/cvmfs not accessible on one of the compute nodes

HPC Support HPC issues
1

Primary informations

Username: coppinp
Cluster: Baobab

Description

Jobs fail on cpu350 with the message:
cd: /cvmfs/dampe.cern.ch/centos7/opt//releases/DmpSoftware-6-0-18_geant4_10_06p03/: No such file or directory

Steps to Reproduce

#!/bin/bash
REPO=/cvmfs/dampe.cern.ch
NODE=$(hostname -s)
if [ ! -d "$REPO" ]; then
    echo "$NODE NO_DIR"
elif timeout 20 ls "$REPO" >/dev/null 2>&1; then
    echo "$NODE OK"
else
    echo "$NODE FAIL_OR_TIMEOUT"
fi

Expected Result

I checked a few other nodes in the private-dpnc-cpu queue, which all seem fine. Just cpu350 return the output: cpu350 NO_DIR