cannot restart from restart file
Csilla Varnai
cv... at cam.ac.uk
Fri Jul 18 16:46:09 UTC 2008
Dear All,
Unfortunately it's happened recently quite often that my jobs are
killed due to last minute-announced updates etc. I'm trying to restart
it from the saved restart files, but I can't, because during the MD
simulation some atoms move too close to one another, and CP2K aborts
during the neighbour list building.
Here's a file I used to test if I could restart the calculation. I
performed 5 steps and then I tried to run the restart files to see if
I could get the same trajectory.
When I try to run any of the saved restart files, the program aborts
immediately (see error message below). I checked the geometry, and the
atoms are closer than the threshold. If I increase the EMAX_SPLINE,
the temperature goes up to > 100 000 K and aborts with basically the
same error message.
However, running the original input file (please find attached), I
don't even get warning messages, although the atoms get closer than
the threshold for the spline construction.
What's the difference in the 2 cases and why can't I obtain the same
trajectory? I would expect to be able to. Am I wrong?
I would really appreciate if anyone could help me!
Thanks very much,
Csilla
KINETIC ENERGY[hartree] = 0.522663555689E+02
0.541626244398E+02
TEMPERATURE[K] = 115820.281
120022.341
*******************************************************************************
WARNING| Particles: 58 30 at distance [au]: 1.20668842
less than: 1.36797301; increase EMAX_SPLINE.
*
*** ERROR in build_neighbor_lists ***
*
*** GEOMETRY wrong or EMAX_SPLINE too small! ***
===== Routine Calling Stack =====
7 build_neighbor_lists
6 build_fist_neighbor_lists
5 list_control
4 fist_force_control
3 velocity_verlet
2 qs_mol_dyn_low
1 CP2K
Csilla Varnai
cv... at cam.ac.uk
Engineering Department
University of Cambridge
Trumpington Street, CB2 1PZ
United Kingdom
-------------- next part --------------
A non-text attachment was scrubbed...
Name: restart-test.inp
Type: application/octet-stream
Size: 13704 bytes
Desc: not available
URL: <https://lists.cp2k.org/archives/cp2k-user/attachments/20080718/8d08ed81/attachment.obj>
-------------- next part --------------
More information about the CP2K-user
mailing list