Dr David Green, QCIF’s eResearch Analyst at The University of Queensland, has some tips for those using more than one high-performance computer via QRIScloud.
There are many users who have access to more than one high-performance computer within QRIScloud. These HPCs include Awoonga and FlashLite (and for UQ staff and students, Tinaroo).
A wrapper script for the qstat command was deployed earlier this year that combines status information from all clusters and reports it to you. This is useful for people who have jobs running across the clusters.
The qstat command (i.e. /usr/local/bin/qstat) produces a comprehensive list of all your jobs across all three clusters.
If any HPC cluster is not responding then it is skipped over in the output.
If you just want to know what is happening on a specific server — e.g. the new Portable Batch System (PBS) file on Tinaroo — you should use the “bare” command: /opt/pbs/bin/qstat @tinmgr2
The server names are awongmgmr1, awonmgr2, flm1, flashmgr2, tinmgmr1 and tinmgr2.
The “1’s” run the older version of PBS Pro and are being decommissioned soon. The “2’s” are the newer version and will soon be the default servers for all HPC nodes.
Any queries, please contact: support@qriscloud.org.au.
This article was first published on 17/04/2019.