az-hop
az-hop copied to clipboard
Azure
The Azure HPC On-Demand Platform provides an HPC Cluster Ready solution
Allow multiple queues to be defined for OpenPBS with options to cover : - priority - maximum allowed walltime - maximum queue time if possible
VM allocation and start is now taking less than 60 seconds, however node configuration can takes 3 mn, This need to be tuned. Check the log files in /opt/cycle/jetpack/logs for...
Add a dashboard to monitor the health of all key services like job scheduler, cyclecloud, apache httpd24, etc...
``` # Add I_MPI_HYDRA_BOOTSTRAP=ssh in the IMPI module otherwise it will break how PBS launch jobs if ! grep I_MPI_HYDRA_BOOTSTRAP /usr/share/modules/modulefiles/mpi/impi-2021 ; then echo "setenv I_MPI_HYDRA_BOOTSTRAP ssh" >> /usr/share/modules/modulefiles/mpi/impi-2021 fi...
on hardened system audit is configure but log files are rotating and never purged. /var/log/audit is filled by lots of audit logs. https://joelitechlife.ca/2022/02/17/redhat-linux-audit-log-rotation/
This is a warning showing up on all PBS jobs. Some explanations from here https://openpbs.atlassian.net/wiki/spaces/PD/pages/51901651/PP-516+Direct+write+of+the+job+s+stdout+err+files.
## In what area(s)? > /area administration /area ansible > /area autoscaling > /area configuration > /area cyclecloud > /area documentation > /area image > /area job-scheduling > /area monitoring...
