Strange Behavior of VASP with MPICH-G2

Questions regarding the compilation of VASP on various platforms: hardware, compilers and libraries, etc.


Moderators: Global Moderator, Moderator

Post Reply
Message
Author
swan2925
Newbie
Newbie
Posts: 7
Joined: Tue May 31, 2005 3:07 am
License Nr.: NOT REGISTERED

Strange Behavior of VASP with MPICH-G2

#1 Post by swan2925 » Fri Jul 01, 2005 2:50 am

I am using PGF compiler and MPICH-G2.
VASP can be built with it, and the VASP can run too.

however, its behavior is quite strange. normally a vasp run will produce 3 outputs, one is print to stdout, and the other two are files OSZICAR and OUTCAR.

In the current build, the vasp only produces OSZICAR and OUTCAR file but nothing goes to stdout. While the OSZICAR and OUTCAR file seem to be the main output files, i dunno if it matters to get nothing from stdout.

I would like to ask, is this behavior normal? :(
Last edited by swan2925 on Fri Jul 01, 2005 2:50 am, edited 1 time in total.

admin
Administrator
Administrator
Posts: 2921
Joined: Tue Aug 03, 2004 8:18 am
License Nr.: 458

Strange Behavior of VASP with MPICH-G2

#2 Post by admin » Fri Jul 01, 2005 8:48 am

no, but this is not a VASP error. stdout is written for sure, the error is either
in your job script or in the MPICH installation.
Last edited by admin on Fri Jul 01, 2005 8:48 am, edited 1 time in total.

swan2925
Newbie
Newbie
Posts: 7
Joined: Tue May 31, 2005 3:07 am
License Nr.: NOT REGISTERED

Strange Behavior of VASP with MPICH-G2

#3 Post by swan2925 » Fri Jul 01, 2005 9:55 am

Thanks for the admin reply first!!

An update to my situation.

I found that after two hours, the VASP process died and write the STDOUT with: (the STDOUT isn't write immediately, but a long time after my execution.)

running on 1 nodes
each image running on 1 nodes
distr: one band on 1 nodes, 1 groups
vasp.4.6.6 17Mar03 complex
01/POSCAR found : 2 types and 112 ions
LDA part: xc-table for Ceperly-Alder, standard interpolation
00/POSCAR found : 2 types and 112 ions
02/POSCAR found : 2 types and 112 ions
POSCAR, INCAR and KPOINTS ok, starting setup
WARNING: random wavefunctions but no delay for mixing, default for NELMDL
entering main loop
N E dE d eps ncg rms rms(c)
RMM: 1 0.233067147194E+04 0.23307E+04 -0.57989E+04 233 0.500E+02
RMM: 2 0.518796907502E+03 -0.18119E+04 -0.20512E+04 233 0.136E+02
RMM: 3 -0.124820787106E+03 -0.64362E+03 -0.75070E+03 233 0.923E+01
RMM: 4 -0.406320164275E+03 -0.28150E+03 -0.24987E+03 233 0.640E+01
RMM: 5 -0.505258096910E+03 -0.98938E+02 -0.88900E+02 233 0.380E+01
RMM: 6 -0.547644628146E+03 -0.42387E+02 -0.36483E+02 233 0.251E+01
RMM: 7 -0.565223159093E+03 -0.17579E+02 -0.16433E+02 233 0.162E+01
RMM: 8 -0.573558439154E+03 -0.83353E+01 -0.77003E+01 233 0.113E+01
RMM: 9 -0.581337024918E+03 -0.77786E+01 -0.76466E+01 568 0.754E+00
RMM: 10 -0.581532041125E+03 -0.19502E+00 -0.31986E+00 626 0.170E+00
RMM: 11 -0.581550839222E+03 -0.18798E-01 -0.16174E-01 452 0.236E-01
RMM: 12 -0.581551729438E+03 -0.89022E-03 -0.86315E-03 346 0.722E-02 0.266E+01
RMM: 13 -0.557108052554E+03 0.24444E+02 -0.41863E+01 466 0.701E+00 0.580E+00
RMM: 14 -0.555554885937E+03 0.15532E+01 -0.67386E+00 468 0.287E+00 0.146E+00
RMM: 15 -0.555368151510E+03 0.18673E+00 -0.75750E-01 490 0.941E-01 0.633E-01
RMM: 16 -0.555301393502E+03 0.66758E-01 -0.15882E-01 477 0.420E-01 0.373E-01
RMM: 17 -0.555273152835E+03 0.28241E-01 -0.11832E-01 470 0.321E-01 0.198E-01
RMM: 18 -0.555280469276E+03 -0.73164E-02 -0.40410E-02 455 0.186E-01 0.128E-01
RMM: 19 -0.555286112792E+03 -0.56435E-02 -0.12094E-02 411 0.112E-01 0.823E-02
RMM: 20 -0.555299509301E+03 -0.13397E-01 -0.73399E-03 376 0.849E-02 0.479E-02
RMM: 21 -0.555314094720E+03 -0.14585E-01 -0.24868E-03 310 0.457E-02 0.270E-02
RMM: 22 -0.555323862449E+03 -0.97677E-02 -0.19439E-03 296 0.371E-02 0.132E-02
RMM: 23 -0.555325990698E+03 -0.21282E-02 -0.37713E-04 199 0.125E-02 0.559E-03
RMM: 24 -0.555326700037E+03 -0.70934E-03 -0.15842E-04 170 0.736E-03
1 F= -.55532670E+03 E0= -.55532670E+03 d E =-.555327E+03
curvature: 0.00 expect dE= NaN dE for cont linesearch NaN
trial: gam= 0.00000 g(F)= NaN g(S)= NaN ort = NaN (trialstep = NaN)
search vector abs. value= NaN
bond charge predicted
REAL_OPTLAY: internal error (1) 0 800

Finally, I get a NaN. I am using PGI Fortran compiler!!
Is there anything goes wrong?
Or is it possible that it maybe due to my wrong input files?

I am looking forward to replies, thanks!!
Last edited by swan2925 on Fri Jul 01, 2005 9:55 am, edited 1 time in total.

Post Reply