I have a .gro configuration which ran perfectly fine previously. However, I don’t know what I have changed in the .mdp file and I can never make the simulation run again, it returns
" Fatal error:There is no domain decomposition for 6 ranks that is compatible with the given
box and a minimum cell size of 17.3445 nm"
It was previously run with the same mpi setting.
I have spent hours on figuring out what went wrong. I wonder if there is a way to recover the .mdp file from the cpt, log, or the tpr file?
Most mdp setting are given at the beginning of the md.log file. Do a diff of your working and not working md.log files. That should show what settings differ.
Note that 17 nm as a minimum cell size requirement is enormous. My guess is that there is a PBC issue and atoms that should be close together ended up in different periodic images and this is somehow not recognized.
Thank you hess for your response. I have figured out what went wrong. Like you said, it is a pbc problem. I have a membrane between two charmm-36 metal slabs , somehow I have to use pbc-molecules instead of pbc for this configuration. pbc works fine with just the membrane. I am guessing the edges of the metal slabs have to be bonded with its image.