My Molecular Dynamics Simulation (93 Mio. particles) fails after approx. 222900 timesteps with the ERROR 'Cannot open restart file 'filename' (.../write_restart.cpp:272). According to the LAMMPS documentation this error is self-explanatory, but I do not understand why this occurs after so many timesteps. I write out these .restart-files every 1000 timesteps, toggling between two filenames and apparently this stops working.
I use LAMMMPS version 3Mar20 on a Linux Cluster, parallelized on 960 cores.
I did test the script in a scaled down scenario (all dimensions reduced). For this, everything worked just as expected.