Several tools can be used to monitor the usage of the UL HPC platform:
- Monika shows the current status of the clusters and the submitted jobs, also allowing you to see only the subset of compute nodes that have specific properties (e.g. are GPU-enabled).
- Drawgantt shows a Gantt chart with the submitted jobs’ timeline between specific dates.
- Ganglia is the traditional HPC clusters monitor, displaying a breadth of metrics for each compute node (e.g. cpu/disk/network usage) and aggregated cluster-wide metrics.
- CDash is a software testing server, whose web-based dashboard shows the tests of the MPI stacks we provide through Environment Modules.
The Grid’5000 status page must be used for the G5k cluster, selecting Luxembourg under the specific tool - Monika, Drawgantt, Ganglia.
Information on monitoring your submitted jobs from the cluster can be found in this section of the OAR page.