<p>28/09/2021<br />
Dear Users,</p>
<p>the maintenance operations on MARCONI100 have been completed and the cluster<br />
is now back in production.</p>
<p>During the maintenance, Slurm scheduler has been updated to version 21.08.1 .<br />
Unfortunately, this caused an issue that we are investigating and that may<br />
have affected your jobs. In case of failure with an error log about GPU<br />
device unavailability, as a workaround please change the directive<br />
"--gres=gpu:xx" with "--gpus-per-node=xx" in your jobscript.</p>
<p>We apologies for the inconvenience and we will let you know as soon as the<br />
problem is solved.</p>
<p>Regards,<br />
Alessandro</p>