As the number of resources on the grid tends to be very dynamic and potentially large, it becomes increasingly important to discover and monitor them efficiently. This paper concentrates on the scalability, reliability and performance of grid discover and monitoring services, using autonomous concepts. Application scenarios are developed where the notification rate of data is dynamically modified, based on the overhead costs at the MDS3 Index Service. Previously collected performance benchmarks are also utilised to implement two self-adaptive algorithms which are the basis of the scenarios. The objective of these feedback algorithms is to sustain an adequate level of service for clients while minimising costs at the MDS. Different types of workload models are also used to assess the efficiency of the algorithms. Experimental results are subsequently shown when varying notification update mechanisms and decision parameters are used, ensuring that the MDS is scalable with accruing concurrent clients. Therefore, a policy is proposed where the notification rate is computed dynamically, thereby creating an autonomous grid monitoring service.
30 October 2018 in Electronic Notes in Theoretical Computer Science
In this paper we present the outline of a novel electrostatic, second order Particle-in-Cell (PIC) algorithm, that makes use of 'ghost particles' located around true particle positions in order to represent a charge distribution. We implement our algorithm within EMPIRE-PIC, a PIC code developed at Sandia National Laboratories. We test the performance of our algorithm on a variety of many-core architectures including NVIDIA GPUs, conventional CPUs, and Intel's Knights Landing. Our preliminary results show the viability of second order methods for PIC applications on these architectures when compared to previous generations of many-core hardware. Specifically, we see an order of magnitude improvement in performance for second order methods between the Tesla K20 and Tesla P100 GPU devices, despite only a 4× improvement in the theoretical peak performance between the devices. Although these initial results show a large increase in runtime over first order methods, we hope to be able to show improved scaling behaviour and increased simulation accuracy in the future.