ИСТИНА

In this presentation, we will describe our approaches for performing Operational Data Analytics (ODA) using the example of the Supercomputing Center in Lomonosov Moscow State University. The goal of ODA is to allow for the continuous monitoring, storing, and analysis of the information about different aspects of HPC system behavior. We will focus mainly on the topic of performance analysis and show our ODA tools that have been developed and are being used in MSU on an ongoing basis, as well as describe subtleties that need to be taken into account while performing ODA, technical difficulties arising and ways for solving them, and useful insights that can be obtained.