User- and job-centric monitoring: Analysing and presenting large amounts of monitoring data
Henrik Eichenhardt, R. Müller-Pfefferkorn, Reinhard Neumann, Thomas William · 2008
For data analysis or simulations (e.g. in particle physics) single users submit hundreds or thousands of jobs to the Grid. This puts a new burden on the users side - keeping an overview on the status and performance of the jobs in a distributed environment. In this paper, a user- and job-centric monitoring system is presented. It collects job and system specific information, analyses them and presents them graphically to the user. In the analysis of the monitoring data job problems are classified to give hints to the user. The graphical and interactive presentation allows the user to easily keep an overview as well as to dig into more detailed monitoring data.