From: yudaqi (dqyu_at_ctb.pku.edu.cn)
Date: Fri Dec 18 2009 - 21:02:50 CST
Hi all,
I want to use Amber force field (ff99SB and ff03) in NAMD 2.7b2. I have
compiled the multicore and MPI (MPICH1, MPICH2, and OPENMPI) version,
and downloaded the TCP version for NAMD 2.7b2. I have tried two systems
with 24316 and 48456 atoms respectively. The smaller system with 24316
atoms runs ok for all versions. Unfortunately, the bigger system with
48458 atoms often crashed with memory issues for MPI and TCP versions,
although it works for multicore version. The error does not happen every
time, but quite frequently. After I adjusted the maximum locked memory
in the system, from 32 to unlimited, the error happens rarely in TCP
version. However, the failure remains in the MPI version. I guess that
there are some memory problems in the communication part of NAMD 2.7b2.
Would you please help me to find out the problems and fix them?
The TCP version returns the following errors:
------------- Processor 0 Exiting: Caught Signal ------------
Signal: segmentation violation
Suggestion: Try running with '++debug', or linking with '-memory paranoid'.
Fatal error on PE 0> segmentation violation
The MPI version returns:
------------- Processor 0 Exiting: Caught Signal ------------
Signal: 11
[0] MPI Abort by user Aborting program !
[0] Aborting program!
Killed by signal 2.
Killed by signal 2.
Killed by signal 2.
The scripts used are similar to the following pastes:
amber yes
parmfile ub_solvated.prmtop
ambercoor ub_solvated.inpcrd
readexclusions yes
exclude scaled1-4
1-4scaling 0.833333
scnb 2.0
switching off
cutoff 8.0
PME on
set temperature 300
set outputname mined
set restartname mined.restart
wrapAll on
cellBasisVector1 67.569747 0 0
cellBasisVector2 0 67.569747 0
cellBasisVector3 0 0 67.569747
cellOrigin 33.7848735 33.7848735 33.7848735
PMEGridSizeX 64
PMEGridSizeY 64
PMEGridSizeZ 64
rigidBonds all
rigidTolerance 0.00001
temperature $temperature
seed 12345
timestep 1.0
outputname $outputname
restartname $restartname
outputEnergies 20
restartfreq 500
DCDfreq 500
binaryoutput yes
binaryrestart yes
outputTiming 100
minimize 10000
Thank you very much!
Yours sincerely
Daqi Yu
Peking University
This archive was generated by hypermail 2.1.6 : Wed Feb 29 2012 - 15:53:37 CST