Fermi: jobs failure due to the reboot of some I/O nodes

Dear Users,

due to the current problems on the scratch filesystem, the I/O nodes on 4 of the 10 racks of FERMI appear to be in an error state and need to be immediately rebooted to guarantee their proper functioning. In order to perform the reboot, the jobs now running on those racks will be killed. We invite the users to check the status of their running jobs and to resubmit them in case they were affected by this reset.

We apologize for the inconvenience,

HPC User Support @ CINECA