Performance evaluation of measurement data acquisition mechanisms in a distributed computing environment integrating remote laboratory instrumentation

摘要

The usage of laboratory and measurement instrumentation of any kind, from large complex equipment to networks of sensors that collectively appear as a distributed measurement device, has become of relevant importance in all branches of experimental sciences. Owing to the increasing networking capacity and access ubiquity, this bulk of instrumentation is ever more frequently accessed remotely by users who want to perform experiments, collect and process experimental data, analyze and interpret results. With reference to a remote instrumentation architecture deeply rooted in distributed computing paradigms such as grids and clouds, we evaluate the performance of mechanisms for the collection of data generated by instruments, in order to assess the capabilities of remote instrumentation services. In the presence of instruments generating measurements at high rate, which must be delivered to a multiplicity of users, publish/subscribe dispatching (push) mechanisms are shown to outperform pull-based ones.