Leonardo: ongoing access and storage issues

Dear Users,

we are currently experiencing issues affecting access to Leonardo login nodes and the storage areas. The slurm controller was rebooted to align its cache to the SlurmDB and no job failures have been caused by this event. However, many jobs are currently in a pending state with reason “PartitionConfig”. This is a temporary condition and will be resolved as soon as the scheduler is able to reassess all queued jobs.

We are working to restore the cluster’s full functionality and we will provide further updates as soon as more information becomes available.

We apologize for the inconvenience,

HPC User Support – CINECA