgs. Lyngby, Denmark October 19–20, 2017}}, publisher = {DTU Compute Technical Report 2017-06}, year = 2017, ISSN = {1601-2321}, annote = {\url{http:///AWPN\_2011.shtml}} }
This material is presented to ensure timely dissemination of scholarly and technical work. Copyright and all rights therein are retained by authors or by other copyright holders. All persons copying this information are expected to adhere to the terms and constraints invoked by each author's copyright. In most cases, these works may not be reposted without the explicit permission of the copyright holder.