[gmx-users] Simulation Terminated on Clusters Due to "unable to write checkpoint file"

[Ray Chao]

D
​ear gmx users,

I am running my cases on the cluster. The simulation was always terminated
after running for a while. The error was always "Unable to write checkpoint
file: maybe you are out of disk space". However, I have 1TB of disk
allocation, which is enough,  and the disk is in good condition. Besides, I
created the .top .gro .mdp files on my own workstation and also changed the
permission from "read-only" to "read and write".  The gromacs version is
5.0.2 or 5.1.2 on the clusters.

Is there any specific requirement about running gromacs on clusters? Your
response is very much appreciated.​

You all have a good Valentine's day.