[gmx-users] Problem with MPIRUN and Ewald

[Matteo Busato]

Good morning to everyone,


I am trying to perform a simple classical dynamic of a box of 500 TIP3P water molecules with an Ag+ ion in its center. However, when running the dynamic with 8 physical cores but no multithread with mpirun it crashes and the output of the job reports:


"The number of PME grid lines per rank along x is 3. But when using OpenMP threads, the number of grid lines per rank along x should be > = pme_order (6) or pmeorder-1. To resolve this issue, use fewer ranks along x (and possibly more along y and/or z) by specifyng -dd manually"


The md.log file says instead:


"Fatal error: the size of the domain decomposition grid (125) does not match the number of ranks (8). The total number of ranks is 8".


The strange fact is that it works on supercomputer CINECA resources but it doesn't on our local cluster.


This is the .mdp file I'm using:


; Run control
integrator               = sd       ; Langevin dynamics
tinit                    = 0
dt                       = 0.002
nsteps                   = 500000   ; 1 ns
nstcomm                  = 100
; Output control
nstxout                  = 500
nstvout                  = 500
nstfout                  = 0
nstlog                   = 500
nstenergy                = 500
nstxout-compressed       = 0
; Neighborsearching and short-range nonbonded interactions
cutoff-scheme            = verlet
nstlist                  = 20
ns_type                  = grid
pbc                      = xyz
rlist                    = 1.0
; Electrostatics
coulombtype              = PME
rcoulomb                 = 1.0
; van der Waals
vdwtype                  = cutoff
vdw-modifier             = potential-switch
rvdw-switch              = 0.9
rvdw                     = 1.0
; Apply long range dispersion corrections for Energy and Pressure
DispCorr                  = EnerPres
; Spacing for the PME/PPPM FFT grid
fourierspacing           = 0.12
; EWALD/PME/PPPM parameters
pme_order                = 6
ewald_rtol               = 1e-06
epsilon_surface          = 0
; Temperature coupling
; tcoupl is implicitly handled by the sd integrator
tc_grps                  = system
tau_t                    = 1.0
ref_t                    = 300
; Pressure coupling is on for NPT
Pcoupl                   = Parrinello-Rahman
tau_p                    = 1.0
compressibility          = 4.5e-05
ref_p                    = 1.0
; Free energy control stuff
free_energy              = yes
init_lambda_state        = 0
delta_lambda             = 0
calc_lambda_neighbors    = 1        ; only immediate neighboring windows
; Vectors of lambda specified here
; Each combination is an index that is retrieved from init_lambda_state for each simulation
; init_lambda_state        0    1    2    3    4    5    6    7    8    9    10   11   12   13   14   15   16   17   18   19   20
vdw_lambdas              = 0.00 0.10 0.20 0.30 0.40 0.50 0.60 0.70 0.80 0.90 1.00 1.00 1.00 1.00 1.00 1.00 1.00 1.00 1.00 1.00 1.00
coul_lambdas             = 0.00 0.00 0.00 0.00 0.00 0.00 0.00 0.00 0.00 0.00 0.00 0.10 0.20 0.30 0.40 0.50 0.60 0.70 0.80 0.90 1.00
; We are not transforming any bonded or restrained interactions
bonded_lambdas           = 0.00 0.00 0.00 0.00 0.00 0.00 0.00 0.00 0.00 0.00 0.00 0.00 0.00 0.00 0.00 0.00 0.00 0.00 0.00 0.00 0.00
restraint_lambdas        = 0.00 0.00 0.00 0.00 0.00 0.00 0.00 0.00 0.00 0.00 0.00 0.00 0.00 0.00 0.00 0.00 0.00 0.00 0.00 0.00 0.00
; Masses are not changing (particle identities are the same at lambda = 0 and lambda = 1)
mass_lambdas             = 0.00 0.00 0.00 0.00 0.00 0.00 0.00 0.00 0.00 0.00 0.00 0.00 0.00 0.00 0.00 0.00 0.00 0.00 0.00 0.00 0.00
; Not doing simulated temperting here
temperature_lambdas      = 0.00 0.00 0.00 0.00 0.00 0.00 0.00 0.00 0.00 0.00 0.00 0.00 0.00 0.00 0.00 0.00 0.00 0.00 0.00 0.00 0.00
; Options for the decoupling
sc-alpha                 = 0.5
sc-coul                  = no       ; linear interpolation of Coulomb (none in this case)
sc-power                 = 1.0
sc-sigma                 = 0.3
couple-moltype           = AG       ; name of moleculetype to decouple
couple-lambda0           = none     ; no interactions at lambda=0
couple-lambda1           = vdw-q    ; turn on both vdW and Coulomb
couple-intramol          = no
nstdhdl                  = 10
; Do not generate velocities
gen_vel                  = no
; options for bonds
constraints              = h-bonds  ; we only have C-H bonds here
; Type of constraint algorithm
constraint-algorithm     = lincs
; Constrain the starting configuration
; since we are continuing from NPT
continuation             = yes
; Highest order in the expansion of the constraint coupling matrix
lincs-order              = 12


Thank you in advance for your suggestions.

Matteo Busato